'Math on Trial' - Leila Schneps, 'Thoughtful' on PMF

'Math on Trial' - Leila Schneps, 'Thoughtful' on PMF

Postby Sarah » Wed Mar 06, 2013 3:57 pm

Math on Trial: How Numbers Get Used and Abused in the Courtroom
http://www.amazon.com/Math-Trial-Numbers-Abused-Courtroom/dp/0465032923

Maths on Trial - Blob
http://mathsontrial.blogspot.co.at/2011/05/welcome-to-maths-on-trial.html

"Publication Date: March 12, 2013
In the wrong hands, math can be deadly. Even the simplest numbers can become powerful forces when manipulated by politicians or the media, but in the case of the law, your liberty—and your life—can depend on the right calculation.

In Math on Trial, mathematicians Leila Schneps and Coralie Colmez describe ten trials spanning from the nineteenth century to today, in which mathematical arguments were used—and disastrously misused—as evidence. They tell the stories of Sally Clark, who was accused of murdering her children by a doctor with a faulty sense of calculation; of nineteenth-century tycoon Hetty Green, whose dispute over her aunt’s will became a signal case in the forensic use of mathematics; and of the case of Amanda Knox, in which a judge’s misunderstanding of probability led him to discount critical evidence—which might have kept her in jail. Offering a fresh angle on cases from the nineteenth-century Dreyfus affair to the murder trial of Dutch nurse Lucia de Berk, Schneps and Colmez show how the improper application of mathematical concepts can mean the difference between walking free and life in prison.

A colorful narrative of mathematical abuse, Math on Trial blends courtroom drama, history, and math to show that legal expertise isn’t always enough to prove a person innocent."

PMF poster 'Thoughtful' has a book coming out. It will be interesting to analyse the book to try and understand why someone so book educated is so stongly affected by confirmation bias.
User avatar
Sarah
Site Admin
 
Posts: 3542
Joined: Thu May 06, 2010 11:23 pm

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 06, 2013 9:23 pm

Google offers a much better preview than amazon.

http://books.google.com/books?id=PFAIb6 ... 22&f=false
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 06, 2013 9:57 pm

She writes:

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


Except this is what they wrote:

"[R]ight from the outset we must resist the attempt -- reflected in [literally "pursued by"] the entire structure of the defense case, but out of place in the context of this decision -- to involve the [Cassazione] panel in endorsing the hypothesis that others, Raffaele Sollecito and Amanda Knox, were responsible for the murder, aggravated by sexual violence, of Meredith Kercher. The decision that this Court is called to make concerns only the responsibility of Guede with regard to the act under dispute; any participation of others in the crime will be taken into account only to the extent that this circumstance bears on the matter that is our only concern: the modification or confirmation of the judgement of responsibility of the defendant -- the latter being wholly agreed upon by the judges of the first and second level."
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby Annella » Wed Mar 06, 2013 10:08 pm

'The Italian concept of judicial truth does not trouble itself with reality; it controls the narrative by controlling the past"
User avatar
Annella
 
Posts: 1429
Joined: Tue Oct 25, 2011 6:19 pm

Re: 'Math on Trial'

Postby Grayhawker » Wed Mar 06, 2013 11:41 pm

MichaelB wrote:She writes:

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


Except this is what they wrote:

"[R]ight from the outset we must resist the attempt -- reflected in [literally "pursued by"] the entire structure of the defense case, but out of place in the context of this decision -- to involve the [Cassazione] panel in endorsing the hypothesis that others, Raffaele Sollecito and Amanda Knox, were responsible for the murder, aggravated by sexual violence, of Meredith Kercher. The decision that this Court is called to make concerns only the responsibility of Guede with regard to the act under dispute; any participation of others in the crime will be taken into account only to the extent that this circumstance bears on the matter that is our only concern: the modification or confirmation of the judgement of responsibility of the defendant -- the latter being wholly agreed upon by the judges of the first and second level."


Ooops! Nasty stuff truth. So inconvenient. Especially when you put a contradiction in print.
Paolo Micheli stated with regard to Amanda and Raffaele: "We do not need evidence, common sense and logic tell us that they dated each other to commit this crime."
User avatar
Grayhawker
 
Posts: 2387
Joined: Thu Oct 06, 2011 4:13 pm
Location: Kansas City

Re: 'Math on Trial'

Postby Annella » Thu Mar 07, 2013 1:06 am

Grayhawker wrote:
MichaelB wrote:She writes:

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


Except this is what they wrote:

"[R]ight from the outset we must resist the attempt -- reflected in [literally "pursued by"] the entire structure of the defense case, but out of place in the context of this decision -- to involve the [Cassazione] panel in endorsing the hypothesis that others, Raffaele Sollecito and Amanda Knox, were responsible for the murder, aggravated by sexual violence, of Meredith Kercher. The decision that this Court is called to make concerns only the responsibility of Guede with regard to the act under dispute; any participation of others in the crime will be taken into account only to the extent that this circumstance bears on the matter that is our only concern: the modification or confirmation of the judgement of responsibility of the defendant -- the latter being wholly agreed upon by the judges of the first and second level."


Ooops! Nasty stuff truth. So inconvenient. Especially when you put a contradiction in print.



So there will be questions for the author then. :)
'The Italian concept of judicial truth does not trouble itself with reality; it controls the narrative by controlling the past"
User avatar
Annella
 
Posts: 1429
Joined: Tue Oct 25, 2011 6:19 pm

Re: 'Math on Trial'

Postby european neighbour » Thu Mar 07, 2013 2:47 am

This is weird! I love their approach, really! They should just apply their skills for Massei's reasoning!!

What's about: "MATH ON MASSEI'S TRIAL" Vol.1, 2, 3.....
european neighbour
 
Posts: 935
Joined: Wed May 26, 2010 9:28 am

Re: 'Math on Trial'

Postby suomisat » Thu Mar 07, 2013 5:30 am

Yeah I'd love to see a professional logician tackle Massei and Galati.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby roteoctober » Thu Mar 07, 2013 6:46 am

Math on Trial page 85 wrote:What Hellmann is saying is that the fact of having run the experiment independently two times and obtained the same result
twice does not increase the reliability of the result. "The sum of two results, both unreliable...cannot give a reliable result."
This sentence shows a complete misundertanding of the probabilistic result...




No.

Hellmann is not doing a probabilistic calculation, he is applying logical-juridical arguments.

If you go on reading page 85 you'll read about the tossed coins analogy made by the author to justify her statement about Hellmann misunderstanding the probabilistic result.

First of all she makes an example arbitrarily attributing to the correctness of the DNA profile on the knife a probability of 80-90% .

Well, attributing to them an 80-90% correctness is arbitrary at best.

Since Hellmann defines it "seriously weakened" as evidence, one could rather think of a 20-30% probability of correctness, once considered that the method (procedure) used to obtain it does not satisfy the methods considered as established by a broad consensus in the scientific community.

But Hellmann's task is not that of calculating probabilities: his point is logical-juridical as I said before and can be expressed as "I have to consider only results which are the products of procedure which meet the criteria considered appropriate by the majority of the scientific community and this for each one of them considered independently. I still consider a result not satisfying these criteria as a possible very weak circumstantial evidence and I could even admit a second one, but even a couple of them would be in contradiction with all the others known elements concerning the knife which are telling us that knife is not the murder weapon, so even TWO small traces of the victim's DNA on the blade would not change my conclusions."

The DNA got all the limelight in the debate and apparently also at trial, but H&Z put its value in the correct perspective in their report: whatever you can make of it scientifically, a couple of DNA traces (moreover in extremely small quantities, being at the extreme limits of the possibilities of the machines, if not beyond) don't make a murder weapon out of a knife that everything else we know about it is excluding as the murder weapon.
And indeed this last point would maintain its strength even with a DNA with a 100% reliability.

In conclusion, the calculations made in the book concerning the knife are correct (and they are so easy that it could not be otherwise), the point is that the assumptions (both the 80-90% probability and that Hellmann did a probabilistic calculation) are arbitrary and at least for what concerns the second, even plainly wrong.
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby suomisat » Thu Mar 07, 2013 7:20 am

So Thoughtful has either made a dumb mistake or is deliberately obscuring the issue. That's what I thought.

A whole chapter of her book, basically an over-literalist mistake. Exactly as I suspected at first glance, but it's good to get my hunch confirmed.

Thoughtful is a perfect guilter in this regard.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby MichaelB » Thu Mar 07, 2013 7:24 am

The daughter and co-author is available via this website to be an "expert" for tv on crime, law & order

http://www.findatvexpert.com/sector/crime-law-order-3.htm

:wow: Who needs John Douglas or Steve Moore's opinion when you can get this girl, a maths tutor.

Image
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby MichaelB » Thu Mar 07, 2013 7:25 am

The google links for these two are in the hundreds.

http://www.theaccessproject.org.uk/peop ... ie-colmez/
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby geebee2 » Thu Mar 07, 2013 7:41 am

To examine the issue of testing the DNA found in sample I, it seems fair to quote Galati

2.1 The Order of September 7, 2011 – Rejection of fresh expert examination
Contradiction and manifest illogicality of the reasoning. Missing acquisition of decisive piece of evidence – Article 606(e) and (d) Criminal Procedure Code

The CAA, by the order of September 7, 2011 has rejected, amongst other things, the Procurator-General’s request to provide a supplement of expert testimony with the aim of carrying out genetic analysis on the quantity of human DNA extracted from the new trace sampled by the experts on the blade of the [28] knife (trace 1 according to the Vecchiotti-Conti report) in proximity to the point in which Dr Stefanoni of the Scientific Police had sampled the trace from which she had extracted the DNA attributed to Meredith Kercher.

Although the first question formulated by the CAA inherently encompassed new samplings and analyses of probable DNA found, Dr Vecchiotti had considered not proceeding with the analysis of the extracted DNA and its consequent attribution in as much, on her advice, the quantity was not sufficient to carry out a reliable analysis. It was a case, that is, of low copy number, a term used in the literature in the field to indicate, in fact, a small quantity of DNA.

The prosecution consultant, Prof. Giuseppe Novelli (undisputed luminary of human genetics), on the contrary, had evidenced during the course of his depositions (cf. hearing transcript of September 6, 2011, pp52 and following) how, just like at the time the unrepeatable technical tests carried out by Dr Stefanoni (2007/2008), it was possible to analyse low copy number DNA traces with totally reliable results and how, with the new machines introduced in the meantime and amply tested, one could have successfully proceeded with the extraction, amplification and attribution of DNA having at one’s disposition even just one cell (10, 15 picograms) and, therefore, of a quantity a great deal less than that sampled by the experts (approximately 100 picograms). The CAA, inexplicably, rejected the request, holding that ‚To leave aside the sustainability or otherwise of the gaps set out by the Procurator General, and shared by the civil parties, the tests effected by the experts and the evaluative elements proposed by the parties’ consultants allow this Court to form its own reasoned [29] conclusion‛.

It is quite true that the Supreme Court’s jurisprudence holds the second instance court’s reasoning to a less stringent standard when it is deciding to reject a preliminary investigation submission, but it is also true that, in the case with which we are occupied, the reasoning can be said to exist exclusively under a graphic profile and therefore purely apparent.

Given, in fact, the expert examination requested by the defence, a new element had appeared at the completion of the technical investigations susceptible of biochemical analysis, that is, further traces of DNA, not collected before.

The discovery ought to have induced the CAA, in concordance with the prior decision, to order new tests, to widen the scientific enquiry for a more complete evaluation of the traces, in particular those found during the course of the expert examination.

The Procurator-General, in this regard, had, as already mentioned, evidenced that new biogenetic investigative techniques would have allowed the analysis of this unexpectedly occurring piece of evidence, even if of scanty quantity; this was shown to the Court with total precision and by the authority of consultant Prof Novelli, and on the basis of the declarations of this latter, the Public Prosecutor had requested a supplementary expert examination.

In the face of the discovery of new evidence and a request to test it, specifically argued under the scientific and procedural aspect, the Court adopted a provision to reject, privy of any real reasoning and in obvious contradiction to the necessity of further biological understanding explicitly affirmed in the dispositive order for the expert examination.

In the first place, because if the reasoned and full deliberation involves [30] the outcomes of expert tests, it cannot be based on the guidance of elements that have not been acquired by the experts’ unilateral decision and not shared with the parties’ consultants. It would otherwise, in fact, have been the case if the experts would have reported not having found any biological trace and the Procurator-General’s request had been that of proceeding to with a fresh sampling: in this case, in fact, a reasoning that would have rejected the request for a consideration of the value of the already-obtained negative result would have been congruent. But the case before us is quite different: the fresh biological trace exists, but there was no proceeding to go to an analysis of it and to extract the respective profile from it.

In the second place, because the second instance Court [the CAA] negates the relevance of a test that it itself has disposed with the first of the formulated questions: ‚the College of experts to ascertain: whether, by means of new technical testing, the attribution and the level of reliability of the probable attribution of the DNA present on exhibits 165b (bra-clasp) and 36 (knife) is possible; if it is not possible to proceed with a new technical test, to evaluate, on the basis of the record, the level of reliability of the genetic tests carried out by the Scientific Police on the above-mentioned exhibits, with reference also to probable contaminations‛. With the experts recovering a new trace containing human DNA on the blade of the knife (a certainly odd presence, if one has in mind the ordinary use of the utensil, and overall significant – even of itself, and to leave aside the attribution – and useful to compare with what was found by Forensics on the same blade), they could not exempt themselves from carrying the test to a conclusion, so being able to evaluate the unreliability or otherwise of the result only on completed procedures, just as the same question was explicitly suggesting. The refusal is even further unjustified if one considers [31] the state of technology at the time of the experts’ response to the questions and the trenchant declarations of Prof Novelli, never contradicted by anyone, if not the same second instance Court [the CAA], which affirms in judgment, contrary to the truth and to what emerges from the hearing transcripts, that the procedures of which Prof Novelli spoke are still in an experimental phase (page 84 of the judgment). Apart from the obvious error of having held the concepts of ‚leading edge‛ and ‚in experimental phase‛ to be identical when, in reality, they refer to completely separate situations: a technical certainty is experimental when its reliability has not yet been confirmed; it is at the leading edge when the method is innovative with respect to those preceding it; what counts more, though, is that Prof Novelli defined leading edge techniques as those with which: ‚we are able to analyse, to produce profiles, with respect to the standard kits, in the order of 10 picograms, so we are already beyond the famous 100 picograms which people talk about today.

Procurator-General: So below the quantity which has been found in these traces?

NOVELLI: It’s possible, see, even I, in my field of geneticist I often find myself analysing small traces of DNA even as regards diagnostic questions, diseases, diagnoses on human embryos where we would have one cell only, we’re speaking of 7 or 8 picograms, you can see how we have to be precise and accurate‛ (page 53 of the hearing transcript of September 6, 2011 already referred to).

It is clear, therefore, from the thrust of Prof Novelli’s declarations, the techniques, which he described as leading edge –and for this reason put into practice by well-equipped and qualified laboratories – which allow the analysis of quantities of DNA at levels well below those with respect to which [32] was recovered by the experts appointed by the Court, whose reasoning on the specific point is, therefore, clearly illogical.

As can be quite well seen, the non-acquisition of an item of evidence that the Court itself had considered necessary, disposing the re-opening of oral argument and formulating that specific question for the experts, vitiates the judgment not only under the reasoning aspects placed into evidence so far, but also under the separate aspect of the non-acquisition of a decisive element of proof: and in truth, if the DNA sampled by the experts had been attributed to the victim, the decisive value of this result would have been obvious, it would have totally negated the experts’ conclusions on the unreliability of the analyses carried out by the Scientific Police on the murder weapon. And, on the other side, it is the Court of Cassation’s instruction the parties’ recourse as of right to contrary evidence, even in cases where the evidence that is intended to be tested has been ex officio acquired by the court. The relevant jurisprudence, with a constant line, on this point has affirmed: ‚On the matter of oral argument instruction, the court which – in the exercise of exceptional power pursuant to Article 507 Criminal Procedure Code – admits new evidence, but also admits contrary evidence.‛51

It must be held that the principle can and must be applied even in cases like the instant, in consideration of, – even though involving expert examinations –, it would have had as its object not only evaluation, but also acquisition, of data (the profile extracted from the DNA recovered by the experts from the blade of the knife) which would have constituted further evidence.

Documents referrable:
Appendix 6: CAA hearing transcript of September 6, 2011 – examination of Prof Giuseppe Novelli, technical expert for the prosecution.
[33] Appendix 8: CAA hearing transcript of September 7, 2011


But I think the issue is procedural - the prosecution cannot keep coming back asking for new evidence to be tested, delaying the trial.
Especially when the decision not to proceed to amplification was taken with the agreement of the prosecution experts.
User avatar
geebee2
 
Posts: 5159
Joined: Sun Oct 16, 2011 12:39 pm
Location: Gloucester, United Kingdom

Re: 'Math on Trial'

Postby B_Real » Thu Mar 07, 2013 8:01 am

A reasonable person is not obsessed with the idea of Amanda plunging a kitchen knife into Meredith's throat as part of some bizarre group sacrifice.

The fantasy of Amanda killing Meredith did not even originate from evidence or statements. It came purely from imagination before there was any evidence.

A reasonable person has trouble believing this lurid complex scenario without solid, reliable evidence that it happened, of which there is none and never will be. Indeed to believe the scenario requires all manner of other preposterous assumptions to explain things like the lack of other evidence, why is the knife there anyway, not throwing the knife away, nobody fingering Guede, and on and on.

A piece of decisive evidence should not require the construction of additional elaborate theories. That's the opposite of what decisive evidence is supposed to be. And that is basically Hellmann's argument.

The police probably didn't understand that when they seized the 'murder weapon' from Raffaele's drawer.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby Jstanz » Thu Mar 07, 2013 9:07 am

suomisat wrote:So Thoughtful has either made a dumb mistake or is deliberately obscuring the issue. That's what I thought.

A whole chapter of her book, basically an over-literalist mistake. Exactly as I suspected at first glance, but it's good to get my hunch confirmed.

Thoughtful is a perfect guilter in this regard.


I think after the acquittal Thoughtful faced quite a dilemma. After all the work they'd put in on the book....what to do? what to do? I don't think she can afford to admit she was wrong with a book on the line.
All my grammar and spelling mistakes are the result of auto-correct. If auto-correct is not used here, I still blame it.
Jstanz
 
Posts: 2781
Joined: Mon Aug 08, 2011 8:13 am

Re: 'Math on Trial'

Postby suomisat » Thu Mar 07, 2013 9:21 am

Jstanz wrote:
suomisat wrote:So Thoughtful has either made a dumb mistake or is deliberately obscuring the issue. That's what I thought.

A whole chapter of her book, basically an over-literalist mistake. Exactly as I suspected at first glance, but it's good to get my hunch confirmed.

Thoughtful is a perfect guilter in this regard.


I think after the acquittal Thoughtful faced quite a dilemma. After all the work they'd put in on the book....what to do? what to do? I don't think she can afford to admit she was wrong with a book on the line.


Maybe so. I can see why they may have raced to get it published before the High Court decision.

The only thing that irks me is that the prosecution may cite the book before the High Court and say something about internationally acclaimed mathematicians say Hellmann messed up blah blah (even though Thoughtful's analysis misses the point entirely).
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby MichaelB » Thu Mar 07, 2013 9:29 am

suomisat wrote:
Jstanz wrote:
suomisat wrote:So Thoughtful has either made a dumb mistake or is deliberately obscuring the issue. That's what I thought.

A whole chapter of her book, basically an over-literalist mistake. Exactly as I suspected at first glance, but it's good to get my hunch confirmed.

Thoughtful is a perfect guilter in this regard.


I think after the acquittal Thoughtful faced quite a dilemma. After all the work they'd put in on the book....what to do? what to do? I don't think she can afford to admit she was wrong with a book on the line.


Maybe so. I can see why they may have raced to get it published before the High Court decision.

The only thing that irks me is that the prosecution may cite the book before the High Court and say something about internationally acclaimed mathematicians say Hellmann messed up blah blah (even though Thoughtful's analysis misses the point entirely).


!!!!!

It doesn't work like that. It's a half day hearing and there's no waving some silly book written by a woman in Paris and her daughter, a maths tutor.
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby erasmus44 » Thu Mar 07, 2013 11:51 am

A few thoughts. 1. Mathematicians will almost always prefer more data regardless of its reliability because it does tend to reduce uncertainty. Economists weigh the costs of the additional data (notice the debate about various medical tests). The legal system generally adopts a threshold level of reliability for "evidence" and excludes everything below that threshold because of the danger that unreliable evidence (although having "some" probative value) will bias the forum and lead it in the wrong direction. Low copy DNA falls in this area and the legal system generally excludes it (below a certain level) but it has "some" probative value (like, for example, the fact that we know Amanda was not in the United States at the time of the crime - this could be verified with data from Immigration and Naturalization and would be somewhat "probative" in the sense that it is more likely that Amanda committed the crime if she was not in the United States at the time of the crime than if she was in the United States at the time of the crime).
2. Even if I knew for sure that they had committed the crime (but didn't know exactly how) and the low copy DNA on the knife turned out to be Meredith's, I would still assign a fairly high probability to it being the result of contamination. Here's why. We know that at least one other knife was used in the crime and it appears to have been disposed of. For this knife to be used, several difficult assumptions have to be made. First, that they brought the knife over to the cottage which implies a premeditated plan to commit the crime, but most of the theories of the crime assume a degree of spontaneity and "things getting out of control" and it is hard to get the knife over to the cottage consistent with such a theory. Massei gets around this by arguing that Amanda carried the knife around with her for protection at all times, but if you believe that, then of course the knife was frequently over at the cottage and could have picked up Meredith's DNA from something else in Amanda's bag, could have been used at the cottage, or could have picked up DNA when Amanda's bag was open. Second, you have to assume that the perps disposed on one knife but kept this knife and left it on top of the pile of knives in Raffeale's kitchen drawer - a rather bizarre assumption. Given the generally sloppy forensic work in this case, and even assuming they committed the crime, I would tend to think that the probability of the DNA being the result of either 1. contamination or 2. innocent picking up of DNA while in AK's purse is higher than the probability that it picked up DNA during the course of the crime.
3. Once you reach this conclusion, then the finding of DNA on the knife - especially low copy DNA which makes contamination more plausible is not very probative at all. One question the Supreme Court may ask itself is whether, if such a finding were made, it could tilt the balance to "guilt beyond a reasonable doubt" - if it could not possibly do so, then it doesn't make sense to do the test. In my opinion, it could not possibly do so.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby roteoctober » Thu Mar 07, 2013 2:54 pm

The quarrel concerning the picograms is not really stuff for the SC, what they may consider is that if on the one hand we have Novelli saying that he is able to get profiles from less than 10 picograms in its work with embryos, on the other hand we have the ample literature quoted by Hellmann, expressing the rather established agreement that LCN DNA is of limited utility in forensic science.
It's not just about Novelli or someone else being able to extract a profile from 10 picograms with new machines, the point is "how much this has been done or is being done and is accepted in the field of forensic science concerning crimes and trials and not pure research?"

But indeed, and erasmus44 nailed it, the point is about "and even if Meredith's profile was there, then what?"

According to article 606 section d, refusing to allow contrary evidence is reason for a new trial if it is "decisive", and indeed Galati makes a reference to this:

if the DNA sampled by the experts had been attributed to the victim, the decisive value of this result would have been obvious, it would have totally negated the experts’ conclusions on the unreliability of the analyses carried out by the Scientific Police on the murder weapon


Now, deciding what is decisive and justifies a new trial is up to the SC judges, but can the presence of the victim's DNA on that knife overturn the acquittal (to be decisive it must be such as to change by itself the outcome of the trial)?

Another way of putting the question is "if Meredith's DNA is on the blade they are guilty, while if it is not there they are innocent?"...it's up to the SC judges to decide but they will have to consider both contamination issues (the smaller the quantity of DNA, the higher the chance of a contamination) and above all (I've written this too many times by now :lol: ) that there are many other elements concerning that knife which point to the knife not being the murder weapon.

Too much emphasis has been placed on picograms of DNA, it is time for judges who are not lab technicians to look at this matter in a wider perspective.
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby geebee2 » Thu Mar 07, 2013 2:57 pm

Could the supreme court order the test to be carried out, and then decide what to do next depending on the result?

I don't they will, because it would set a pretty ridiculous precedent. Surely the prosecution cannot go chasing after possible new evidence forever.
User avatar
geebee2
 
Posts: 5159
Joined: Sun Oct 16, 2011 12:39 pm
Location: Gloucester, United Kingdom

Re: 'Math on Trial'

Postby roteoctober » Thu Mar 07, 2013 3:13 pm

geebee2 wrote:Could the supreme court order the test to be carried out, and then decide what to do next depending on the result?

I don't they will, because it would set a pretty ridiculous precedent. Surely the prosecution cannot go chasing after possible new evidence forever.


No, the SC can just order a new trial, they can't acquire new evidence of any kind, that's reserved to the lower grades.

The prosecution would be right if the case could really be decided by that DNA and if Novelli's new techniques could be considered well established in the field of forensic genetics (which is not the same of a lab working on embryos where the quality and purity of the sample is certain).
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby geebee2 » Thu Mar 07, 2013 6:24 pm

Ok. I don't think they will act, because all the other tests and evidence are negative - there is absolutely no reason to think testing sample I would produce a positive result.

I presume that if the supreme court rubber stamp Hellmann ( as seems likely, except maybe for the calunnia ), that really is the end of it, there is no further appeal possible.
User avatar
geebee2
 
Posts: 5159
Joined: Sun Oct 16, 2011 12:39 pm
Location: Gloucester, United Kingdom

Re: 'Math on Trial'

Postby Chris_Halkides » Thu Mar 07, 2013 9:14 pm

I think if anyone argues for another round of testing, he or she should be expected to explain what the interpretation of each possible result will be and why they think so. Would a negative result fall into the exculpatory category? If not, why not?
Chris_Halkides
 
Posts: 1848
Joined: Tue Nov 02, 2010 4:33 pm

Re: 'Math on Trial'

Postby sept79 » Thu Mar 07, 2013 9:59 pm

Continued testing on the knife is simply a non-computer version of 'garbage in, garbage out'.
sept79
 
Posts: 1412
Joined: Wed May 26, 2010 7:16 am

Re: 'Math on Trial'

Postby roteoctober » Fri Mar 08, 2013 2:01 am

geebee2 wrote:Ok. I don't think they will act, because all the other tests and evidence are negative - there is absolutely no reason to think testing sample I would produce a positive result.

I presume that if the supreme court rubber stamp Hellmann ( as seems likely, except maybe for the calunnia ), that really is the end of it, there is no further appeal possible.


Indeed my point is that even a positive result would not be of much importance and certainly not decisive.

Yes if the acquittal is confirmed by the SC that really is the end.
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby european neighbour » Fri Mar 08, 2013 2:25 am

geebee2 wrote:I presume that if the supreme court rubber stamp Hellmann ( as seems likely, except maybe for the calunnia ), that really is the end of it, there is no further appeal possible.

What's about Strasbourg?
european neighbour
 
Posts: 935
Joined: Wed May 26, 2010 9:28 am

Re: 'Math on Trial'

Postby roteoctober » Fri Mar 08, 2013 5:48 am

There could be an appeal to the ECHR if the conviction for calumny is upheld.
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby ClintNine » Sat Mar 09, 2013 3:07 pm

roteoctober wrote:
Math on Trial page 85 wrote:What Hellmann is saying is that the fact of having run the experiment independently two times and obtained the same result
twice does not increase the reliability of the result. "The sum of two results, both unreliable...cannot give a reliable result."
This sentence shows a complete misundertanding of the probabilistic result...




No.

Hellmann is not doing a probabilistic calculation, he is applying logical-juridical arguments.

If you go on reading page 85 you'll read about the tossed coins analogy made by the author to justify her statement about Hellmann misunderstanding the probabilistic result.

First of all she makes an example arbitrarily attributing to the correctness of the DNA profile on the knife a probability of 80-90% .

Well, attributing to them an 80-90% correctness is arbitrary at best.

Since Hellmann defines it "seriously weakened" as evidence, one could rather think of a 20-30% probability of correctness, once considered that the method (procedure) used to obtain it does not satisfy the methods considered as established by a broad consensus in the scientific community.

But Hellmann's task is not that of calculating probabilities: his point is logical-juridical as I said before and can be expressed as "I have to consider only results which are the products of procedure which meet the criteria considered appropriate by the majority of the scientific community and this for each one of them considered independently. I still consider a result not satisfying these criteria as a possible very weak circumstantial evidence and I could even admit a second one, but even a couple of them would be in contradiction with all the others known elements concerning the knife which are telling us that knife is not the murder weapon, so even TWO small traces of the victim's DNA on the blade would not change my conclusions."

The DNA got all the limelight in the debate and apparently also at trial, but H&Z put its value in the correct perspective in their report: whatever you can make of it scientifically, a couple of DNA traces (moreover in extremely small quantities, being at the extreme limits of the possibilities of the machines, if not beyond) don't make a murder weapon out of a knife that everything else we know about it is excluding as the murder weapon.
And indeed this last point would maintain its strength even with a DNA with a 100% reliability.

In conclusion, the calculations made in the book concerning the knife are correct (and they are so easy that it could not be otherwise), the point is that the assumptions (both the 80-90% probability and that Hellmann did a probabilistic calculation) are arbitrary and at least for what concerns the second, even plainly wrong.


I agree completely with your post. I think this example in Thoughtful's book, "Maths on Trial" illustrates what is wrong with many arguments on PMF and TJMK. They try too hard to force a conclusion based on suppositions. They attempt to give weight to their suppositions using any means possible: Intimidation, bullying, exaggeration and most heinous of all, they pretend what they are asserting is LOGICAL. There is no logic to what happened that night. But certainly no proof beyond a reasonable doubt that AK and RS were involved. Thoughtful's example is a thought experiment heavily weighted to her own point of view. It's a completely arbitrary and false argument that does not at all characterize the reality of Hellmann's decision and the validity and rationality of its jurisprudence. But she passes off her critique as if it were unassailable because she is a professor of mathematics at a university in Paris. It's nonsense.
ClintNine
 
Posts: 556
Joined: Fri Nov 16, 2012 1:31 pm

Re: 'Math on Trial'

Postby ClintNine » Sat Mar 09, 2013 3:22 pm

erasmus44 wrote:A few thoughts. 1. Mathematicians will almost always prefer more data regardless of its reliability because it does tend to reduce uncertainty. Economists weigh the costs of the additional data (notice the debate about various medical tests). The legal system generally adopts a threshold level of reliability for "evidence" and excludes everything below that threshold because of the danger that unreliable evidence (although having "some" probative value) will bias the forum and lead it in the wrong direction. Low copy DNA falls in this area and the legal system generally excludes it (below a certain level) but it has "some" probative value (like, for example, the fact that we know Amanda was not in the United States at the time of the crime - this could be verified with data from Immigration and Naturalization and would be somewhat "probative" in the sense that it is more likely that Amanda committed the crime if she was not in the United States at the time of the crime than if she was in the United States at the time of the crime).
2. Even if I knew for sure that they had committed the crime (but didn't know exactly how) and the low copy DNA on the knife turned out to be Meredith's, I would still assign a fairly high probability to it being the result of contamination. Here's why. We know that at least one other knife was used in the crime and it appears to have been disposed of. For this knife to be used, several difficult assumptions have to be made. First, that they brought the knife over to the cottage which implies a premeditated plan to commit the crime, but most of the theories of the crime assume a degree of spontaneity and "things getting out of control" and it is hard to get the knife over to the cottage consistent with such a theory. Massei gets around this by arguing that Amanda carried the knife around with her for protection at all times, but if you believe that, then of course the knife was frequently over at the cottage and could have picked up Meredith's DNA from something else in Amanda's bag, could have been used at the cottage, or could have picked up DNA when Amanda's bag was open. Second, you have to assume that the perps disposed on one knife but kept this knife and left it on top of the pile of knives in Raffeale's kitchen drawer - a rather bizarre assumption. Given the generally sloppy forensic work in this case, and even assuming they committed the crime, I would tend to think that the probability of the DNA being the result of either 1. contamination or 2. innocent picking up of DNA while in AK's purse is higher than the probability that it picked up DNA during the course of the crime.
3. Once you reach this conclusion, then the finding of DNA on the knife - especially low copy DNA which makes contamination more plausible is not very probative at all. One question the Supreme Court may ask itself is whether, if such a finding were made, it could tilt the balance to "guilt beyond a reasonable doubt" - if it could not possibly do so, then it doesn't make sense to do the test. In my opinion, it could not possibly do so.


Also an excellent post. It shows the quagmire of SC trying to resolve the DNA evidence on the knife. I can't imagine Cassation saying to have another go at the knife - it could go on forever and the side with the answer when the music stops is right? Can't work that way. So therefore the principle of reasonable doubt - whether based on possible contamination or that the knife was not the murder weapon - is the only correct decision.

I don't understand what is being argued so I guess I need to read this book in order to comment. But I do know that jurisprudence cannot be based on running an experiment 100 times when in this case, there wasn't even enough of the mystery substance for one test.
ClintNine
 
Posts: 556
Joined: Fri Nov 16, 2012 1:31 pm

Re: 'Math on Trial'

Postby ClintNine » Sat Mar 09, 2013 3:32 pm

MichaelB wrote:The daughter and co-author is available via this website to be an "expert" for tv on crime, law & order

http://www.findatvexpert.com/sector/crime-law-order-3.htm

:wow: Who needs John Douglas or Steve Moore's opinion when you can get this girl, a maths tutor.

Image


Actually the daughter is riding on the mother's (the poster Thoughtful on PMF) coat tails on this one. The mother is a maths professor in Paris - long career and Harvard educated and quite brilliant. The mother has also written a lot of crime novels under a pen name. The daughter is no doubt brilliant also.

I just think that probability wouldn't be a basis for attacking Hellmann's ruling- but I have to read the book.
ClintNine
 
Posts: 556
Joined: Fri Nov 16, 2012 1:31 pm

Re: 'Math on Trial'

Postby Dedj » Sat Mar 09, 2013 4:25 pm

ClintNine wrote:
roteoctober wrote:
Math on Trial page 85 wrote:What Hellmann is saying is that the fact of having run the experiment independently two times and obtained the same result
twice does not increase the reliability of the result. "The sum of two results, both unreliable...cannot give a reliable result."
This sentence shows a complete misundertanding of the probabilistic result...



Thoughtful's example is a thought experiment heavily weighted to her own point of view. It's a completely arbitrary and false argument that does not at all characterize the reality of Hellmann's decision and the validity and rationality of its jurisprudence. But she passes off her critique as if it were unassailable because she is a professor of mathematics at a university in Paris. It's nonsense.


Spot on.

Hellmann is not saying that a second independent experiment cannot contribute to the probability that the result of the first experiment is the true result - he's clearly saying that theres no guarantee about the reliability of the experiments and therefore no guarantee that either result is an accurate reflection of the subject being tested.

This is elementary GCSE Science - if your method is wonky, you can repeat it a hundred times, you've still no idea if you got the results due to the poor methods or if you correctly tested what was there.
Dedj
 
Posts: 385
Joined: Tue Jul 10, 2012 2:10 pm

Re: 'Math on Trial'

Postby Didaktylos » Sun Mar 10, 2013 5:00 am

Dedj wrote:
ClintNine wrote:
roteoctober wrote:
Math on Trial page 85 wrote:What Hellmann is saying is that the fact of having run the experiment independently two times and obtained the same result
twice does not increase the reliability of the result. "The sum of two results, both unreliable...cannot give a reliable result."
This sentence shows a complete misundertanding of the probabilistic result...



Thoughtful's example is a thought experiment heavily weighted to her own point of view. It's a completely arbitrary and false argument that does not at all characterize the reality of Hellmann's decision and the validity and rationality of its jurisprudence. But she passes off her critique as if it were unassailable because she is a professor of mathematics at a university in Paris. It's nonsense.


Spot on.

Hellmann is not saying that a second independent experiment cannot contribute to the probability that the result of the first experiment is the true result - he's clearly saying that theres no guarantee about the reliability of the experiments and therefore no guarantee that either result is an accurate reflection of the subject being tested.

This is elementary GCSE Science - if your method is wonky, you can repeat it a hundred times, you've still no idea if you got the results due to the poor methods or if you correctly tested what was there.


As we say in IT - GIGO
Come together, raise up your voices
This time my song of love and life won't go away
I'll sing forever here in the sunshine
I've lived to see the sun break through the storm
And I'm so glad I'm standing here today
Didaktylos
 
Posts: 576
Joined: Sat Nov 05, 2011 11:18 am

Re: 'Math on Trial'

Postby suomisat » Sun Mar 10, 2013 5:56 am

Thoughtful's blatant error (or deliberate distortion) is yet another variation of the usual guilter reasoning: EGIGO.

Enough garbage in, gold out.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby erasmus44 » Sun Mar 10, 2013 1:03 pm

In fairness, if the problem is a tendency of the test to produce false positives with a known error rate, then conducting a second, completely independent, test could reduce the probability of error. On the other hand, if you adopt my chain of reasoning and posit that a key problem is the potential that MK's DNA may have gotten on the knife due to either contamination or exposure consistent with Massei's theory that AK carried the knife at all times and therefore it was often over at the cottage, then testing again adds nothing. The problem for the prosecution is that the only plausible way the knife gets over to the cottage is Massei's reasoning that AK carried it at all times but if she carried it at all times then it was frequently over at the cottage and intermingled with other belongings in her bag which could pick up MK's DNA because they lived together and shared a kitchen and bathroom.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby Dougm » Sun Mar 10, 2013 1:20 pm

erasmus44 wrote:In fairness, if the problem is a tendency of the test to produce false positives with a known error rate, then conducting a second, completely independent, test could reduce the probability of error. On the other hand, if you adopt my chain of reasoning and posit that a key problem is the potential that MK's DNA may have gotten on the knife due to either contamination or exposure consistent with Massei's theory that AK carried the knife at all times and therefore it was often over at the cottage, then testing again adds nothing. The problem for the prosecution is that the only plausible way the knife gets over to the cottage is Massei's reasoning that AK carried it at all times but if she carried it at all times then it was frequently over at the cottage and intermingled with other belongings in her bag which could pick up MK's DNA because they lived together and shared a kitchen and bathroom.


In fairness, Massei could have surmised that Amanda and Raffaele took the knife from his place that night only, because they had decided to kill Meredith with it. But not even Massei was willing to invent premeditation out of whole cloth. And if they had wanted to do that, there were plenty of knives at the cottage.

There is no plausible narrative that makes sense for them to have taken that knife over there. It doesn't fit the wounds. It doesn't match the bloody imprint on the bed. It didn't have any blood or even human or animal tissue on it. Massei came up with his wacky theory about Amanda carrying it around in her bag because he had decided she was guilty, and had to figure out some facts to match that. He should have approached it the other way around.
When you berate someone and push them and confuse them and lie to them and convince them that they're wrong you're not finding the truth.

Amanda Knox
Dougm
Moderator
 
Posts: 3189
Joined: Sat Jan 01, 2011 5:00 pm

Re: 'Math on Trial'

Postby Chris_Halkides » Sun Mar 10, 2013 2:14 pm

“The prosecution consultant, Prof. Giuseppe Novelli (undisputed luminary of human genetics), on the contrary, had evidenced during the course of his depositions (cf. hearing transcript of September 6, 2011, pp52 and following) how, just like at the time the unrepeatable technical tests carried out by Dr Stefanoni (2007/2008), it was possible to analyse low copy number DNA traces with totally reliable results and how, with the new machines introduced in the meantime and amply tested, one could have successfully proceeded with the extraction, amplification and attribution of DNA having at one’s disposition even just one cell (10, 15 picograms) and, therefore, of a quantity a great deal less than that sampled by the experts (approximately 100 picograms). The CAA, inexplicably, rejected the request, holding that ‚To leave aside the sustainability or otherwise of the gaps set out by the Procurator General, and shared by the civil parties, the tests effected by the experts and the evaluative elements proposed by the parties’ consultants allow this Court to form its own reasoned [29] conclusion‛.”


There are several problems that have occurred to me in reading through this passage from Galati. One is that it fails to address whether or not low template DNA is useful forensically, even when carried out under the most stringent conditions. That is Dr. Budowle’s criticism, as I understand it. Two is that the knife should never have been brought to a lab that performs standard forensic DNA typing if low template work were going to be performed IMO. “Very few laboratories perform low template DNA typing properly, because it requires dedicated facilities and great experience, although there are several published methods for the interpretation of such profiles.” This is a quote from a paper coauthored by Dr. Novelli. Three is that the idea of retesting fails to address the lack of blood on the knife.
Chris_Halkides
 
Posts: 1848
Joined: Tue Nov 02, 2010 4:33 pm

Re: 'Math on Trial'

Postby B_Real » Tue Mar 12, 2013 5:43 pm

thoughtful wrote:God forbid that there should be too many jurors out there who think that even DNA with 100% reliability should be thrown out in the face of any other evidence.


It's very simple.

Hellmann said the knife did not match any of the wounds on the victim. That's where we are at the moment, judicially speaking.

Obviously a sane person would start by trying to show the knife actually did match at least one wound, before worrying about the DNA.

There can be a lot of explanations for DNA, including planting it.

There is only one explanation for a knife not fitting the wounds.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 13, 2013 12:29 am

Hello PMF :wave:

Thoughtless is dishonest and her book is poorly researched if her chapter on the case is anything to go by. It's that simple. Thoughtless is entitled to her opinion but not to print falsehoods espically when the chapter implies innocent people got away with murder.

Here is what she wrote. There's a lot more mistakes in the chapter, eg like who broke down Meredith's door.

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


1. The Giodarno report does not decree that Rudy didn't act alone.

2. The autopsy reports do not "clearly indicate Meredith had been attacked by more than one person".

3. There was no DNA on the knife and it was completely discredited as evidence.

Leila Schneps is no better than a tabloid journalist and should stick to teaching maths in universities and not writing books about murder cases.

:twocents:
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 12:39 am

MichaelB wrote:Hello PMF :wave:

Thoughtless is dishonest and her book is poorly researched if her chapter on the case is anything to go by. It's that simple. Thoughtless is entitled to her opinion but not to print falsehoods espically when the chapter implies innocent people got away with murder.

Here is what she wrote. There's a lot more mistakes in the chapter, eg like who broke down Meredith's door.

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


1. The Giodarno report does not decree that Rudy didn't act alone.

2. The autopsy reports do not "clearly indicate Meredith had been attacked by more than one person".

3. There was no DNA on the knife and it was completely discredited as evidence.

Leila Schneps is no better than a tabloid journalist and should stick to teaching maths in universities and not writing books about murder cases.

:twocents:


Are you quoting directly from her book?
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 13, 2013 12:40 am

Bruce Fischer wrote:
MichaelB wrote:Hello PMF :wave:

Thoughtless is dishonest and her book is poorly researched if her chapter on the case is anything to go by. It's that simple. Thoughtless is entitled to her opinion but not to print falsehoods espically when the chapter implies innocent people got away with murder.

Here is what she wrote. There's a lot more mistakes in the chapter, eg like who broke down Meredith's door.

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


1. The Giodarno report does not decree that Rudy didn't act alone.

2. The autopsy reports do not "clearly indicate Meredith had been attacked by more than one person".

3. There was no DNA on the knife and it was completely discredited as evidence.

Leila Schneps is no better than a tabloid journalist and should stick to teaching maths in universities and not writing books about murder cases.

:twocents:


Are you quoting directly from her book?


Yep, from the google books preview the day I read it.
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 12:42 am

MichaelB wrote:
Bruce Fischer wrote:
MichaelB wrote:Hello PMF :wave:

Thoughtless is dishonest and her book is poorly researched if her chapter on the case is anything to go by. It's that simple. Thoughtless is entitled to her opinion but not to print falsehoods espically when the chapter implies innocent people got away with murder.

Here is what she wrote. There's a lot more mistakes in the chapter, eg like who broke down Meredith's door.

Their verdict was in flat contradiction to the Italian Supreme Court verdict decreeing that Rudy Guede didn't act alone, and with autopsy reports that clearly indicted that Meredith had been attacked by more than one person. But above all, the verdict was in contradiction with the DNA found on the knife.


1. The Giodarno report does not decree that Rudy didn't act alone.

2. The autopsy reports do not "clearly indicate Meredith had been attacked by more than one person".

3. There was no DNA on the knife and it was completely discredited as evidence.

Leila Schneps is no better than a tabloid journalist and should stick to teaching maths in universities and not writing books about murder cases.

:twocents:


Are you quoting directly from her book?


Yep, from the preview the day I read it.


Wow. I actually expected better from someone that followed the case for this long.
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby suomisat » Wed Mar 13, 2013 1:59 am

Laila Schneps’s chapter on the Kercher murder is trivial and even deceptive. The real issues here have to do with judicial reasoning regarding the weighting and evaluation of evidence and technical issues regarding the collection and testing of evidence. She totally misses the big picture.

To put it into perspective, a DNA scientist already knows (or could learn in half an hour) all the math in Schneps’s book.

A judge or legal theorist also already knows (or could learn in an hour or two, at most a few days if s/he's old) all the math in Schneps’s book.

It would take Schneps, on the other hand, years of study to be able to meaningfully contribute to the legal and techncial issues involved in DNA testing and its use in trials.

Schneps's other chapters may be better, but her Kercher chapter is not a meaningful contribution. It is just a description of the crime and trial (from a guilt-bias) that includes errors of fact, and that uses a valid but marginally relevant point of math to try to gain an aura of authority, where no actual authority is present. Her trivial math point sidesteps the real arguments and issues to which she pretends to respond.

This book could be an entertaining way to teach some statistical concepts. And it might help people see through the deceptive pseudo-mathematical arguments of lawyers, but when it turns to the Kercher case it becomes a piece of deceptive rhetoric that exploits math to support a false conclusion. It becomes exactly what Laila Schneps claims to oppose.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby RoseMontague » Wed Mar 13, 2013 3:08 am

Thoughtful is a friend of mine and I admire her thoughts on the case. Even if we don't agree. I believe she would be a valuable asset to the board. I have encouraged her to participate here both in the past and again recently. I am interested in how she will respond to the criticism of her chapter on the Kercher case.
User avatar
RoseMontague
 
Posts: 3526
Joined: Sun May 30, 2010 7:04 am

Re: 'Math on Trial'

Postby Jstanz » Wed Mar 13, 2013 6:36 am

I am interested in how Thoughtful will respond to the charges that she has important facts about the case wrong. That the information about the case in her book is incorrect. Which would logically lead to incorrect conclusions. Surely if this is true, there is recourse available? I think this is a valid discussion here.
All my grammar and spelling mistakes are the result of auto-correct. If auto-correct is not used here, I still blame it.
Jstanz
 
Posts: 2781
Joined: Mon Aug 08, 2011 8:13 am

Re: 'Math on Trial'

Postby diocletian » Wed Mar 13, 2013 7:11 am

This is a kind of professional cowardice. It's easy, when there is no one there to point out your mistakes, to write a piece putting forth a false narrative of a case clothed in mathematical theory. But if she were to attempt to make these points by testifying as an expert witness in this case, she would be shredded within minutes of taking the stand. That's why we have jury trials and people don't get convicted and sent to jail by kooky math professors.
diocletian
 
Posts: 1275
Joined: Thu Dec 15, 2011 10:09 am

Re: 'Math on Trial'

Postby diocletian » Wed Mar 13, 2013 8:19 am

It's actually very interesting to me that this person appears to be very good at math, yet at the same time, doesn't have the critical thinking skills necessary to assess evidence. It's almost like one part of her brain isn't wired the right way.
diocletian
 
Posts: 1275
Joined: Thu Dec 15, 2011 10:09 am

Re: 'Math on Trial'

Postby suomisat » Wed Mar 13, 2013 8:30 am

diocletian wrote:It's actually very interesting to me that this person appears to be very good at math, yet at the same time, doesn't have the critical thinking skills necessary to assess evidence. It's almost like one part of her brain isn't wired the right way.


That's typically how math people are. "Thoughtful" is full-on crazy in the same way a moon-landing denier is crazy. They can function OK in every other area but have a massive blind-spot or incapacity to see context.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby charlatan » Wed Mar 13, 2013 10:17 am

suomisat wrote:
diocletian wrote:It's actually very interesting to me that this person appears to be very good at math, yet at the same time, doesn't have the critical thinking skills necessary to assess evidence. It's almost like one part of her brain isn't wired the right way.


That's typically how math people are. "Thoughtful" is full-on crazy in the same way a moon-landing denier is crazy. They can function OK in every other area but have a massive blind-spot or incapacity to see context.

It's actually not uncommon for experts in a given field to believe their expertise applies directly to another field, or to ignore the contributions of researchers in another field altogether. For example, cutting edge research in physics is increasingly computational in nature, yet there's surprisingly little cross pollination between physicists and computer scientists.

There is much more to the field of research and experimental methodology than basic probability and statistics. I have a 750 page tome on my bookshelf that concerns the interpretation of numbers, and the design of valid experiments across multiple disciplines (the forensic investigation that precedes a trial can be viewed as a kind of experiment) -- this is just one of the many books and research papers on the topic. And yet even with all this analytical firepower at our disposal we still have to be reminded to use our brains:

Measurement, Design and Analysis wrote:Above All: Think!

The obvious bears, indeed requires, repeating: To be meaningful, any activity, including the reading of research reports, has to emanate, first and foremost, from sound critical thinking.


In the end, the prosecution was brought down not by numerical trickery on Hellmann's part but by its own sloppy methodology and faulty reasoning. Any rational appeal judge would have reached the same conclusion.
charlatan
 
Posts: 480
Joined: Fri Oct 07, 2011 12:41 am

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 10:29 am

Jstanz wrote:I am interested in how Thoughtful will respond to the charges that she has important facts about the case wrong. That the information about the case in her book is incorrect. Which would logically lead to incorrect conclusions. Surely if this is true, there is recourse available? I think this is a valid discussion here.


My personal view is that it's straight up libel in English law. Slamdunk case.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby Jstanz » Wed Mar 13, 2013 10:43 am

B_Real wrote:
Jstanz wrote:I am interested in how Thoughtful will respond to the charges that she has important facts about the case wrong. That the information about the case in her book is incorrect. Which would logically lead to incorrect conclusions. Surely if this is true, there is recourse available? I think this is a valid discussion here.


My view is that it's straight up libel in English law. Slamdunk case.

They're playing Russian roulette bringing this inaccurate and highly defamatory guilter crap out this late in the day. This book belongs in 2009 not 2013.


I guess it will depend on exactly what is said. I admit, I have no intention of reading it to find out.
All my grammar and spelling mistakes are the result of auto-correct. If auto-correct is not used here, I still blame it.
Jstanz
 
Posts: 2781
Joined: Mon Aug 08, 2011 8:13 am

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 10:52 am

Jstanz wrote:
B_Real wrote:
Jstanz wrote:I am interested in how Thoughtful will respond to the charges that she has important facts about the case wrong. That the information about the case in her book is incorrect. Which would logically lead to incorrect conclusions. Surely if this is true, there is recourse available? I think this is a valid discussion here.


My view is that it's straight up libel in English law. Slamdunk case.

They're playing Russian roulette bringing this inaccurate and highly defamatory guilter crap out this late in the day. This book belongs in 2009 not 2013.


I guess it will depend on exactly what is said. I admit, I have no intention of reading it to find out.


It's just one more in a huge pile of defamatory books really, pretty much just recycling standard material, so probably nothing will ever happen.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby Katody » Wed Mar 13, 2013 10:58 am

So, Leila Schneps aka 'thoughtful' put in her book that the police "found a speck of Meredith's blood" on the knife. Nice. And that's just one way she utterly and mendaciously misreports the whole argument about the knife.

Funny moment in her book - According to Schneps Rudy's timing of Meredith's death is unreliable because he had the time she came home off by, wait for it.... half an hour!

Anyway, congratulations for cashing in on the still fresh tragedy, Leila. Classy.
Katody
 
Posts: 563
Joined: Sun Mar 06, 2011 2:31 pm

Re: 'Math on Trial'

Postby suomisat » Wed Mar 13, 2013 11:10 am

Katody wrote:So, Leila Schneps aka 'thoughtful' put in her book that the police "found a speck of Meredith's blood" on the knife. Nice. And that's just one way she utterly and mendaciously misreports the whole argument about the knife.

Funny moment in her book - According to Schneps Rudy's timing of Meredith's death is unreliable because he had the time she came home off by, wait for it.... half an hour!

Anyway, congratulations for cashing in on the still fresh tragedy, Leila. Classy.


100% lie. <-- how's that for statistics? Intended to make an innocent person appear guilty.

Truly Machineworthy, repeating known-disproved faktoids to persuade the uninformed and gullible.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby sept79 » Wed Mar 13, 2013 11:13 am

suomisat wrote:.
.
This book could be an entertaining way to teach some statistical concepts. And it might help people see through the deceptive pseudo-mathematical arguments of lawyers, but when it turns to the Kercher case it becomes a piece of deceptive rhetoric that exploits math to support a false conclusion. It becomes exactly what Laila Schneps claims to oppose.


Sort of like that old mathematical axiom ‘lying with statistics'
sept79
 
Posts: 1412
Joined: Wed May 26, 2010 7:16 am

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 11:14 am

suomisat wrote:
Katody wrote:So, Leila Schneps aka 'thoughtful' put in her book that the police "found a speck of Meredith's blood" on the knife. Nice. And that's just one way she utterly and mendaciously misreports the whole argument about the knife.

Funny moment in her book - According to Schneps Rudy's timing of Meredith's death is unreliable because he had the time she came home off by, wait for it.... half an hour!

Anyway, congratulations for cashing in on the still fresh tragedy, Leila. Classy.


100% lie. <-- how's that for statistics? Intended to make an innocent person appear guilty.

Truly Machineworthy, repeating known-disproved faktoids to persuade the uninformed and gullible.


Like I was saying, just that one phrase is a slam dunk libel case if Amanda felt like pursuing it.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby suomisat » Wed Mar 13, 2013 11:25 am

B_Real wrote:
suomisat wrote:
Katody wrote:So, Leila Schneps aka 'thoughtful' put in her book that the police "found a speck of Meredith's blood" on the knife. Nice. And that's just one way she utterly and mendaciously misreports the whole argument about the knife.

Funny moment in her book - According to Schneps Rudy's timing of Meredith's death is unreliable because he had the time she came home off by, wait for it.... half an hour!

Anyway, congratulations for cashing in on the still fresh tragedy, Leila. Classy.


100% lie. <-- how's that for statistics? Intended to make an innocent person appear guilty.

Truly Machineworthy, repeating known-disproved faktoids to persuade the uninformed and gullible.


Like I was saying, just that one phrase is a slam dunk libel case if Amanda felt like pursuing it.


It could get quite funny. "Thoughtful’s" defence council could argue that there's no plausible connection between the knife and the murder, so it's not libel.
My grammar master Yoda is.
suomisat
 
Posts: 1407
Joined: Tue Feb 07, 2012 3:03 am

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 11:57 am

We have had some recent concern from members about the photo posted here of Thoughful's daughter. I do not necessarily agree with the comments that accompany the photo but the photo was not moderated because Thoughtful and her daughter are named authors. The photo was posted on a public website. No one was outed. It would be no different than posting a photo of Candace Dempsey, Douglas Preston, or even me for that matter. The photo below is posted on Amazon. It's in the public already. I would have no complaint if it was posted elsewhere. Of course copyright is an entirely different issue. We will be happy to remove any images on the board if claims of copyright are made.


Image
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby Leila » Wed Mar 13, 2013 1:17 pm

Thoughtful here.

A few responses to comments above. And I'm willing to respond to other specific questions.

Jstanz wrote:
"I think after the acquittal Thoughtful faced quite a dilemma. After all the work they'd put in on the book....what to do? what to do? I don't think she can afford to admit she was wrong with a book on the line."

As a matter of fact, I was hesitating about including this case in the book before the acquittal, because the mathematics that was seriously annoying me at the time was the continued repetition of "There's only a 1% chance that the knife blade DNA belongs to Meredith" etc. that was dominating the subject in the US news. I find these statements, several of which are quoted in the chapter, to be nothing less than purposeful misinformation. I notice that no one here has responded to that part of the chapter. But since that misuse of math wasn't happening in court, I didn't think the case belonged in the book.

However, after Hellmann's motivations report appeared, I was simply astounded by what I saw there. Hellmann had no other reasoning but the one I quote for rejecting a retest. He does make a series of remarks about why it cannot be shown that the knife was really the murder weapon by other considerations than the DNA, but his remarks certainly don't prove that it was not the murder weapon. What he says about the unlikelihood of taking the knife back to the kitchen if it had been really used for a murder is very unconvincing - far more grotesque things have happened in this world. His argument that the knife is not compatible with the wounds because the blade is too long for the depth of the wound is unconvincing, because the knife may well have gone in at an angle or nearly transversally. His argument that the presence of the knife in the cottage would indicate premeditation is also unconvincing, firstly since there might have been premeditation, and secondly because there could have been other reasons for the knife to be in the cottage; Raffaele had cooked there, or it could have been part of a Halloween disguise similar to the meat cleaver picture. I'm looking at Hellmann now (p. 63-65, starting 43 in the PMF translation) and those are all of his reasons for doubting that the knife is the murder weapon. There aren't any others, and none of them come near excluding the knife as the murder weapon. Hellmann then goes on to observe explicitly that if the DNA on the knife is as claimed, that is tantamount to a proof that the knife is the murder weapon. So, according to all this, the question of whether the DNA comes from the blade and whether it is Meredith's, as it undoubtedly is according to the electropherogram, is indeed a major point.

MichaelB wrote:
"It doesn't work like that. It's a half day hearing and there's no waving some silly book written by a woman in Paris and her daughter"

Quite. I only hope that they have time to wave around another silly book crammed with contradictions to previously given evidence and documents: Honor Bound. But in fact, I believe the Court of Cassation will only decide whether the appeal judgment was valid or not, and nothing else.

roteoctober wrote:
"But indeed, and erasmus44 nailed it, the point is about "and even if Meredith's profile was there, then what?" "

That's a good question, and I'll answer it the way I see it: if Meredith's profile is there for certain, it's for a jury to decide whether or not that means the knife was used in her murder, and in fact there's not a chance they would acquit. Only the unreliability of an LCN sample and the possibility of contamination could prevent that, but if a second test showed DNA on the blade that was undoubtedly Meredith's, I believe the case would be closed. Indeed, there is no chance that Meredith's DNA could have gotten onto the knife by contamination during collection or transport in places where Meredith had never been. It beggars belief that amongst all the "bits of DNA floating around" in a forensic laboratory, precisely Meredith's would appear on the knife if the contamination occurred in the lab. If, as was more reasonably suggested, the contamination did not occur on the knife but in the machine, because there was actually nothing on the knife, then a second test would have established that fact clearly as well. As far as the pursuit of truth was concerned, it was a great error not to do it.

Chris wrote:
"I think if anyone argues for another round of testing, he or she should be expected to explain what the interpretation of each possible result will be and why they think so. Would a negative result fall into the exculpatory category? If not, why not?"

It depends what negative result. If there's nothing left on the knife, it would be neither exculpatory nor the contrary: a null result. However, Conti and Vecchiotti did find a tiny quantity on the knife that Novelli claimed was testable. What other negative results could there be? Someone else's DNA?

roteoctober wrote:
"First of all she makes an example arbitrarily attributing to the correctness of the DNA profile on the knife a probability of 80-90% . Well, attributing to them an 80-90% correctness is arbitrary at best."

Sure, it's arbitrary. Looking at the electropherogram, the DNA is 99.99999% certain to be Meredith's. I lowered the probability purposely below the "reasonable doubt" range because LCN samples are generically open to doubt, not because this particular sample is open to doubt. Indeed, if the crude mathematics used in this kind of DNA analysis, that says "peaks below 50RFU shouldn't be counted because they can come from background noise" were replaced by proper mathematics that would consider the average height of background noise peaks in deciding whether a given large peak is "stutter" or a real peak, this sample would not be open to any doubt at all, but Conti and Vecchiotti chose to cite very general safeguard rules without any reference to the example at hand.

roteoctober also wrote:
"The DNA got all the limelight in the debate and apparently also at trial, but H&Z put its value in the correct perspective in their report: whatever you can make of it scientifically, a couple of DNA traces (moreover in extremely small quantities, being at the extreme limits of the possibilities of the machines, if not beyond) don't make a murder weapon out of a knife that everything else we know about it is excluding as the murder weapon.
And indeed this last point would maintain its strength even with a DNA with a 100% reliability.

Nothing whatsoever excludes the knife as the murder weapon; it could have made the wounds (by not going straight in), it could have been in the cottage, it could have been put back in the drawer, even if these things all appear unlikely to some. And if a jury were told that the DNA is 100% reliable, they would certainly agree with me here.

ClintNine wrote:
"But she passes off her critique as if it were unassailable because she is a professor of mathematics."

Our conclusion is unassailable because it's correct: the second test should have been done. Please note that in our book chapter, we have not critiqued Hellmann's judgment in its entirety nor his conclusion. Those who appear to think that our condemnation of his rejection of the second test translates to a rejection of his entire conclusion are welcome to their opinion, but that is not what we wrote, because our book has a narrow focus on mathematical errors.

Dedj wrote:
"Hellmann is not saying that a second independent experiment cannot contribute to the probability that the result of the first experiment is the true result - he's clearly saying that theres no guarantee about the reliability of the experiments and therefore no guarantee that either result is an accurate reflection of the subject being tested."

The phrase "no guarantee about the reliability" makes no mathematical sense. The problem with results from LCN samples is not that there is "no guarantee" but that they are not fully reliable, in that it is possible to misread some peaks as human that are background noise, and vice versa. Thus, in the presence of a readable LCN result, they should be said to be reliable up to a certain percentage that would depend on the quality of the result. In a very legible result such as the knife DNA, the percentage is relatively high and a second test would make it higher.

Dedj further wrote:
"This is elementary GCSE Science - if your method is wonky, you can repeat it a hundred times, you've still no idea if you got the results due to the poor methods or if you correctly tested what was there."

What you say may be elementary GCSE science, but more advanced science will tell you that if you repeat your test a hundred times and compare the results, you can obtain quite a good idea of exactly how wonky your test is, and exactly what results are really present.

erasmus44 wrote:
"In fairness, if the problem is a tendency of the test to produce false positives with a known error rate, then conducting a second, completely independent, test could reduce the probability of error. On the other hand, if you adopt my chain of reasoning and posit that a key problem is the potential that MK's DNA may have gotten on the knife due to either contamination or exposure consistent with Massei's theory that AK carried the knife at all times and therefore it was often over at the cottage, then testing again adds nothing."

This would make sense, and almost be convincing if Amanda had said she carried the knife in her bag at all times. Unfortunately, both she and Raffaele have firmly stated that she did nothing of the kind and that the knife did not leave Raffaele's place. So, if your reasoning is right, it means that they are lying, which doesn't go far to convince anyone of their innocence.

erasmus44 further wrote:
"The problem for the prosecution is that the only plausible way the knife gets over to the cottage is Massei's reasoning that AK carried it at all times."

Well, I can think of at least a couple of other plausible ways, mentioned above.

Dougm wrote:
"It didn't have any blood or even human or animal tissue on it."

Oh? So the human DNA that looked just like Meredith's came from onions?

Chris wrote:
"One is that it fails to address whether or not low template DNA is useful forensically, even when carried out under the most stringent conditions. That is Dr. Budowle’s criticism, as I understand it."

It is obvious that low template DNA is useful forensically, as long as the correct statistical methods are applied to distinguish real peaks from false ones, and a proper criterion of this kind is used to determine when a given peak cannot be determined one way or the other. I realize that those statistical methods for LCN DNA have not been completely developed yet. However, it is clear that in the particular knife DNA sample, we would have to be looking at random background noise peaks that appear quite precisely in the very places where Meredith's peaks should be. The probability of this happening is absurdly low.

B_Real wrote:
"Hellmann said the knife did not match any of the wounds on the victim. That's where we are at the moment, judicially speaking.

Obviously a sane person would start by trying to show the knife actually did match at least one wound, before worrying about the DNA.

There can be a lot of explanations for DNA, including planting it.

There is only one explanation for a knife not fitting the wounds."

Hellmann's conclusion about the knife not fitting the large wound is open to a lot more doubt than the identification of the knife DNA electropherogram with Meredith. But what you say about planting the DNA does indicate your line of thought quite clearly.

MichaelB wrote:
"1. The Giodarno report does not decree that Rudy didn't act alone.

2. The autopsy reports do not "clearly indicate Meredith had been attacked by more than one person".

3. There was no DNA on the knife and it was completely discredited as evidence."

How can I answer these three absolutely wrong points? Go read the autopsy report or its summary in Massei. Learn to spell Giordano and read his report, too. They have been conveniently translated for you over on PMF.

suomisat wrote:
"Her trivial math point sidesteps the real arguments and issues to which she pretends to respond."

You and I clearly don't agree on what the "real arguments and issues" are. It's subjective anyway.

Rose wrote:
"I am interested in how she will respond to the criticism of her chapter on the Kercher case."

This post is why I'm responding.

Jstanz wrote:
"I am interested in how Thoughtful will respond to the charges that she has important facts about the case wrong. That the information about the case in her book is incorrect. Which would logically lead to incorrect conclusions. Surely if this is true, there is recourse available? I think this is a valid discussion here."

I don't think any of the facts I may have wrong in my description of the case has any bearing on my mathematical critique of Hellmann's rejection of the 2nd test. However, I am still concerned by the idea that I may have some facts wrong. It seems that a very unfortunate "speck of blood on the knife" crept into the text, that should have read "speck of DNA". This is indeed an error, but I clearly state in the chapter that tests for blood on the knife sample came up negative, so it is not too serious. I am perfectly willing to discuss other factual errors, if any.

diocletian wrote:
" It's easy, when there is no one there to point out your mistakes, to write a piece putting forth a false narrative of a case clothed in mathematical theory. But if she were to attempt to make these points by testifying as an expert witness in this case, she would be shredded within minutes of taking the stand."

You think? Read some of the other cases in the book. You'd be surprised what kind of math makes its way into trials.
Leila
 
Posts: 659
Joined: Sun Mar 10, 2013 3:48 pm

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 1:25 pm

Leila wrote:Thoughtful here.

A few responses to comments above. And I'm willing to respond to other specific questions.


Let's start with your claim that Meredith's blood was found on the knife. What's the source for that?

Sorry, I see you answered it. What do you plan to do to correct that defamatory false claim since your book has already been printed?

It doesn't matter with regard to that particular claim if you happened to contradict yourself elsewhere in the book.

Welcome to the forum by the way.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby Jstanz » Wed Mar 13, 2013 1:27 pm

Thanks for responding Thoughtful. I think a reasonable discussion between the opposing views is long overdue. I also think you're the person to do it. :smile: I have to be honest with you though, I'm wondering why here now? I'm sure you realize people are going to question you publicly on Amazon.
All my grammar and spelling mistakes are the result of auto-correct. If auto-correct is not used here, I still blame it.
Jstanz
 
Posts: 2781
Joined: Mon Aug 08, 2011 8:13 am

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 1:35 pm

Hi Thoughtful. Welcome to Injustice Anywhere.

Were you able to obtain documents independently or did you rely solely on PMF for your research?
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 1:38 pm

I will say that it's difficult for me to believe the error already noted (DNA not blood) was non-intentional due to the fact that it's been a PMF/TJMK false talking point for years. I have not read Thoughtful's book so my participation will be limited until I have the chance.
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby Dougm » Wed Mar 13, 2013 1:58 pm

Leila wrote:
Dougm wrote:
"It didn't have any blood or even human or animal tissue on it."

Oh? So the human DNA that looked just like Meredith's came from onions?


Welcome Leila, aka Thoughtful. Thanks for posting.

Re: the quote above and your reply. I guess it depends on which "experts" we believe. It is my understanding that H&Z's findings were that there were no human or animal tissue, and no blood, on the knife. They said the only thing they found was starch, vegetable in nature.

Of course, it can be argued that Stefanoni found human tissue, and used it all up in the first test, but it is my understanding that her tests were negative for blood as well.

It is also my understanding that the only way that Stefanoni was able to measure Meredith's DNA was to amplify what she found beyond the recommended limits of the testing equipment, and to then make some semi-arbitrary decisions on what was a legitimate peak and what was noise.

All of this seems to be way too speculative to use as evidence against anyone in a murder trial.
When you berate someone and push them and confuse them and lie to them and convince them that they're wrong you're not finding the truth.

Amanda Knox
Dougm
Moderator
 
Posts: 3189
Joined: Sat Jan 01, 2011 5:00 pm

Re: 'Math on Trial'

Postby erasmus44 » Wed Mar 13, 2013 2:01 pm

"Thoughtful" seems to concede that if Massei's theory is true and AK carried the knife at all times (which would mean that it had been at the cottage numerous times before the night of the crime) then there were opportunities for MK's DNA to get onto the knife completely independently of the homicide. Thoughtful's response seems to be along the following lines - 1. AK denied carrying the knife at all times and AK must have been telling the truth and 2. they may have premeditated the murder and carried the knife over there in order to use it to kill MK or 3. they may have used the knife as part of a Halloween costume.
Wow! So I guess Thoughtful rejects Massei's hypothesis about the knife. First of all, it is possible that AK could have lied about carrying the knife and still be innocent because she thought that admitting carrying the knife would be incriminating. The Halloween theory doesn't work because Halloween was over by the time of the crime and a kitchen knife did not really fit into AK's Halloween costume. Premeditation has been rejected by Mignini and Massei as implausible and given their proclivity to bend the facts in every possible direction against these defendants, it must be REALLY REALLY implausible if Massei and Mignini don't even try to argue for it.
So we are left with three alternatives (if the test is positive) - none of which are particularly plausible. 1. It was a premeditated and planned crime - they threw out their bloody clothes and the "other knife" but not this one - and Mignini and Massei failed to pick up on the fact that it was premeditated (which likely would have been the basis for a harsher sentence) or 2. AK carried the knife at all times, lied about it, and either the knife picked up DNA innocently or the knife was used in the crime and picked up DNA at that time and was not disposed of, although the bloody clothes and the cell phones and the "other knife" were apparently disposed of, or 3. The knife picked up DNA somewhere in the process of handling between the police, the lab etc.

I will admit that none of these explanations are particularly plausible but it is also the case that a finding of DNA on the knife (while constituting incriminating evidence) does not at all lead to the conclusion that they are guilty beyond a reasonable doubt.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 2:22 pm

Leila wrote:Hellmann's conclusion about the knife not fitting the large wound is open to a lot more doubt than the identification of the knife DNA electropherogram with Meredith. But what you say about planting the DNA does indicate your line of thought quite clearly.


But you have to refute Hellmann's conclusion convincingly first otherwise you're wasting your time.

Nobody disputes that a printout was produced which appears to match Meredith's profile.

The problem is the printout is scientifically worthless.

Nonsense claims of finding cells in a groove on the blade, no evidence any DNA was ever recovered from the knife, no recognised procedures followed, all peaks below recognised limits, some peaks ignored.

And didn't they play pass the parcel with the knife in the good old Perugia Questura before sending it off to Rome in a calendar box? What was that about?

I definitely would never want to suggest there is any corruption in the Perugia police. I wouldn't want to get arrested by Monica Napoleoni or her hefty sidekick Lorena Zugarini, would I? Oh wait... ::razz::
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby diocletian » Wed Mar 13, 2013 2:38 pm

Leila wrote:You think? Read some of the other cases in the book. You'd be surprised what kind of math makes its way into trials.


Indeed I do. An expert who relies on a erroneous foundation is highly vulnerable. And, of course math comes into play in many trials (as well as many other aspects of life), but if math were detrminative of trials then we would conduct trial by computer and not by jury. It doesn't work that way for a reason.
diocletian
 
Posts: 1275
Joined: Thu Dec 15, 2011 10:09 am

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 2:55 pm

Leila wrote: The problem with results from LCN samples is not that there is "no guarantee" but that they are not fully reliable, in that it is possible to misread some peaks as human that are background noise, and vice versa. Thus, in the presence of a readable LCN result, they should be said to be reliable up to a certain percentage that would depend on the quality of the result. .


They have zero reliabilty if you don't do them properly in the right lab.

This is what you ignore. It's not about whether LCN is valid in principle. It's whether Stefanoni followed any recognised procedures.

The experts say she did not.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby roteoctober » Wed Mar 13, 2013 3:00 pm

Leila-Thoughtful wrote: roteoctober also wrote:
"The DNA got all the limelight in the debate and apparently also at trial, but H&Z put its value in the correct perspective in their report: whatever you can make of it scientifically, a couple of DNA traces (moreover in extremely small quantities, being at the extreme limits of the possibilities of the machines, if not beyond) don't make a murder weapon out of a knife that everything else we know about it is excluding as the murder weapon.
And indeed this last point would maintain its strength even with a DNA with a 100% reliability.

Nothing whatsoever excludes the knife as the murder weapon; it could have made the wounds (by not going straight in), it could have been in the cottage, it could have been put back in the drawer, even if these things all appear unlikely to some. And if a jury were told that the DNA is 100% reliable, they would certainly agree with me here.


It is going to be very long, so let's start from somewhere.

Nothing excludes the knife as the murder weapon? Not with a 100% probability but with a rather high one yes. I'm not going to give numbers because they would not be very meaningful and certainly they are not used by judges in Italian rulings.

Concerning the two smaller wounds they are not compatible with the kitchen knife, because if inserted to the required depth it would cause a too wide wound.

Hellmann wrote: Firstly, in fact, said knife has not been deemed compatible with the other wounds present on the body of Meredith Kercher, clearly inflicted with a smaller knife, insomuch as the Corte di Assise of first level had to hypothesize the use of multiple knives, and the presence of multiple aggressors, because otherwise the supposed compatibility of the seized knife with the wounds would have failed.

Secondly, on the mattress cover, in the victim’s room, a bloody imprint was detected which clearly matches a knife of lesser dimensions.


About the major wound:

Hellmann wrote: But even the major wound (in the left laterocervical region of the neck) – the one that, according to the Corte di Assise of first level, should have been produced by the knife under discussion – would however not have been caused, according to the defense consultants, by that knife, since the path [of the wound] appears to be noticeably shorter (8 cm as opposed to the 17.5 cm length of the blade) and the presence of an ecchymotic area [bruise] beneath the wound, corresponding to the entry of the blade, would mean that the handle ended up hitting [sbattere] [the flesh] at that point, the blade having been inserted in its entire length.
These arguments, in truth, seem more convincing than those produced by the consultants of the Public Minister – who explain it [ecchymotic area] with a direct action of the aggressors’ hands – which does not seem compatible with the placement of the bruise exactly next by the wound, and even less [compatible] with the explanation proposed by the civil-party attorney, Ms. Perna, during the rebuttal: having been produced by the impact in that place not of the handle but of the fingers of the hand holding the weapon. And in truth, a longer blade, like that of the seized knife, should have, logically, kept the hand at a greater distance from the victim’s body, it not being plausible that the hand could have been nearer to the blade than the handle itself. [That is, the hand holding the handle can’t get anywhere nearer to the blade than the point where the handle touches the blade.]

Moreover – the defense consultants, and particularly professor Torre, observe – since in reality the main wound results from the reiteration of multiple strokes, so much so that it laterally appears widely diastasized - it seems even more difficult to hypothesize that a knife with a 17.5 cm blade was introduced multiple times for a length of only 8 cm.


So, while it is not ruled out as absolutely impossible, it is certainly deemed highly unlikely.

About the logic applied by Hellmann to "calculate" (improper term because they certainly did not do any calculation of that kind) the probability of the correctness of the attribution to Meredith Kercher of the DNA on the blade, he seems to apply for most of Section 11 of the ruling a binary logic: if the profile is obtained through multiple amplifications then probability=1 (100%), if it is obtained through single amplification then probability=0 (0%). He applies the same logic to each sample or trace taken separately and obviously with events with 0 probability, having two of them does not change much (or if you prefer he applied a boolean AND to the two zeros).

This is, in my opinion, the logic applied in the following passage:

Hellmann wrote:This also explains why the Expert Panel did not proceed further in analysing the sample it collected from the blade of the knife: the quantity was determined to be wholly insufficient (again, LCN) for performing two amplifications, so that if they had continued further, the court-appointed experts would have committed the same error found in the investigations [accertamenti] performed by the Scientific Police. Furthermore, it is clear from the concepts outlined above that the need to divide the sample into several parts applies to each individual trace, the intent being to guarantee the reliability of the result of the analysis of that trace. Hence, analysing two individual traces, both LCN, without subjecting each to that procedure which guarantees the result, does not compensate for the failure to repeat the procedure on each individual trace: the sum of two unreliable results, neither having been subjected to a scientifically correct procedure, cannot give a reliable result, regardless of the possible similarities.


Of course the concept of applying a binary logic may appear extreme to some and so Hellmann, after having advocated it once more:

Hellmann wrote: It should again be observed that, since the unreliability of the result primarily renders the circumstantial evidence which should be represented by that result materially non-existent, rather than equivocal in its significance (in fact, it is difficult to understand how one can be scientific by halves: the result is scientifically accurate or it is not, in a true sense, a result) any attempt to clarify its significance by placing it in the context of other results from the proceedings should be resisted, once its unreliability has been determined.


accepts to analyze the other known elements concerning the knife to see if they somehow can support the idea that the knife may be a piece of circumstantial evidence:

Hellmann wrote: Nonetheless, if the aforementioned result is placed in the context of other results from the present proceedings, its unreliability remains definitively proven.


and then he goes on quoting starch, absence of blood, failure by the cytomorphological test to detect any cellular material and so on.

This in terms of Italian law means to point out that even if that DNA were really there, it would not be in agreement (concordant) with the other known elements concerning the knife (in the meaning of article 192 of the penal procedural code).

Now, this is rather important, since for a piece of not admitted evidence to justify the decision of a new trial by the SC (here the not admitted evidence is the testing requested by the prosecution of trace I) it has to be, under article 606 section D of the penal procedural code, what is termed "decisive" evidence.
In the end it is up to the SC to decide what is decisive and what is not, however an element of evidence conflicting with other known elements concerning that evidence is unlikely to be considered "decisive".

This said, you claim that a 100% certain profile of Meredith on the knife would make a jury agree with you. A popular jury impressed by the CSI series possibly yes, an SC panel much less, also because they mainly reason in juridical rather than scientific (or pseudoscientific) terms.

But in any case since you used many times the word "could" (the knife could have been carried to the cottage, there could have been premeditation etc.), well I'll use it myself and I'll say that that DNA could have ended there through contamination/secondary transfer.

Trace I is made of 50pg of DNA (according to Vecchiotti in the hearing of July 5 2011) and the one found by Stefanoni was not quantified but in all probabilty in the same range (some even doubt that it ever existed, but that's another story), now from Professor Novelli himself we know that 8-10pg is the quantity of DNA in a cell: that means that at best (taking for good Stefanoni's trace and assuming that trace I will give Meredith's DNA) there was something like a dozen Meredith's cells on that knife.

To justify the presence of a dozen cells on that knife improper application of procedures by the ordinary police (the knife was handled entirely by ordinary cops, one of them having been at Via della Pergola on November 6, before being sent to the labs in Rome) would be enough, however with so small quantities secondary transfer through Amanda, even if the knife never left Sollecito's flat, could have been ruled out only if all the flatware at Sollecito's had been tested and found negative.

As a last note, "could" is not enough to convict, at least not under Italian law:

Article 530 section 2 penal procedural code wrote: The judge pronounces a judgement of acquittal also when the evidence is missing, insufficient or contradictory


where "contradictory" means that some elements of evidence are in favor and some are against the defendant.
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby Katody » Wed Mar 13, 2013 3:03 pm

Leila wrote:His argument that the knife is not compatible with the wounds because the blade is too long for the depth of the wound is unconvincing, because the knife may well have gone in at an angle or nearly transversally.

The depth of the wound was 8 cm, so was its length at the entrance. There is no angle that would accommodate a 17.5 cm blade.

Both the angle and the depth of the wound was known from the autopsy and no one raised such an argument. A slender girl's neck might be 10 cm across. This argument of yours is completely detached from reality and from what was shown and discussed in the courtroom, Leila.
Katody
 
Posts: 563
Joined: Sun Mar 06, 2011 2:31 pm

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 3:19 pm

There are many great points here. Let's give Leila a chance to respond to the posts we currently have so that we can have a good discussion without anyone's post being overlooked. keep in mind that we have a group of people posting to one person.
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby Dedj » Wed Mar 13, 2013 3:24 pm

"Dedj further wrote:
"This is elementary GCSE Science - if your method is wonky, you can repeat it a hundred times, you've still no idea if you got the results due to the poor methods or if you correctly tested what was there."

What you say may be elementary GCSE science, but more advanced science will tell you that if you repeat your test a hundred times and compare the results, you can obtain quite a good idea of exactly how wonky your test is, and exactly what results are really present."

But this only holds if you have a reliable method for understanding why your results were off and a means of working out how much.

If the test is off by the same amount each time, you can recalibrate to compensate. If the test is accurate, but your methods of sourcing the test subject means you cannot make an estimation of the reliability of the subject being a true-unbiased subject, then your test can be 100% accurate at all times, but still not be fit to support a hypothesis.

You cannot argue that the interpretation of the results must be valid because the tests are and that the sampling stage must be valid because the tests are - you have to show that the whole process of pre-test and test is valid. If your sampling is off, you'll only know if your test shows something unusual, like unexpected peaks on the bra-profile, otherwise you'll get the same useless result twice.
Dedj
 
Posts: 385
Joined: Tue Jul 10, 2012 2:10 pm

Re: 'Math on Trial'

Postby Dedj » Wed Mar 13, 2013 3:27 pm

B_Real wrote:It's not about whether LCN is valid in principle.


Bingo!

Arguing about the probability of the LCN DNA results being/not being Merediths is a totally different topic to arguing about whether or not the LCN DNA results are indicative of murder or contamination when taking into account the test and pre-test proceedures.
Dedj
 
Posts: 385
Joined: Tue Jul 10, 2012 2:10 pm

Re: 'Math on Trial'

Postby sept79 » Wed Mar 13, 2013 3:31 pm

Maybe not what anyone wants to hear, but I find it incomprehensible that the knife DNA is still being argued as a viable piece of evidence against Amanda Knox:

-- Have we all forgotten that Stefanoni lied about the sample size in front of Judge Micheli at the pretrial hearing? I don't give her the benefit of the doubt; she lied. Why? Was it because she didn't wish to defend her testing procedures which didn't even come close to adhering to international standards for LCN DNA?

-- Why did Stefanoni conceal her DNA testing procedures from Judge Massei's court? Why did she conceal the TMB testing which was performed on the footprints in the hallway and else where? Why was she being so deceitful? Why did Massei's court believe anything that she said?

-- Why did Stefanoni continue testing for DNA on the knife after receiving numerous 'too low' read outs?

-- After Massei ordered Stefanoni to release DNA testing information to the court (at least 6 months into the hearings), why did Stefanoni still withhold the electronic data files from the court?

-- After ordering an independent study of the DNA collection/testing, why did Stefanoni delay in getting her testing information to Stefano Conti and Carla Vecchiotti?

-- Why did the prosecution protest so vehemently about the opening up of the knife which Conti-Vecchiotti requested? Is it because they knew the knife to be a fraud?

-- Again, why is the knife still in play when the Conti-Vecchiotti study concluded that that the technical analysis is not reliable for the following reasons:

1. There does not exist evidence which scientifically confirms that trace B (blade of knife) is the product of blood.

2. The electrophoretic profiles exhibited reveal that the sample indicated by the letter B (blade of knife) was a Low Copy Number (LCN) sample, and, as such, all of the precautions indicated by the international scientific community should have been applied.

3. Taking into account that none of the recommendations of the international scientific community relative to the treatment of Low Copy Number (LCN) samples were followed, we do not accept the conclusions regarding the certain attribution of the profile found on trace B (blade of knife) to the victim Meredith Susanna Cara Kercher, since the genetic profile, as obtained, appears unreliable insofar as it is not supported by scientifically validated analysis;

4. International protocols of inspection, collection, and sampling were not followed;

5. It cannot be ruled out that the result obtained from sample B (blade of knife) derives from contamination in some phase of the collection and/or handling and/or analyses performed.

-- Let's not forget that the knife does not match all of the wounds found on the victim and the knife does not match the bloody silhouette of the actual knife used to murder Meredith Kercher!

-- And, finally, a question for anyone arguing the efficacy of the knife as the murder weapon, "Would you personally find it fair and honorable to be convicted of murder as the result of a finding comparable to Item 36, trace B which for all practical purposes convicted Amanda Knox in Massei's court?" I didn't think so!
sept79
 
Posts: 1412
Joined: Wed May 26, 2010 7:16 am

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 3:37 pm

i have worked in laboratories before.

Not with DNA machines, but with chromatography machines.

The machine is first calibrated by someone who knows what they are doing, testing samples with known quantities and tweaking settings until an accurate result is obtained.

You do not mess with the settings once they have been calibrated. You cannot touch them unless you yourself are calibrating the machine.

The idea of somebody just turning dials up in order to get a 'result' must surely be laughable to anyone who has worked in a lab.

And thoughtful's statement that Stefanoni immediately began comparing the printout from the machine with Meredith's DNA profile is exactly what you're not supposed to do. You're supposed to analyse the printout before even looking at the target profile.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby erasmus44 » Wed Mar 13, 2013 4:11 pm

A couple of points from the book. 1. p. 64 - "The results of the autopsy revealed that she (Meredith) had been attacked by more than one person." This is simply incorrect. In fact, the number of attackers is a hotly contested issue with expert testimony on both sides of the question - Massei, while concluding that there were multiple attackers, concedes that the forensic evidence was perfectly consistent with a single attacker. 2. The break-in was a hotly contested issue and I think Hendry's analysis is the most credible. The police did not do the detailed type of crime scene investigation of the area where the break in occurred (or was staged) which would allow a firm conclusion to be reached. 3. The police did not do a really good job of preserving and collecting evidence at the crime scene and this casts doubt upon much of the other evidence submitted. 4. The refusal of the crime lab to make its data available to the defense and the court also tends to cast doubt upon the results - why do this?
Mathematics is useful but it is important to respect other disciplines - thus, the question of the validity, danger of contamination, correct procedure, etc. for DNA testing is not really a mathematical question - it is a question for the DNA testing experts.
Finally, a great deal depends upon your approach to calculating the probability of various results. The Bayesian approach suggests the development of a preliminary probability and the application of new data to modify that probability. If Thoughtful read PMF and TJMK and used them as her sources to reach the conclusion that all sorts of contested issues were settled against the defendants, then she approached the knife with the assumption that there was a high probability that they are guilty.
I have demonstrated above that a positive DNA test on the knife (even if absolutely impeccable as a matter of DNA testing) could be consistent with guilt or innocence. It is very likely that your conclusion as to the relative likelihood of these two possibilities depends heavily upon the preliminary probability you assign to guilt.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 4:30 pm

Leila wrote:However, Conti and Vecchiotti did find a tiny quantity on the knife that Novelli claimed was testable. What other negative results could there be? Someone else's DNA?


They found 'something'. The 'something' tested negative for human DNA, blood, you name it. That's my understanding.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby Leila » Wed Mar 13, 2013 4:44 pm

Let me respond to some of these. I’ll do my best, but there are too many comments and I am not going to be able to keep responding at this rhythm for long. It takes a lot of time.


B_Real wrote: “Let's start with your claim that Meredith's blood was found on the knife. What's the source for that?
Sorry, I see you answered it. What do you plan to do to correct that defamatory false claim since your book has already been printed?”

I don’t intend to take any particular action. If a second edition of the book comes out, I will try to have the “speck of blood” corrected to “speck of DNA”. If the issue comes up in any public place, I will explain. I do point out that the chapter also contains the passage: “She took swabs along these streaks, managed to obtain an infinitesimal quantity of biological material – human cells, in fact, but precious few of them – and proceeded to perform two analyses on it: one to determine whether or not the cells were blood cells, and the other to obtain a DNA identification. She also swabbed the handle of the knife. The DNA on the handle turned out to be Amanda’s, but this was not considered incriminating, as she had used the knife while cooking at Raffaele’s house. The DNA on the blade, however, was a different story. The first test could not establish whether or not the cells were from blood. They were human cells, but they might have been from skin or tissue.”

Jstanz wrote: “I have to be honest with you though, I'm wondering why here now? I'm sure you realize people are going to question you publicly on Amazon.”

Well, people are questioning me here and now, and not on Amazon. Anyway, I don’t consider Amazon a place to have a discussion. If people on Amazon want to talk to me about the book, they are welcome to post on our blog “Maths on Trial”. By the way, I saw on another thread that someone appears to believe that I suppressed the comments posted on that blog by a certain “Eric”. In fact, Eric twice posted comments and then suppressed his own comments almost immediately. I have no idea why. But since Coralie gets the comments in her email, she sent them on to me, and I responded anyway. There was nothing particularly offensive about them, and I would certainly not have deleted them myself.

Bruce Fischer wrote: “Were you able to obtain documents independently or did you rely solely on PMF for your research?”

Presumably you are only referring to the chapter on Meredith’s murder. Each case in the book represented a significant document hunt. For Meredith, I obtained the electropherograms, Massei’s motivations and what court transcripts I have from sources other than PMF. I got Hellmann, Galati, Conti-Vecchiotti and Giordano from PMF. Information from the press came from all over the place. I was in personal contact with Antioco Fois. I visited Perugia myself, for other reasons, but I took the time to explore the distances, check the views, assess the possibility of climbing the cottage wall (I’m no expert, but it seems just about impossible to get up and in even once you’ve got a hand on the windowsill, though Reinhold Messner could probably do it with ease) and so forth. No need for PMF to see what happens on Oprah Winfrey, The View etc. not to mention Oggi. All this being said, I can’t see why this is important.

Bruce also wrote: “I will say that it's difficult for me to believe the error already noted (DNA not blood) was non-intentional due to the fact that it's been a PMF/TJMK false talking point for years.”

Well, it really was unintentional. I’ve just checked back in our original draft and it said “a speck of Meredith on the blade”. At some point in the very long and repetitive process of correcting and recorrecting proofs, which involved us two as well as two proofreaders from the publisher, somebody did not like this phrase, which now that I re-read it, is not very elegant. It got modified, and it got past me. I regret that.

Dougm wrote: “It is my understanding that H&Z's findings were that there were no human or animal tissue, and no blood, on the knife. They said the only thing they found was starch, vegetable in nature.”

This is not right, and in fact, this is one of the misunderstandings that pervade pro-Amanda discussions. I welcome this opportunity to set it right. Here is what Hellmann says about the experts’ findings:

From Hellmann (p. 74 original, p. 50 translation): “In what concerns exhibit 36 (the knife), the expert team, after proceeding to take samples on swabs and perform the DNA extraction, in itself certainly correct and not criticized by anyone, found that the DNA extracted during the course of the expert operations and quantified according to currently used methods was not useful, because of the fact that the quantity was too scarce to perform the successive steps represented by amplification and electrophoresis.”

Professor Novelli claimed subsequently that the quantity found by Conti & Vecchiotti was testable by his up-to-date methods. It was not zero.

Dougm further wrote: “It is also my understanding that the only way that Stefanoni was able to measure Meredith's DNA was to amplify what she found beyond the recommended limits of the testing equipment, and to then make some semi-arbitrary decisions on what was a legitimate peak and what was noise.”

Your description is more or less correct in what concerns the bra clasp DNA. Indeed, there are extra peaks there which are high enough to be considered, and yet she chose to view them as background. This is disputable, yet at the same time Raffaele’s Y-haplotype profile is not only clear on the bra clasp, but in a significantly greater quantity (represented by higher peaks) than those other spurious peaks that could represent other contributors.

But none of this is the case for the knife electropherogram. Have you seen it? Stefanoni did amplify beyond the recommended limits, but the machine did the job. The result she obtained is clear as a bell; no arbitrary decisions were needed. The only big peaks correspond to Meredith’s; there are no other individuals present. Look at the graph.

Erasmus44 wrote: “First of all, it is possible that AK could have lied about carrying the knife and still be innocent because she thought that admitting carrying the knife would be incriminating. The Halloween theory doesn't work because Halloween was over by the time of the crime and a kitchen knife did not really fit into AK's Halloween costume.”

Not for a single second do I believe that Amanda would have accepted a 26 year prison sentence so as not to admit that she carried the knife around in her bag. And people can still have Halloween fun the next day, it being a holiday. And, although this is nothing but my imagination and just one possibility among many others, I was envisioning the knife as part of a costume Raffaele might use, like the meat-cleaver picture.

As for your three possibilities which are really four: (1) it was premeditated, (2) Amanda carried the knife around, (3) It was used in the crime but not disposed of although other things were disposed of, (4) it was contaminated, I find (3) the most likely. I don’t believe there was a lot of throwing away of bloody clothes. I believe there was another knife (if not two others), that left the knife print on the bed, and that Rudy had it and threw it away along with his own bloody shoes.

B_Real wrote: “But you have to refute Hellmann's conclusion convincingly first otherwise you're wasting your time.”

In your opinion. As far as Coralie and I are concerned, raising awareness of the issue of mathematical errors in trials is not a waste of time, in fact it’s a goal that transcends the desire to convince anyone of what I may think happened in one particular case. Refuting Hellmann’s conclusion is not part of the purpose of our book. Certainly, I would have much to say about other parts of Hellmann’s report (and nor do I agree with everything assumed by Massei), but that is neither here nor there, since this thread is discussing the book.

B_Real further wrote: “The problem is the printout is scientifically worthless. Nonsense claims of finding cells in a groove on the blade, no evidence any DNA was ever recovered from the knife, no recognised procedures followed, all peaks below recognized limits, some peaks ignored. And didn't they play pass the parcel with the knife in the good old Perugia Questura before sending it off to Rome in a calendar box? What was that about?”

The electropherogram is undoubted evidence that DNA was recovered from the blade. And if you do not consider Hellmann’s report above of what Conti and Vecchiotti further found on the blade as evidence that something was there, then we have no basis for discussion and will simply have to agree to disagree.

No peaks are ignored. As for your statement that all peaks are below recognized limits, the problem there is that such peaks could come from background “stutter”. But it is impossible that random background stutter could have produced bumps in the 26 exact places where Meredith’s alleles are positioned and nowhere else (in fact 34 places; Italians analyze 17 peak pairs where a standard American analysis uses 13). The guideline you mention is useful when there are unexplained peaks that could be human or stutter and it is not known which they are. That is simply not the case in the knife electropherogram.

I have heard that the knife was packed in an ordinary new envelope or package, which they then placed in a box to be transferred to the lab. This is not standard procedure, but there is obviously no way that Meredith’s DNA was on a new envelope used by police in Raffaele’s house. If contamination occurred because of poor transport conditions or poor lab cleanliness or whatever, it could have been anyone else’s DNA, but I really don’t see how it could have been Meredith’s.

Diocletian wrote: “If math were determinative of trials then we would conduct trial by computer and not by jury. It doesn't work that way for a reason.”

It doesn’t work that way because of history and tradition. But DNA analysis is new and cannot be performed by traditional methods of argumentation. You will probably see in your lifetime a combination of Bayesian analysis by computer together with regular jury deliberation enter the courts. There’s going to be no choice about this with increasingly delicate forensic science being used in the courtroom.

Katody wrote: “The depth of the wound was 8 cm, so was its length at the entrance. There is no angle that would accommodate a 17.5 cm blade.”

There were experts arguing both sides of this at the first trial, so as far as I am concerned, I have to use my own judgment. And my own judgment says that a big knife can make a huge, jagged, wide-open wound. I think (in accordance with some of the experts and not others) that it was partly in, and was sawing back and forth because of the struggle, and causing abrasions on the edge.

Anyway, between the certainty of a DNA sample on the blade and the total uncertainty of your claiming to know every shape of every wound that can be made by one struggling person pushing a large knife against another struggling person, I, speaking as a potential jury member, am going to go with the DNA sample every time.

Roteoctober wrote: “About the logic applied by Hellmann to "calculate" (improper term because they certainly did not do any calculation of that kind) the probability of the correctness of the attribution to Meredith Kercher of the DNA on the blade, he seems to apply for most of Section 11 of the ruling a binary logic: if the profile is obtained through multiple amplifications then probability=1 (100%), if it is obtained through single amplification then probability=0 (0%). He applies the same logic to each sample or trace taken separately and obviously with events with 0 probability, having two of them does not change much (or if you prefer he applied a boolean AND to the two zeros).

If Hellmann really assumed a reliability of 0% for a single amplification, then he’d be far wronger, mathematically, than I even thought. A single amplification gives good results; a second one is only for extra certainty. This said, I don’t think you’re doing him justice. Here’s what he actually says: “In substance, this Court holds that the risk of obtaining a result that is not particularly reliable, not having been obtained via correct methodology (in particular because two amplifications were not done) though the quantity of extract was small (LCN), it could be accepted merely for orienting purposes in a 360° investigation – as one says – but cannot be accepted when it is a question of basing a proof of guilt beyond every reasonable doubt on the result of the genetic expertise.”

I agree with this assessment of Hellmann, and it is why Coralie and I chose to assess the reliability of the result as around 80-90%. “Not particularly reliable”, “acceptable for orienting purposes” and “not accepted as proof beyong every reasonable doubt” were the ideas that we wanted to express in attributing a numerical percentage to the reliability of the test.

Roteoctober wrote: “This said, you claim that a 100% certain profile of Meredith on the knife would make a jury agree with you. A popular jury impressed by the CSI series possibly yes, an SC panel much less, also because they mainly reason in juridical rather than scientific (or pseudoscientific) terms.”

I agree with this. The SC has only to judge whether the appeal court did a proper job or not.

B_Real wrote: “And thoughtful's statement that Stefanoni immediately began comparing the printout from the machine with Meredith's DNA profile is exactly what you're not supposed to do. You're supposed to analyse the printout before even looking at the target profile.”

I have no idea in what order Stefanoni proceeded, but I can assure you that no matter what she did, the form of the printout speaks for itself. What do you want to analyse on this printout? No drop-out, no extras, no background, no interpretation necessary. Write down the peak locations and you’re done.

Erasmus 44 wrote: “Mathematics is useful but it is important to respect other disciplines - thus, the question of the validity, danger of contamination, correct procedure, etc. for DNA testing is not really a mathematical question - it is a question for the DNA testing experts.”

There are two parts to DNA testing: using the machines, which is not that difficult, but yes, you need to know what you are doing in the forensic lab, and interpreting the output, which is neither more nor less than mathematics. It is obvious that the mathematical techniques used by DNA experts are not sufficiently well-developed yet. It’s a cutting-edge science, but it will come.

Erasmus44 further wrote: “I have demonstrated above that a positive DNA test on the knife (even if absolutely impeccable as a matter of DNA testing) could be consistent with guilt or innocence. It is very likely that your conclusion as to the relative likelihood of these two possibilities depends heavily upon the preliminary probability you assign to guilt.”

It seems to have escaped everyone’s notice that the purpose of the chapter was not to assign guilt or innocence, and that we did not do so. The actual focus of the chapter is Hellmann’s wrong decision about the second test, through which crucial information might possibly have been obtained. It belongs in the context of an entire book with ten chapters all devoted to various kinds of mathematical problems in trials.
Leila
 
Posts: 659
Joined: Sun Mar 10, 2013 3:48 pm

Re: 'Math on Trial'

Postby roteoctober » Wed Mar 13, 2013 4:47 pm

B_Real wrote:
Leila wrote:However, Conti and Vecchiotti did find a tiny quantity on the knife that Novelli claimed was testable. What other negative results could there be? Someone else's DNA?


They found 'something'. The 'something' tested negative for human DNA, blood, you name it. That's my understanding.


Of course it could be someone else's DNA, at the very least Amanda's and/or Raffaele's, or the maid's.

By the way, reading the tables from the CV Report, I understand that trace I was divided in three wells and one of them tested positive for Y-chromosome, so it is a mixture.
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby MustBeQuantum » Wed Mar 13, 2013 4:56 pm

Thank you, Thoughtful, for your explanations and analysis.

As always, Billy Preston said it best:

Nothing from nothing means nothing.

Roteoctober, thank you for putting things so perfectly.
MustBeQuantum
 
Posts: 1528
Joined: Fri Oct 28, 2011 9:57 am

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 5:02 pm

Leila wrote:I don’t intend to take any particular action. If a second edition of the book comes out, I will try to have the “speck of blood” corrected to “speck of DNA”. If the issue comes up in any public place, I will explain.


Good luck with that approach.

Didn't work too well for the publishers who insinuated that one of the wrongly convicted Birimingham Six might not have been wrongly convicted.

Your falsehood about the blood seems more libelous to me by the way, because it's about a specific fact. Of course I am not a lawyer.



http://www.bbc.co.uk/news/uk-northern-ireland-11523913" onclick="window.open(this.href);return false;" onclick="window.open(this.href);return false;" onclick="window.open(this.href);return false;" onclick="window.open(this.href);return false;" onclick="window.open(this.href);return false;

Editor and publisher apologise to 'Birmingham six' man

A man wrongly convicted of the 1974 Birmingham pub bombings has accepted undisclosed libel damages following a suggestion there were grounds to believe he was guilty.

Paddy Hill spent 16 years in prison before his conviction was quashed in 1991.

He took High Court proceedings over a reference to him in a book of Irish poetry.

His lawyer said it was seriously defamatory to Mr Hill.

Mr Hill took legal action after he was mentioned in 'An Anthology Of Modern Irish Poetry'.

In the preface to a selection of poems by Pearse Hutchinson - including British Justice dedicated to Mr Hill - editor Wes Davis wrote that Mr Hill was "one of the six men accused, perhaps wrongly, of bombing a pub in Birmingham, England".

Lucy Moorman told Mr Justice Eady, in the court in London that as well as being seriously defamatory to Mr Hill, it had also been extremely damaging to the reputation of the Miscarriages of Justice Organisation (MOJO) which he had set up following his release.

"Mr Hill fought for years to establish his innocence and to rebuild his life and restore his good name," she said.



That was all it took, Leila. Claiming he was accused, perhaps wrongly.

Again, i am not a lawyer.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby roteoctober » Wed Mar 13, 2013 5:10 pm

Leila wrote:Roteoctober wrote: “About the logic applied by Hellmann to "calculate" (improper term because they certainly did not do any calculation of that kind) the probability of the correctness of the attribution to Meredith Kercher of the DNA on the blade, he seems to apply for most of Section 11 of the ruling a binary logic: if the profile is obtained through multiple amplifications then probability=1 (100%), if it is obtained through single amplification then probability=0 (0%). He applies the same logic to each sample or trace taken separately and obviously with events with 0 probability, having two of them does not change much (or if you prefer he applied a boolean AND to the two zeros).

If Hellmann really assumed a reliability of 0% for a single amplification, then he’d be far wronger, mathematically, than I even thought. A single amplification gives good results; a second one is only for extra certainty. This said, I don’t think you’re doing him justice. Here’s what he actually says: “In substance, this Court holds that the risk of obtaining a result that is not particularly reliable, not having been obtained via correct methodology (in particular because two amplifications were not done) though the quantity of extract was small (LCN), it could be accepted merely for orienting purposes in a 360° investigation– as one says – but cannot be accepted when it is a question of basing a proof of guilt beyond every reasonable doubt on the result of the genetic expertise.”

I agree with this assessment of Hellmann, and it is why Coralie and I chose to assess the reliability of the result as around 80-90%. “Not particularly reliable”, “acceptable for orienting purposes” and “not accepted as proof beyong every reasonable doubt” were the ideas that we wanted to express in attributing a numerical percentage to the reliability of the test.


I think he assumed a 0% reliability for a single amplification not in a strictly mathematical sense (i used the binary logic to try to represent what IMO was his reasoning but under a logical-juridical point of view), but in the sense "a single amplification can be used merely for orienting purposes in a 360° investigation, but cannot be accepted when it is a question of basing a proof of guilt beyond every reasonable doubt", in this sense, and only in this one it is to be seen as 0%.

You instead attribute to that a 80-90%...you see: from different assumptions you get different results...

In another post in this thread I used a 20-30% figure (and I was generous) considering that later he says "the failure of this reliability for the reasons outlined, or at least its very serious weakening".

Indeed the result you get as output depends on what you input. :smile:
roteoctober
Tech Director
 
Posts: 2433
Joined: Thu Nov 10, 2011 4:01 pm
Location: Turin - Italy

Re: 'Math on Trial'

Postby erasmus44 » Wed Mar 13, 2013 5:14 pm

I think it is important to be clear about the various possibilities here. Assuming an absolutely reliable DNA test of the knife showing MK's DNA on the knife without any DNA testing issues. We have the following possibilities - 1. pre-test contamination; 2. a premeditated crime in which the knife was purposely taken over there to commit the crime and then, either AK and RF got no blood on their clothes(very unlikely) or got blood on their clothes and threw their clothes out but decided to keep the knife(also unlikely), 3. Amanda carried the knife with her at all times, used it in the crime, and lied about it - still have the two unlikely possibilities with respect to the clothes; 4. Amanda carried the knife with her at all times, had nothing to do with the crime, lied about carrying the knife because she thought that telling the truth about it would incriminate her and increase the probability of her conviction and the knife picked up DNA while at the cottage or from other items in Amanda's bag(admittedly unlikely, but not impossible), 5. Although Halloween is not an Italian holiday and Raffeale apparently did not go out on Halloween night, he dressed up in a costume involving the knife the next night and went over to the cottage in that costume and then committed the homicide (again with the two possibilities about getting blood on the clothes)(almost absurdly unlikely). 6. Amanda carried the knife around at all times, lied about it, committed the crime, but the DNA on the knife was generated in a way not connected with the crime(again, unlikely but not impossible).
I am not suggesting that Amanda accepted a 27 year sentence rather than admit carrying the knife around with her. Just the opposite - there is a possibility that she would lie in the hope that lying about carrying the knife around would reduce the chance that she would be convicted. She might have an incentive to tell this lie whether she was actually guilty or not.
In my opinion, we have at least 5 very unlikely, but not absolutely impossible, possibilities here. Your assessment of their relative likelihood depends a great deal on the initial probability you assign to guilt.

In that sense, it is like what I call the Red Bus Problem. You are hit by a red bus (paint residue is left on your clothing) but you do not know which bus company operated the bus. There are only two bus companies in town - the RED BUS company which operates only red buses and the Technicolor Bus Company which operates buses of a number of colors only 10% of which are red. What is the probability you were hit by a RED BUS company bus?
In fact, it is helpful to know that the RED BUS company actually operates only 5 buses but the Technicolor Bus Company operates 1000 and so the TBC has 100 red buses and it is much more likely you were hit by one of theirs.
Once you admit these various possibilities, then the initial probability you assign to guilt becomes critical.

None of us like to admit uncertainty and it is tempting to think we can reach mathematical certainty by using a DNA test. But in this case it is an illusion and a dangerous one at that.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby Leila » Wed Mar 13, 2013 5:52 pm

B_real wrote: "Good luck with that approach.

Didn't work too well for the publishers who insinuated that one of the wrongly convicted Birimingham Six might not have been wrongly convicted.

Your falsehood about the blood seems more libelous to me by the way, because it's about a specific fact. Of course I am not a lawyer."

You mean you think that I will be sued for the shirt off my back because it says "speck of blood" instead of "speck of DNA"? Sorry, but I cannot take this seriously. Who do you imagine will do the suing? What is the amount of damages you assess?

I'm ready to answer reasonable and interesting questions in a carefully thought-out way, but this strikes me as over-the-top ranting and not worth the bother.
Leila
 
Posts: 659
Joined: Sun Mar 10, 2013 3:48 pm

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 13, 2013 6:02 pm

Leila

What do you think this means?

"[R]ight from the outset we must resist the attempt -- reflected in [literally "pursued by"] the entire structure of the defense case, but out of place in the context of this decision -- to involve the [Cassazione] panel in endorsing the hypothesis that others, Raffaele Sollecito and Amanda Knox, were responsible for the murder, aggravated by sexual violence, of Meredith Kercher. The decision that this Court is called to make concerns only the responsibility of Guede with regard to the act under dispute; any participation of others in the crime will be taken into account only to the extent that this circumstance bears on the matter that is our only concern: the modification or confirmation of the judgement of responsibility of the defendant -- the latter being wholly agreed upon by the judges of the first and second level."

Or if you prefer the PMF translation of the same paragraph. Page 15

In the meantime it is now necessary to escape the attempt, pursued by the overall setting of
the defence, but out of place in the context of this decision, to involve the Court in supporting
the thesis of the responsibility of others, namely Raffaele Sollecito and Amanda Knox, for the
murder aggravated by the sexual assault of Meredith Kercher. The decision to which this
court is called concerns uniquely the responsibility of Guede regarding the deed with which
he is charged, and the possible participation of others in the crime should be taken into
account only to the extent to which such a circumstance would have an impact on the
exclusive commitment of the Court to either modifying or confirming the verdict of guilt of
the defendant, which was entirely shared by the courts of first and second instance.


What do you think Massei means writing this?

The consultants and forensic scientists have asserted that from the point of view of forensic science, it cannot be ruled out that the author of the injuries could have been a single attacker, because the bruises and the wounds from a pointed and cutting weapon are not in themselves incompatible with the action of a single person. With regard to this, it is nevertheless observed that the contribution of each discipline is specifically in the domain of the specific competence of that discipline, and in fact the consultants and forensic experts concentrated their attention on the aspects specifically belonging to forensic science: time of death, cause of death, elements indicating sexual violence, the injuries present on the body of the victim, and the causes and descriptions of these.

The answer given above concerning the possibility of their being inflicted by the action of a single person or by more than one was given in relation to these specific duties and questions, which belong precisely to the domain of forensic science, and the meaning of this answer was thus that there are no scientific elements arising directly from forensic science which could rule out the injuries having been caused by the action of a single person."
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby B_Real » Wed Mar 13, 2013 6:11 pm

Leila wrote:B_real wrote: "Good luck with that approach.

Didn't work too well for the publishers who insinuated that one of the wrongly convicted Birimingham Six might not have been wrongly convicted.

Your falsehood about the blood seems more libelous to me by the way, because it's about a specific fact. Of course I am not a lawyer."

You mean you think that I will be sued for the shirt off my back because it says "speck of blood" instead of "speck of DNA"? Sorry, but I cannot take this seriously. Who do you imagine will do the suing? What is the amount of damages you assess?

I'm ready to answer reasonable and interesting questions in a carefully thought-out way, but this strikes me as over-the-top ranting and not worth the bother.


I'm sure you know who would sue you. As to the damages, for this kind of defamation, could be a million or so plus costs in England.

Anyway it's just a side issue. I'm just saying you are potentialy liable, in my view, for publishing this falsehood about the blood. I am not predicting you will be sued. Let's move on.
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby erasmus44 » Wed Mar 13, 2013 6:19 pm

MichaelB wrote:Leila

What do you think this means?

"[R]ight from the outset we must resist the attempt -- reflected in [literally "pursued by"] the entire structure of the defense case, but out of place in the context of this decision -- to involve the [Cassazione] panel in endorsing the hypothesis that others, Raffaele Sollecito and Amanda Knox, were responsible for the murder, aggravated by sexual violence, of Meredith Kercher. The decision that this Court is called to make concerns only the responsibility of Guede with regard to the act under dispute; any participation of others in the crime will be taken into account only to the extent that this circumstance bears on the matter that is our only concern: the modification or confirmation of the judgement of responsibility of the defendant -- the latter being wholly agreed upon by the judges of the first and second level."

Or PMF translation if you prefer of the same paragraph.

In the meantime it is now necessary to escape the attempt, pursued by the overall setting of
the defence, but out of place in the context of this decision, to involve the Court in supporting
the thesis of the responsibility of others, namely Raffaele Sollecito and Amanda Knox, for the
murder aggravated by the sexual assault of Meredith Kercher. The decision to which this
court is called concerns uniquely the responsibility of Guede regarding the deed with which
he is charged, and the possible participation of others in the crime should be taken into
account only to the extent to which such a circumstance would have an impact on the
exclusive commitment of the Court to either modifying or confirming the verdict of guilt of
the defendant, which was entirely shared by the courts of first and second instance.


What do you think Massei means writing this?

The consultants and forensic scientists have asserted that from the point of view of forensic science, it cannot be ruled out that the author of the injuries could have been a single attacker, because the bruises and the wounds from a pointed and cutting weapon are not in themselves incompatible with the action of a single person. With regard to this, it is nevertheless observed that the contribution of each discipline is specifically in the domain of the specific competence of that discipline, and in fact the consultants and forensic experts concentrated their attention on the aspects specifically belonging to forensic science: time of death, cause of death, elements indicating sexual violence, the injuries present on the body of the victim, and the causes and descriptions of these.

The answer given above concerning the possibility of their being inflicted by the action of a single person or by more than one was given in relation to these specific duties and questions, which belong precisely to the domain of forensic science, and the meaning of this answer was thus that there are no scientific elements arising directly from forensic science which could rule out the injuries having been caused by the action of a single person."




The last two quotes are from Massei who was not taking great pains to make findings helpful to the defendants. The single attacker/multiple attacker issue was hotly disputed by experts and is open to debate.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby Katody » Wed Mar 13, 2013 6:44 pm

Leila wrote:The electropherogram is undoubted evidence that DNA was recovered from the blade.

Of course not. It's only the evidence there was DNA in the sample tested.

Even if the new test come out as positive it wouldn't remove the unreliability of the evidence. The knife was open to contamination during collection, handling at the police station and in the lab, as the court experts demonstrated. No new tests can magically reverse those mistakes or, even more importantly, make the DNA result coherent with the totality of the kitchen knife evidence and the evidence of the actual murder weapon.

The mental gymnastics you're performing to explain away all these implausibilities and inconsistencies, when there is ample and clear evidence of the actual murder weapon should make you stop and think.
Katody
 
Posts: 563
Joined: Sun Mar 06, 2011 2:31 pm

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 7:01 pm

Thoughtful wrote: I visited Perugia myself, for other reasons, but I took the time to explore the distances, check the views, assess the possibility of climbing the cottage wall (I’m no expert, but it seems just about impossible to get up and in even once you’ve got a hand on the windowsill, though Reinhold Messner could probably do it with ease) and so forth.


I find it interesting that our witnesses that have seen the cottage in person say the climb would be no problem. People from PMF that have seen the cottage in person say the climb would be nearly impossible. It shows how people sometimes see what they want to see. The truth is there is no credible evidence of staging and there is a white powdery substance on the floor and clothes in Filomena's room that appears to be from the outside wall that shows someone entered through that window. It would also be a ridiculous window for a resident of the cottage to use to stage a burglary. Not plausible at all. I believe I remember you discussing your trip to Perugia. Are you the person that entered the cottage while it was occupied by tenants to take photos of the kitchen?

(Regarding Meredith's wounds) Thoughtful wrote: There were experts arguing both sides of this at the first trial, so as far as I am concerned, I have to use my own judgment. And my own judgment says that a big knife can make a huge, jagged, wide-open wound. I think (in accordance with some of the experts and not others) that it was partly in, and was sawing back and forth because of the struggle, and causing abrasions on the edge.


Have you seen photos of the wounds? You are describing details about Meredith's wounds that simply do not exist.

Thoughtful wrote: It seems to have escaped everyone’s notice that the purpose of the chapter was not to assign guilt or innocence, and that we did not do so. The actual focus of the chapter is Hellmann’s wrong decision about the second test, through which crucial information might possibly have been obtained. It belongs in the context of an entire book with ten chapters all devoted to various kinds of mathematical problems in trials.


Really? Amanda and Raffaele are free. You are not claiming in your book that the appeal was flawed due to problems with mathematics? What do you expect your readers to conclude about your position on the case? You have actively campaigned for years on a hate site that operates for the sole purpose of attacking Amanda and Raffaele. I have not read your book, but from what I have seen here, I find it very difficult to believe that you were able to write anything about this case in a neutral fashion.
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby Katody » Wed Mar 13, 2013 7:03 pm

Leila wrote:The actual focus of the chapter is Hellmann’s wrong decision about the second test, through which crucial information might possibly have been obtained.

Apart from the doubtful DNA result there were no elements indicating the kitchen knife was a murder weapon.
In fact many elements indicated it wasn't, confirming the contamination hypothesis. There was no crucial information to be obtained and Hellmann was correct in his decision.
Katody
 
Posts: 563
Joined: Sun Mar 06, 2011 2:31 pm

Re: 'Math on Trial'

Postby Katody » Wed Mar 13, 2013 7:12 pm

Leila wrote:Information from the press came from all over the place. I was in personal contact with Antioco Fois.

That's the "apprentice journalist" who procured the unreliable witnesses; the miracle memory man, the miracle ear lady. Can't recall now but wasn't the heroin addict also result of his work?
Katody
 
Posts: 563
Joined: Sun Mar 06, 2011 2:31 pm

Re: 'Math on Trial'

Postby Dougm » Wed Mar 13, 2013 7:48 pm

There has been a discussion here about Hellmann's refusal to allow a "second test" of the DNA on the knife.

It would actually be a third test, if a "test" is defined as having another set of experts look at it. H&Z were supposed to be giving a "third party" decision that was not allowed by Massei. Massei disallowed a defense request to retest the knife, so Hellmann appointed H&Z to examine the knife and give their point of view. They gave the opinion that the first test was not done per accepted standards, and that there is not enough material left to do a retest (following those same accepted international standards).

So now the SC is supposed to allow a "third" test, outside accepted standards, because Novelli says so?
When you berate someone and push them and confuse them and lie to them and convince them that they're wrong you're not finding the truth.

Amanda Knox
Dougm
Moderator
 
Posts: 3189
Joined: Sat Jan 01, 2011 5:00 pm

Re: 'Math on Trial'

Postby erasmus44 » Wed Mar 13, 2013 8:02 pm

A couple of more points:
1. DNA science involves more than just pressing the button on a machine and calling a mathematician into the room to interpret the results - I am not a DNA expert myself so I tend to view a disagreement among the experts as suggesting doubt in the matter but I am fairly certain that there are different ways to conduct the tests and to interpret the results. As so often happens, an expert in one field tends to underestimate the complexity and judgment involved in a totally different field. Can't I figure out all of this math stuff myself? After all, I have a pocket calculator and the battery is charged.
2. Although Laila posts that she isn't opining on guilt or innocence, in fact she posts earlier that the case would be "closed" if the test she proposed showed MK's DNA. I think that this is ridiculous because, at a minimum, one would have to know the chain of custody of the knife and its plausible involvement in the crime. As my earlier posts suggest, this leaves quite us with a choice of a whole array of possible histories (each of which has its own quirky problems and each of which is unlikely, but not impossible for different reasons) leading to the outcome of MK's DNA on the knife only some of which involve the guilt of these defendants. Without assigning preliminary probabilities to these, the case is not "closed" - in fact, it is a long way from resolution - especially, in the context of a reasonable doubt standard.
3. Laila is also too quick to assert expertise in crime scene forensics, wound analysis, and a whole series of other disciplines which contribute to her likely assigning a high probability to guilt and preferring the possible histories involving guilt and discarding the others as "unlikely". This is, of course, the problem. It is why we see in some false confession cases later demonstrated to be clearly wrongful convictions - testimony on technical issues which is incriminating. If you know in advance that the defendant is guilty, then it makes sense to resolve doubts in the direction of guilt.
4. The result is exactly the kind of "rush" to bad judgment the book cautions against. Bad logic, bad law, bad science. Take a step back, smell the roses, and think it all through carefully.
erasmus44
 
Posts: 3138
Joined: Thu Sep 29, 2011 12:10 pm

Re: 'Math on Trial'

Postby Bruce Fischer » Wed Mar 13, 2013 8:13 pm

Thoughtful,

I would never question your intelligence as I have no doubt you are most likely a brilliant mathematician. Your book may very well be a vital learning tool that has been marred by including this case. It's possible for gifted minds to lack common sense. A good example would be knowing whether it is right or wrong to enter a private residence without consent to take photographs. If my recollection of that occasion is wrong then I apologize in advance. I have not gone back to PMF to double check what my memory is telling me.
"This could happen to any one of you. If you don't believe it could happen, you are either misinformed or in a state of deep denial" -- Debra Milke
User avatar
Bruce Fischer
Site Admin
 
Posts: 4470
Joined: Thu May 06, 2010 4:26 pm
Location: USA

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 13, 2013 8:20 pm

This is an incredible statement. :wow:

Thoughtful wrote: I visited Perugia myself, for other reasons, but I took the time to explore the distances, check the views, assess the possibility of climbing the cottage wall (I’m no expert, but it seems just about impossible to get up and in even once you’ve got a hand on the windowsill, though Reinhold Messner could probably do it with ease) and so forth.


Rudy is directly linked to a break-in via a 16 foot window using a rock but when it comes to the cottage, it's all to hard according to you even though it's 13 foot with a metal grate to stand on and lift himself in.

What is it with people like you that when it comes to Rudy, you have to throw all common sense out the window about who he was, what he said and what he was capable of? It's like the other day, you treat Rudy's statements like he's austisic and just a bystander to the murder who can only offer little clues like the cap embellishment when he was implying it was the guy at the fountain. Or it doesnt matter about the law office because there was a balcony so it shouldn't be compared to the cottage at all even though it establishes his MO of 2nd story window entry with a rock just like the cottage break in.

He was an unemployed criminal with no source of income directly and indirectly linked to multiple crimes in the 6 weeks leading upto the murder, including the robbery of his neighbours house where it burnt down and her cat died, yet in your book you just describe him as a "petty thief". A petty thief is someone who steals from a supermarket or a pickpocket not someone illegally entering residential & commerical properties armed with a knife and burglary.

So Rudy can't get up 13 foot with a grate but......

http://www.dailymail.co.uk/news/article-2164119/Elf-safety-Boy-15-climbs-60ft-tree-save-cat-stuck-5-days--As-firemen-nothing.html

15 year old boy climbs 60 foot tree to rescue his cat after fire & rescue said it was to dangerous.

Image

or a more extreme example of a guy climbing a cliff face with no ropes.

Image Image
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby MichaelB » Wed Mar 13, 2013 9:52 pm

Bruce Fischer wrote:We have had some recent concern from members about the photo posted here of Thoughful's daughter. I do not necessarily agree with the comments that accompany the photo but the photo was not moderated because Thoughtful and her daughter are named authors. The photo was posted on a public website. No one was outed. It would be no different than posting a photo of Candace Dempsey, Douglas Preston, or even me for that matter. The photo below is posted on Amazon. It's in the public already. I would have no complaint if it was posted elsewhere. Of course copyright is an entirely different issue. We will be happy to remove any images on the board if claims of copyright are made.


Image


Thanks Bruce

I have no idea why anyone would be complaining. It's a public photo of the co-author of the book we're discussing who also advertises herself as an expert for tv discussing crime, law & order. No one here would make a nasty comment about her based on her photograph. My comment was about how absurd it is for a 25 year old with no qualifications, experience or background in law enforcement or the criminal justice system to consider herself an expert for tv compared to real experts like Steve Moore & John Douglas.

Leila, it's great to see you posting here. If you get a chance to answer these questions, that would be great :) What do you think about the PMF gallery where thousands of photographs have been collected and stored of Amanda, Raffaele, their familes, friends and supporters? There's dozens of photos of Madison Paxton..... what does PMF need them for? What do you think about all the private photographs that have been stolen from peoples facebook accounts and posted on the forum and then mocked and ridiculed by your friends? What do you think about brmull posting child photos of Raffaele? What purpose does that serve? How come you don't speak out against it? You'd be horrified if people were saying those things about you and your family wouldn't you?
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby B_Real » Thu Mar 14, 2013 2:39 am

Thoughtful wrote: It seems to have escaped everyone’s notice that the purpose of the chapter was not to assign guilt or innocence, and that we did not do so. The actual focus of the chapter is Hellmann’s wrong decision about the second test, through which crucial information might possibly have been obtained. It belongs in the context of an entire book with ten chapters all devoted to various kinds of mathematical problems in trials.


Can you explain how your sentence in the book where you pruriently speculate that Amanda enjoyed telling the court about her sex life, fits into that concept?

Can you provide a source for your claim that Curt and Edda stayed in high end hotels and travelled in limousines while in Perugia? I have never heard that anywhere. And again, can you explan what it has to do with mathematical problems?

Can you explain why you include the alleged quote from Amanda 'Shit happens' which has never been reported anywhere? Why did you do that?

Why do you talk so much about the Knox family? Why are they even in the book at all? What do they have to do with maths in trials?
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby MichaelB » Thu Mar 14, 2013 2:53 am

B_Real wrote:
Thoughtful wrote: It seems to have escaped everyone’s notice that the purpose of the chapter was not to assign guilt or innocence, and that we did not do so. The actual focus of the chapter is Hellmann’s wrong decision about the second test, through which crucial information might possibly have been obtained. It belongs in the context of an entire book with ten chapters all devoted to various kinds of mathematical problems in trials.


Can you explain how your sentence in the book where you pruriently speculate that Amanda enjoyed telling the court about her sex life, fits into that concept?

Can you provide a source for your claim that Curt and Edda stayed in high end hotels and travelled in limousines while in Perugia? I have never heard that anywhere. And again, can you explan what it has to do with mathematical problems?

Can you explain why you include the alleged quote from Amanda 'Shit happens' which has never been reported anywhere? Why did you do that?

Why do you talk so much about the Knox family? Why are they even in the book at all? What do they have to do with maths in trials?


Good questions.
The stupid things Ergon says - THE BEST OF NASEER AHMAD: "Curatolo's testimony is one of the bedrock foundations of my beliefs in this case."
User avatar
MichaelB
 
Posts: 6172
Joined: Thu Jan 05, 2012 10:07 pm
Location: Perryville Prison

Re: 'Math on Trial'

Postby B_Real » Thu Mar 14, 2013 3:22 am

roteoctober wrote:
B_Real wrote:
Leila wrote:However, Conti and Vecchiotti did find a tiny quantity on the knife that Novelli claimed was testable. What other negative results could there be? Someone else's DNA?


They found 'something'. The 'something' tested negative for human DNA, blood, you name it. That's my understanding.


Of course it could be someone else's DNA, at the very least Amanda's and/or Raffaele's, or the maid's.

By the way, reading the tables from the CV Report, I understand that trace I was divided in three wells and one of them tested positive for Y-chromosome, so it is a mixture.


Let's think of it this way.

The knife has now been kicking about for years at the crime lab, being handled by various people and going on trips to court and back.

It seems almost certain that somewhere along the way it has picked up tiny traces of DNA, which would not be a problem normally because they would show up as noise.

Except they are a fatal problem when the trace Leila thinks should be tested again is also at the 'noise' level, (if it was ever there to begin with of course).
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Re: 'Math on Trial'

Postby Leila » Thu Mar 14, 2013 3:38 am

More responses.

Michael B. wrote: “What do you think this means?” [quote from Giordano follows about avoiding naming Amanda and Raffaele specifically as Rudy’s accomplices]

It means that it was not the role of the court to make any assumptions about the identity of the people they believed were with Rudy when he committed the crime. The court contented itself with ruling that he did not commit the crime alone.

Michael B. also wrote: “What do you think Massei means writing this?”

The passage from Massei, taken in context of the whole report, has the following meaning precisely: if you look at Meredith’s wounds without any reference to the way in which the crime was committed, evidence of a struggle etc., let alone evidence of who was in the cottage and so forth – if you purely look at the wounds and bruises themselves, then they could have been caused by a single person. Indeed, if Meredith had been utterly passive during the attack, unconscious for example, the attacker could have grabbed her by the arms, by the throat, stabbed her with a small knife, thrown it onto the bed, grabbed a bigger knife, moved around her and stabbed her throat from the other side, partially removed her clothing, and so on and so forth. There are no “scientific elements arising directly from forensic science” (here meaning “forensic pathology”) that rule out the injuries having been caused by a single attacker. But, as Massei explains in detail, a great deal of evidence not in the domain of the forensic pathologist shows that events did not unroll this way.

Katody wrote: “Even if the new test come out as positive it wouldn't remove the unreliability of the evidence. The knife was open to contamination during collection, handling at the police station and in the lab, as the court experts demonstrated.”

The day someone explains to me how Meredith’s DNA could have landed on the knife during collection in a place where she never set foot will be the day I see something new. Handling at the police station: the knife was not unbagged there. In the lab: with hundreds of samples tested, the probability that Meredith’s of all people would have landed on a knife being tested in the context of her murder – after 6 days in which no samples connected to her had been tested, but many from other cases – is vanishingly small.

Bruce Fischer wrote: “You are describing details about Meredith's wounds that simply do not exist."

This was in response to what I wrote about the knife sawing back and forth. That information is contained in the autopsy report. Within the large wound were several smaller cuts, and at least some of the experts explained them in this manner.

Bruce Fischer wrote: “Really? Amanda and Raffaele are free. You are not claiming in your book that the appeal was flawed due to problems with mathematics?"

You believe that Amanda and Raffaele are innocent. It seems to me that you, too should be concerned about flaws in the appeal motivations reasoning that allow people who suspect that they are guilty to remain unconvinced. In other words, even if they are free, it is not pleasant for them that a sizeable number of people will always believe that they have not been fully proved innocent, because the appeal judgment was flawed. The correctness of a judge’s explanations of his decisions should be of interest to people on all sides. And, believe it or not, the purpose of our book is much more to make this point than to try to prove that Amanda and Raffaele are guilty. Had I written a text with that specific goal, it would have been focused quite differently, and it would have been a lot longer than the single chapter.

Katody wrote: “That's the "apprentice journalist" who procured the unreliable witnesses; the miracle memory man, the miracle ear lady. Can't recall now but wasn't the heroin addict also result of his work?”

Correct. But your denigrating language is disagreeable and I have no wish to respond to comments couched in this tone. There are plenty of courteous people on this thread.

Katody also wrote: “Apart from the doubtful DNA result there were no elements indicating the kitchen knife was a murder weapon. In fact many elements indicated it wasn't, confirming the contamination hypothesis. There was no crucial information to be obtained and Hellmann was correct in his decision.”

Fine. We will agree to disagree, and conversation between us now has no reason to continue.

Dougm wrote: “So now the SC is supposed to allow a "third" test, outside accepted standards, because Novelli says so?”

The experts did not conduct a second test, using up-to-date machines that can deal with quantities as small as a single cell. They could have, but they did not. A second test does not mean reviewing the results of the first test. It means reswabbing the knife (which was done) and running the resulting sample through a machine (which was not done). I do not expect the Court of Cassation to allow anything of the kind. This is a question that will only arise in the hypothetical situation where the Court of Cassation orders a complete new appeal trial.

Erasmus44 wrote: “DNA science involves more than just pressing the button on a machine and calling a mathematician into the room to interpret the results - I am not a DNA expert myself so I tend to view a disagreement among the experts as suggesting doubt in the matter but I am fairly certain that there are different ways to conduct the tests and to interpret the results.”

I did not write “pressing a button”, I wrote that “you need to know what you are doing in a forensic lab”. Misrepresenting what I said is bad faith argument. I am here on this particular thread in order to argue in good faith, and will not respond to any comments that twist mine. If you ask a DNA analyst, they will be the first to tell you that the interpretation portion of their work is the harder part, and that the reasoning involved is essentially statistical; applied statisticians also work using a mixture of rigorous methods and intuition born of long experience.

Erasmus 44 also wrote: “Although Leila posts that she isn't opining on guilt or innocence, in fact she posts earlier that the case would be "closed" if the test she proposed showed MK's DNA.”

I say this speaking as a potential jury member, and I strongly believe that a jury would see it the same way. However, note that I have not said, and am not certain, that a second test really would reveal Meredith’s DNA. I think it was very important to determine whether it would or not, and not doing so was a missed opportunity.

Bruce Fischer wrote: “I would never question your intelligence as I have no doubt you are most likely a brilliant mathematician. Your book may very well be a vital learning tool that has been marred by including this case. It's possible for gifted minds to lack common sense. A good example would be knowing whether it is right or wrong to enter a private residence without consent to take photographs.”

I actually pride myself on my common sense, which leads me in particular to judge whether an act is “right on wrong” by the essential, major criterion of whether it causes any harm. I cannot see any other criterion that comes close in importance to this one. Harmless transgression of formalities and rules does not disturb me.

MichaelB wrote about a boy who climbed a tree. He is seemingly unaware of the difference between a tree, which has branches and a rough trunk that one can cling to, and a sheer wall. It’s easy to climb a tree; I was a champion at this in my youth, and took terrifying risks in the thin branches at the top. But is surprising to me that anyone can believe that a person standing on a metal grate with hands above his head on a window sill can lift himself in and through a jagged hole without disturbing any of the multiple shards of glass on the windowsill or sending even a fragment of glass down to the ground below. Nor is it possible for a normal-sized man to stand on the grate and reach up far enough inside the window to open the catch and swing it inwards.

Personally, I don’t understand why even those who believe in the Lone Wolf scenario don’t think that Rudy was simply let in the door by Meredith on some pretext.

MichaelB also commented on the photo situation. My response is that it does not disturb me at all that photos knowingly made public by people are reused on other sites in the course legitimate discussions, as long as they are correctly labeled and not used for stalking or other invidious purposes. I find it laughable to talk about “stealing” facebook photos. I might as well say that you are “stealing” Math on Trial if you buy a copy. Both the photographs and the book are knowingly made available to strangers.

I would be horrified if people were bandying about pictures that I myself had made public, showing us brandishing meat cleavers or exploding with laughter behind machine guns. But no such pictures of us are available because we don’t actually do those things. In what concerns childhood photos of Raffaele, if they are out there in the public domain and Brmull has shown me nothing which a couple of clicks wouldn’t have let me see for myself, I really can’t see the harm in it. I have seen several childhood pictures of Amanda put out by her own family, and she was a lovely little girl. What about it?

As for PMF collecting every kind of document about Meredith’s murder in one place, I think it is a valuable resource.
Leila
 
Posts: 659
Joined: Sun Mar 10, 2013 3:48 pm

Re: 'Math on Trial'

Postby B_Real » Thu Mar 14, 2013 3:41 am

Leila wrote:MichaelB also commented on the photo situation. My response is that it does not disturb me at all that photos knowingly made public by people are reused on other sites in the course legitimate discussions, as long as they are correctly labeled and not used for stalking or other invidious purposes.


:lol: Have you actually seen the photo gallery at PMF? My hair stood on end the first time I checked it out. Same with everybody else who had a look.

It is the world's largest most comprehensive collection of Knox photos, not to mention all of her friends, family and supporters. It's obsessive, creepy, and serves absolutely no purpose.

There's no gallery here with a vast collection of e.g. Mignini photos. How would it help us discuss the case if we did have one?
User avatar
B_Real
 
Posts: 3756
Joined: Thu Feb 09, 2012 11:12 am

Next

Return to Injustice in Perugia Forum

Who is online

Users browsing this forum: No registered users and 9 guests

cron