Continuation Part 11: Amanda Knox/Raffaele Sollecito

[christianahannah]

Diocletus,

Nice find. I liked this part: "The luminol tests were not performed according to accepted standards...Most importantly, confirmatory tests were not performed to determine whether the "glow" was caused by human blood or something else."

McDaniel had murdered Giddings and placed her body in the bathtub. He came back later and dismembered parts of her body.

What do you think the glow may have been from?

 

[Machiavelli]

So, putting the finding that there is no evidence of Ms Knox in the murder room together with the inferences you have drawn from the evidentiary picture elsewhere, what best prevails? How do the inferences drawn survive the absence of evidence in the room?

Whichever way you look at this, you simply must prove she was in the room. All other evidence for guilt that you cite is predicated on this most important contention.

Apart from the fact that is a lot of evidence in the room: there's the whole autopsy report, a lamp of Knox, and DNA of Sollecito; the simple point is that your argument is completely irrational. The alleged "no evidence" in the room is nothilg like evidence of innocence, it's not like something that can counter-balance something else. It is simply false that you must prove she was in the murder room. It is false both legally and by common sense. And the room is not "more important" to the crime than other places which are equally necessarily involved in the crime, in which the murderer necessarily performed activities. So that: if "absence of physical evidence" is evidence of absence, something that must be won and overcome, than a reverse equivalent argument could be brought in favor of Guede: how do you explain the absence of Guede's DNA or any other proof of him in the small bathroom, where the murderer must have switched on the light, washed himself/herself, his hands and feet, touched the taps and walked barefoot? And also: if physical evidence on the crime scene is an important indicator, how is it that you don't consider the room where there was the breaking a window as part of the crime scene? Why do you ignore the presence of luminol stains 176 and 177 with mixed DNA traces from the victim and with Knox, why don't you "count" them as physical evidence findings?

 

[LondonJohn]

Again with this nonsense. The curve is not Gaussian, it cannot be Gaussian. It must be truncated. The probability quickly drops to zero.

Well, Machiavelli's point has some (very limited, and misunderstood) validity.

The experimental curve for t(lag) is demonstrably a modified bell curve, with a significantly squashed left hand side (owing to the fact that the curve cannot extend beyond t=0) and a shortened tail on the right hand side (owing to the fact that there is a definitive upper limit to t(lag)).

And Machiavelli is correct to say that we are dealing with the far right hand side of this bell curve, where there is (by definition) a shallower gradient, and where therefore there is a lesser step change in probability per unit of time than for the steep middle part of the curve.

But....... the curve at this point still enables cogent, statistically-valid analysis. I can go over the precise data at some point if necessary, but from memory, the curve obeys the following approximate properties:

At t=150, the area under the curve to the right of this point equates to about 2%, meaning that there's a 2% probability that t(lag) is >= 150 minutes.

At t=180, the area under the curve to the right equates to about 0.2%, meaning that a) there's a 2% probability that t(lag) is >=180m, and b) there's a 1.8% (2%-0.2%) probability that t(lag) is between 150m and 180m.

At t=210, the area under the curve to the right equates to about 0.01%, meaning that a) there's a 0.01% probability that t(lag) is >=210m, b) there's a 1.99% (2%-0.01%) probability that t(lag) is between 150m and 210m.

At t=240, the curve has reached zero on the y-axis, showing that there's zero probability of t(lag) >=240m.

Now, with all those in mind, the pertinent questions are actually conditional probability calculations. These are as follows:


Q1: Given that t >=150*, what is the probability that t(lag) is between 150-180m as opposed to >180m?

A1: 90%. In other words, it's 90% likely that Kercher died between 9pm and 9.30pm as opposed to later than 9.30pm.


Q2: Given that t >=150, what is the probability that t(lag) is between 150-210m as opposed to >210m?

A2: 99.5%. In other words, it's 99.5% likely that Kercher died between 9pm and 10pm as opposed to later than 10pm.


Q3: Given that t >=150, what is the probability that t(lag) is between 150-240m as opposed to >240m?

A3: 100%. It's a statistical certainty that Kercher died between 9pm and 10.30pm as opposed to later than 10.30pm.


That's what an analysis of the statistics, and the proper application of conditional probability theory, tell us. I will concede that there's bound to be some margin of error around these numbers, but the overall message will remain absolutely firm: it's very highly likely that Kercher died between 9pm and 9.30pm; it's a virtual certainty that she died between 9pm and 10pm, and its a total certainty that she died before 10.30pm.


* Working on the knowledge that start of Kercher's last meal (which is the relevant starting point here) took place at around 6.30pm, and that she was last knowingly alive just before 9pm (150 minutes later)

 

[Machiavelli]

I am curious, do you believe that American police or prosecutors ever manufacture evidence against defendants?

What kind of question is that? I make no sense of it. I don't "believe" anything. If there is evidence some of them do so, and I can see it, then I will note it on the specific case.

 

[LondonJohn]

What zero? There are even records of 12-hour mistakes based on the calculation of digestion time.
Sophie and friends locate the dinner at 18:00 - 18:30, which - according to defence consultant Introna - would mean that the alleged "most likely" location (21:00 - 21:10) is actually 160 minutes far from dinner as for our testimony evidence in the best scenario. Even if we strain the information and we try to make violence to the testimony, to disregard them, and we decide to locate the time of dinner where it was not, like at 19:30, still the alleged "most likely" time of s.c. "attack" will be located about 100 minutes far away from the time of dinner.
This means it is absolutely far beyond the area where the probability drops quickly. There is nothing like a quick drop in that segment.

No no no no. The critical factor here is that the entirety of Kercher's last meal was still entirely within her stomach when she died.

All of the timing variance problems related to digestive evidence occur either when all of the last meal is already in transit through the intestines, or where part of the last meal has left the stomach while some remains. In such instances, it's very well known (medically) that a) intestinal transit rates differ markedly between people, and b) total stomach emptying times differ markedly too.

But we are dealing with neither of these things here. We are dealing with a victim whose last meal was still entirely within her stomach. And there's good (and sufficient) empirical evidence available about how long human stomachs take after the ingestion of food to pass the first contents out to the duodenum. It's very well known that this time is wholly dependent upon when the meal started (and is independent of the length of time of consumption, or whether or not more food is ingested at some later time), and that it follows a well-defined modified-normal curve.

 

[MDDVS]

What zero? There are even records of 12-hour mistakes based on the calculation of digestion time.
Sophie and friends locate the dinner at 18:00 - 18:30, which - according to defence consultant Introna - would mean that the alleged "most likely" location (21:00 - 21:10) is actually 160 minutes far from dinner as for our testimony evidence in the best scenario. Even if we strain the information and we try to make violence to the testimony, to disregard them, and we decide to locate the time of dinner where it was not, like at 19:30, still the alleged "most likely" time of s.c. "attack" will be located about 100 minutes far away from the time of dinner.
This means it is absolutely far beyond the area where the probability drops quickly. There is nothing like a quick drop in that segment.

We are not talking about "diggestion time" here. We are talking about the time it takes for the stomach to start to empty. And yes, Zero because the stomach must start to empty eventualy.

From the paper by Hellmig et al. (2006) we learn that 160 min is very close to the point where the probabilty reaches zero. For a person Meredith's age that's already stretching a lot as the maximum recorded is closer to 140 min.

It doesn't matter how fast the probabilty drops, it reaches zero at most at 180-190 min, if we talking about healthy adults.

 

[Bill Williams]

Apart from the fact that is a lot of evidence in the room: there's the whole autopsy report, a lamp of Knox, and DNA of Sollecito; the simple point is that your argument is completely irrational. The alleged "no evidence" in the room is nothilg like evidence of innocence, it's not like something that can counter-balance something else. It is simply false that you must prove she was in the murder room. It is false both legally and by common sense. And the room is not "more important" to the crime than other places which are equally necessarily involved in the crime, in which the murderer necessarily performed activities. So that: if "absence of physical evidence" is evidence of absence, something that must be won and overcome, than a reverse equivalent argument could be brought in favor of Guede: how do you explain the absence of Guede's DNA or any other proof of him in the small bathroom, where the murderer must have switched on the light, washed himself/herself, his hands and feet, touched the taps and walked barefoot? And also: if physical evidence on the crime scene is an important indicator, how is it that you don't consider the room where there was the breaking a window as part of the crime scene? Why do you ignore the presence of luminol stains 176 and 177 with mixed DNA traces from the victim and with Knox, why don't you "count" them as physical evidence findings?

Folk need to read this post again and again. This is the theory of the crime for most of the Pro-guilt PR lobby.

Amanda Knox is guilty even if not there.

And an absence of Rudy Guede's forensics in the bathroom, somehow makes up for his DNA inside the victim.



And these folk have the gall to defend the word "osmotic". Ooooops, as per the PMF translation...... "organic".

 

[Desert Fox]

What kind of question is that? I make no sense of it. I don't "believe" anything. If there is evidence some of them do so, and I can see it, then I will note it on the specific case.

I am taking this as an attempt to dodge

 

[kwill]

Well, Machiavelli's point has some (very limited, and misunderstood) validity.

The experimental curve for t(lag) is demonstrably a modified bell curve, with a significantly squashed left hand side (owing to the fact that the curve cannot extend beyond t=0) and a shortened tail on the right hand side (owing to the fact that there is a definitive upper limit to t(lag)).

And Machiavelli is correct to say that we are dealing with the far right hand side of this bell curve, where there is (by definition) a shallower gradient, and where therefore there is a lesser step change in probability per unit of time than for the steep middle part of the curve.

But....... the curve at this point still enables cogent, statistically-valid analysis. I can go over the precise data at some point if necessary, but from memory, the curve obeys the following approximate properties:

At t=150, the area under the curve to the right of this point equates to about 2%, meaning that there's a 2% probability that t(lag) is >= 150 minutes.

At t=180, the area under the curve to the right equates to about 0.2%, meaning that a) there's a 2% probability that t(lag) is >=180m, and b) there's a 1.8% (2%-0.2%) probability that t(lag) is between 150m and 180m.

At t=210, the area under the curve to the right equates to about 0.01%, meaning that a) there's a 0.01% probability that t(lag) is >=210m, b) there's a 1.99% (2%-0.01%) probability that t(lag) is between 150m and 210m.

At t=240, the curve has reached zero on the y-axis, showing that there's zero probability of t(lag) >=240m.

Now, with all those in mind, the pertinent questions are actually conditional probability calculations. These are as follows:


Q1: Given that t >=150*, what is the probability that t(lag) is between 150-180m as opposed to >180m?

A1: 90%. In other words, it's 90% likely that Kercher died between 9pm and 9.30pm as opposed to later than 9.30pm.


Q2: Given that t >=150, what is the probability that t(lag) is between 150-210m as opposed to >210m?

A2: 99.5%. In other words, it's 99.5% likely that Kercher died between 9pm and 10pm as opposed to later than 10pm.


Q3: Given that t >=150, what is the probability that t(lag) is between 150-240m as opposed to >240m?

A3: 100%. It's a statistical certainty that Kercher died between 9pm and 10.30pm as opposed to later than 10.30pm.


That's what an analysis of the statistics, and the proper application of conditional probability theory, tell us. I will concede that there's bound to be some margin of error around these numbers, but the overall message will remain absolutely firm: it's very highly likely that Kercher died between 9pm and 9.30pm; it's a virtual certainty that she died between 9pm and 10pm, and its a total certainty that she died before 10.30pm.


* Working on the knowledge that start of Kercher's last meal (which is the relevant starting point here) took place at around 6.30pm, and that she was last knowingly alive just before 9pm (150 minutes later)

I was a math teacher. This is lovely.

 

[Bill Williams]

Apart from the fact that is a lot of evidence in the room: there's the whole autopsy report, a lamp of Knox, and DNA of Sollecito; the simple point is that your argument is completely irrational. The alleged "no evidence" in the room is nothilg like evidence of innocence, it's not like something that can counter-balance something else. It is simply false that you must prove she was in the murder room. It is false both legally and by common sense. And the room is not "more important" to the crime than other places which are equally necessarily involved in the crime, in which the murderer necessarily performed activities. So that: if "absence of physical evidence" is evidence of absence, something that must be won and overcome, than a reverse equivalent argument could be brought in favor of Guede: how do you explain the absence of Guede's DNA or any other proof of him in the small bathroom, where the murderer must have switched on the light, washed himself/herself, his hands and feet, touched the taps and walked barefoot? And also: if physical evidence on the crime scene is an important indicator, how is it that you don't consider the room where there was the breaking a window as part of the crime scene? Why do you ignore the presence of luminol stains 176 and 177 with mixed DNA traces from the victim and with Knox, why don't you "count" them as physical evidence findings?

"there's the whole autopsy report"​

8 of 9 experts who Judge Massei mentions in his 2010 motivations read that autopsy report and said that a single attacker could not be ruled out.

"The lamp of Knox"​

Forensically useless, and probably brought into the room by a postal police or investigator to get some light on the crime scene. Besides, what was Knox supposed to be doing with the lamp?

"DNA of Sollecito"​

More properly, male-DNA from a group of males from which Raffaele cannot be ruled out.

"The alleged "no evidence" in the room is nothilg like evidence of innocence"​

Strawman argument, meant to reverse the burden of proof. The no evidence of Knox in that room rules her out forensicly.... I suppose that in some alternate universe she still could be guilty.

"It is simply false that you must prove she was in the murder room."​

- Wait a minute. Didn't you just say that the autopsy is evidence? Aren't you implying that the autopsy implies she's in the room? It seems that you're content to argue the possibility Knox might have been involved, rather than prove beyond a reasonable doubt she was involved.

"And the room is not "more important" to the crime than other places which are equally necessarily involved in the crime, in which the murderer necessarily performed activities."​

It is if there is no evidence she was in that room, and all the evidence of her outside that room is because she lived there...... and used that bathroom as a matter of course. Are you trying to argue that the bathroom was forensicly sterile before the murder occurred?

 

[Machiavelli]

Well, Machiavelli's point has some (very limited, and misunderstood) validity.

The experimental curve for t(lag) is demonstrably a modified bell curve, with a significantly squashed left hand side (owing to the fact that the curve cannot extend beyond t=0) and a shortened tail on the right hand side (owing to the fact that there is a definitive upper limit to t(lag)).

And Machiavelli is correct to say that we are dealing with the far right hand side of this bell curve, where there is (by definition) a shallower gradient, and where therefore there is a lesser step change in probability per unit of time than for the steep middle part of the curve.

But....... the curve at this point still enables cogent, statistically-valid analysis. I can go over the precise data at some point if necessary, but from memory, the curve obeys the following approximate properties:

At t=150, the area under the curve to the right of this point equates to about 2%, meaning that there's a 2% probability that t(lag) is >= 150 minutes.

At t=180, the area under the curve to the right equates to about 0.2%, meaning that a) there's a 2% probability that t(lag) is >=180m, and b) there's a 1.8% (2%-0.2%) probability that t(lag) is between 150m and 180m.

At t=210, the area under the curve to the right equates to about 0.01%, meaning that a) there's a 0.01% probability that t(lag) is >=210m, b) there's a 1.99% (2%-0.01%) probability that t(lag) is between 150m and 210m.

At t=240, the curve has reached zero on the y-axis, showing that there's zero probability of t(lag) >=240m.

Now, with all those in mind, the pertinent questions are actually conditional probability calculations. These are as follows:


Q1: Given that t >=150*, what is the probability that t(lag) is between 150-180m as opposed to >180m?

A1: 90%. In other words, it's 90% likely that Kercher died between 9pm and 9.30pm as opposed to later than 9.30pm.


Q2: Given that t >=150, what is the probability that t(lag) is between 150-210m as opposed to >210m?

A2: 99.5%. In other words, it's 99.5% likely that Kercher died between 9pm and 10pm as opposed to later than 10pm.


Q3: Given that t >=150, what is the probability that t(lag) is between 150-240m as opposed to >240m?

A3: 100%. It's a statistical certainty that Kercher died between 9pm and 10.30pm as opposed to later than 10.30pm.


That's what an analysis of the statistics, and the proper application of conditional probability theory, tell us. I will concede that there's bound to be some margin of error around these numbers, but the overall message will remain absolutely firm: it's very highly likely that Kercher died between 9pm and 9.30pm; it's a virtual certainty that she died between 9pm and 10pm, and its a total certainty that she died before 10.30pm.


* Working on the knowledge that start of Kercher's last meal (which is the relevant starting point here) took place at around 6.30pm, and that she was last knowingly alive just before 9pm (150 minutes later)

This calculation is incorrect. Let's try to see where the error is. I will not consider single data figures but I will address instead another concept, which is: the need to change the sample base on which develop error analisis, that means a change in the whole statystical distribution function whenever you change the reference sample. The error in LJs' calculation consists in keeping the same reference sample. LJ does not change the population of reference of the sample.

I already made an example that has to do with age. An average Italian person currently has less than 10% probability to survive up to 88 years.
In fact, less than 50% of individuals survive until 80 years of age. So the average Italian has less than 50% probability to survive beyond 80 years.

However, wht's the probability for an 80 year old person to reach 88 years? The probability is above 50%.

It means if you pick up a random sample of 80 year old people, after 8 years you will find that more than 50% is still alive. This is much more than what you would expect if you considered the general sample.
A 80 year old person has a much higher probability to reach 88 years of age than an average person.
Why this?

Because a sample population of 80 year old people is not the average Italian population.
If you want to calculate the actual probability, once you know a value, you need to consider an appropriate sample.
The distribution becomes more diffuse and the gradient of the curve with this new sample is less steep, because the sample has changed.

In other words, if you want to calculate a probability base on gastric emptying, then you need to re-draw the curve, that shall be a curve that describes the distribution of that new, diferent sample, a sample of cases of gastric emptying later than 100-150 minutes.

If you assume as a known value that the gastric emptying has not started after 100 minutes, you need to consider a curve based on a sample population of people whose gastric emptying didn't start after 100 minutes, not the average case population. Based on a known value, you should do a research focused on a sample of people who share that value, not the average population of cases; than draw another curve based on that sample, and after that you can calculate gradient and areas.

 

[Chris_Halkides]

A list of questions about luminol

McDaniel had murdered Giddings and placed her body in the bathtub. He came back later and dismembered parts of her body.

What do you think the glow may have been from?

If it is true (as I gathered from the links I read) that the police officer did not photograph the reaction, then there might not have even been any glow. However for the sake of argument, I will assume that there was. It may or may not have been from blood, and if it were not blood, then bleach or Comet cleanser would be two other hypotheses. Drawing the inference that it must have been blood is unwarranted in the absence of confirmatory tests. IMO it speaks well of a system of justice that excludes prejudicial evidence, and McDaniel was convicted on the basis of evidence that I have no reason to believe was unsound in any way.

Here is a list of questions about the luminol evidence within this case that I generated on the fly a few days ago.

Why is Meredith's DNA absent from the majority of the luminol-positive areas?
Why was no confirmatory test done?
Why did the luminol-positive areas test negative by TMB?
Why do some jurisdictions (several states in the U.S. and at least one in Australia) not even allow luminol evidence into the courtroom without supporting information, while others allow it in but acknowledge that luminol is only presumptive? Is there any jurisdiction that treats luminol as a confirmatory test?
What evidence rules out iron, manganese, or other transition metal ions as the source of the luminol reactivity?
What errors did Virkler and Lednev make in their conclusion that the areas were probably not blood?
Who made Rep. 180? If it is Amanda's track, then why doesn't the second toe match? If it is not Amanda's track, then what reason is there to suppose that the substance in question is blood?
Why are there only right feet? When and how were the tracks made?

Based partially upon my research today, I would like to add two more:

Did the overapplication of luminol cause a false positive?
Should the overapplication of luminol reduce one's confidence in the people doing and analyzing this test?

 

[kwill]

Apart from the fact that is a lot of evidence in the room: there's the whole autopsy report, a lamp of Knox, and DNA of Sollecito . . .

Oh no. Just when we were getting ourselves sorted out about how evidence is a concept and not an object in a room, here comes a parade of objects in a room claiming to be evidence.

Tho' I'm fairly certain that the autopsy report itself was never in the room, and also that it prompted several pathologists to testify that the manner of death did not require more than one attacker.

However, that may be a concept, which is allowed to be called evidence, but only as long as it's not also an object in the room. If so, it's evidence that Guede was working all alone that night.

 

[Bill Williams]

Oh no. Just when we were getting ourselves sorted out about how evidence is a concept and not an object in a room, here comes a parade of objects in a room claiming to be evidence.

Tho' I'm fairly certain that the autopsy report itself was never in the room, and also that it prompted several pathologists to testify that the manner of death did not require more than one attacker.

However, that may be a concept, which is allowed to be called evidence, but only as long as it's not also an object in the room. If so, it's evidence that Guede was working all alone that night.

There are sometimes when I think Machivaelli is a pro-innocence false-flag poster.

Bruce Fischer.... is it you!?

 

[Kaosium]

So they are all part of a conspiracy, aren't they? Indeed it makes a lot of sense. The Supreme Court annulls Hellmann (who, in your view, would have worked properly and issued the "right" verdict) because they are worried about the "reputation of the judiciary". If they were worried about the "reputation of the judiciary" and they thought Hellmann was sound and good, don't you think it would have been very easy to let it stand?
Is that your theory?

But very embarrassing to the host of judges like Matteini, who imprisoned Raffaele and Amanda on false evidence. Or the idiot who confirmed that who said something moronic and easily dis-proven like 'that window would require Spiderman to climb into' as well as Massei who allowed all that garbage evidence in his court and who failed to punish the perjuries and failures to disclose on the part of the prosecution and who took Curatolo seriously instead of laughing him out of court and having Mignini frog-marched to the brig for ever presenting him in court for the third time as a material witness in a murder case. Plus there's the fact the Squadra Mobile had all received awards for their 'work' on the case and the clueless clown Giobbi who'd put Amanda's picture on the wall in Rome as one of the SCO's greatest 'achievements' and the Police Unions who were out there 'protesting' the acquittal, the Polizia Scientifica who'd used bogus DNA work in court and actually pretended they could get precise forensic evidence from the crime scene they'd already trashed.

That's a lot of people in Italy who need for Raffaele and Amanda to be considered guilty or their own guilt in these matters might just be investigated, or at the very least they must accept their profound humiliation in being party to this travesty. One thing the Court of Cassation of Italy has shown is that they simply do not care what the ECHR has to say about their many and notorious violations and the Republic of Italy shells out hundreds of millions of euros because of it and it's been going on for decades and has never gotten fixed.

 

[MDDVS]

Well, Machiavelli's point has some (very limited, and misunderstood) validity.

The experimental curve for t(lag) is demonstrably a modified bell curve, with a significantly squashed left hand side (owing to the fact that the curve cannot extend beyond t=0) and a shortened tail on the right hand side (owing to the fact that there is a definitive upper limit to t(lag)).

And Machiavelli is correct to say that we are dealing with the far right hand side of this bell curve, where there is (by definition) a shallower gradient, and where therefore there is a lesser step change in probability per unit of time than for the steep middle part of the curve.

But....... the curve at this point still enables cogent, statistically-valid analysis. I can go over the precise data at some point if necessary, but from memory, the curve obeys the following approximate properties:

At t=150, the area under the curve to the right of this point equates to about 2%, meaning that there's a 2% probability that t(lag) is >= 150 minutes.

At t=180, the area under the curve to the right equates to about 0.2%, meaning that a) there's a 2% probability that t(lag) is >=180m, and b) there's a 1.8% (2%-0.2%) probability that t(lag) is between 150m and 180m.

At t=210, the area under the curve to the right equates to about 0.01%, meaning that a) there's a 0.01% probability that t(lag) is >=210m, b) there's a 1.99% (2%-0.01%) probability that t(lag) is between 150m and 210m.

At t=240, the curve has reached zero on the y-axis, showing that there's zero probability of t(lag) >=240m.

Now, with all those in mind, the pertinent questions are actually conditional probability calculations. These are as follows:


Q1: Given that t >=150*, what is the probability that t(lag) is between 150-180m as opposed to >180m?

A1: 90%. In other words, it's 90% likely that Kercher died between 9pm and 9.30pm as opposed to later than 9.30pm.


Q2: Given that t >=150, what is the probability that t(lag) is between 150-210m as opposed to >210m?

A2: 99.5%. In other words, it's 99.5% likely that Kercher died between 9pm and 10pm as opposed to later than 10pm.


Q3: Given that t >=150, what is the probability that t(lag) is between 150-240m as opposed to >240m?

A3: 100%. It's a statistical certainty that Kercher died between 9pm and 10.30pm as opposed to later than 10.30pm.


That's what an analysis of the statistics, and the proper application of conditional probability theory, tell us. I will concede that there's bound to be some margin of error around these numbers, but the overall message will remain absolutely firm: it's very highly likely that Kercher died between 9pm and 9.30pm; it's a virtual certainty that she died between 9pm and 10pm, and its a total certainty that she died before 10.30pm.


* Working on the knowledge that start of Kercher's last meal (which is the relevant starting point here) took place at around 6.30pm, and that she was last knowingly alive just before 9pm (150 minutes later)

No, the curve is not Gaussian, as you can find in most of the relevant literature. We also know that is must be truncated because the stomach must start to empty eventualy.

The curve actualy has a tail that is heavier than a gaussian, that's not the issue. The issue is that it is truncated so it is impossible to go beyond a certain threshold.

 

[Bill Williams]

from Italian journalist Francesco Luna at his Facebook page (original in Italian, this is a translation)

AMANDA AND RAFFAELE, SEVEN YEARS OF INJUSTICE

The 2nd November 2007 was a clear bright day like today. Amanda Knox was twenty years old and was happy, while she crossed Piazza Grimana square and arrived at via della Pergola, in Perugia. She had left her new boyfriend, Raffaele Sollecito, sleeping in bed in the small apartment on Corso Garibaldi and was returning to the little isolated villa that she shared with her friends Meredith, Filomena and Laura. The young American girl from Seattle, in Italy for just a little more than one month, didn’t know that from that moment she would soon enter into a dark tunnel of horror, made up of blood, absurd accusations, bogus evidence, violence, psychological pressure, deception, mass media lynching, years of imprisonment, conviction, acquittal and then conviction again. She didn’t imagine that the poor, oblivious Raffaele was about to end up with her in that deadly grip.

The recounting by the mass media of the court proceedings for the murder of poor Meredith Kercher and the convictions of Amanda Knox and Raffaele Sollecito is an embarrassment for Italian justice and is a black page for the Italian and English newspapers. Rarely in a civilized country has one seen so many errors, so much bullying, abuse, so many lies communicated to the media and published without a minimum of critical corroboration. Two normal kids were depicted as despicable people, condemned in the media, trapped by inexistent evidence and thrown in prison for four years like murderers. Only a courageous verdict delivered in 2011 by Judge Hellmann allowed the two unfortunates to leave prison, but their ordeal was not over. With a new, paradoxical decision, in which they unduly entered into the merits of the case, the Supreme Court annulled the acquittal in 2013. And indeed it ordered a new conviction, challenging all logic, which was promptly provided by the Assize Court of Appeal of Florence in January of this year.

In the meantime, the real, sole murderer of Meredith, Rudy Guede, saw his conviction reduced from thirty to only sixteen years of imprisonment, thanks to his accusations against Amanda Knox, the real victim of this abhorrent witch hunt. And thanks to his opportunistic legal choices, he could leave prison already this year.

Italy has one last chance to close with dignity this small despicable story. The Supreme Court will decide next March 25th. The outlook is not at all good, it’s useless to hide it. It’s clear that the truth about Meredith Kercher’s death is of very little importance. Everyone knows that it was only Rudy Guede that murdered her. All that counts is the reputation of the lousy investigators and many journalists, who abandoned their duty towards critique and truth to obtain a few miserable and slanderous scoops. But above all else, at this point there is a whole system at stake, that has chosen to protect itself and the careers of police, judges and journalists. If they continue to ignore the awful error made, convicting Amanda Knox and Raffaele Sollecito for a murder which they are not responsible for, the system will perhaps save itself, but the life of two innocents will be sacrificed. If they will have the courage to admit the error, perhaps it will be necessary to apologize, but we will have shown, despite everything, to still be a civilized country.

 

[MDDVS]

This calculation is incorrect. Let's try to see where the error is. I will not consider single data figures but I will address instead another concept, which is: the need to change the sample base on which develop error analisis, that means a change in the whole statystical distribution function whenever you change the reference sample. The error in LJs' calculation consists in keeping the same reference sample. LJ does not change the population of reference of the sample.

I already made an example that has to do with age. An average Italian person currently has less than 10% probability to survive up to 88 years.
In fact, less than 50% of individuals survive until 80 years of age. So the average Italian has less than 50% probability to survive beyond 80 years.

However, wht's the probability for an 80 year old person to reach 88 years? The probability is above 50%.
It means if you pick up a random sample of 80 year old people, after 8 years you will find that more than 50% is still alive. This is much more than what you would expect if you considered the general sample.
A 80 year old person has a much higher probability to reach 88 years of age than an average person.
Why this?

Because a sample population of 80 year old people is not the average Italian population.
If you want to calculate the actual probability, once you know a value, you need to consider an appropriate sample.
The distribution becomes more diffuse and the gradient of the curve with this new sample is less steep, because the sample has changed.

In other words, if you want to calculate a probability base on gastric emptying, then you need to re-draw the curve, that shall be a curve that describes the distribution of that new, diferent sample, a sample of cases of gastric emptying later than 100-150 minutes.

If you assume as a known value that the gastric emptying has not started after 100 minutes, you need to consider a curve based on a sample population of people whose gastric emptying didn't start after 100 minutes, not the average case population. Based on a known value, you should do a research focused on a sample of people who share that value, not the average population of cases; than draw another curve based on that sample, and after that you can calculate gradient and areas.

Yes, what's the proability of an 80 year old to reach 150 years of age?

 

[Bill Williams]

But very embarrassing to the host of judges like Matteini, who imprisoned Raffaele and Amanda on false evidence. Or the idiot who confirmed that who said something moronic and easily dis-proven like 'that window would require Spiderman to climb into' as well as Massei who allowed all that garbage evidence in his court and who failed to punish the perjuries and failures to disclose on the part of the prosecution and who took Curatolo seriously instead of laughing him out of court and having Mignini frog-marched to the brig for ever presenting him in court for the third time as a material witness in a murder case. Plus there's the fact the Squadra Mobile had all received awards for their 'work' on the case and the clueless clown Giobbi who'd put Amanda's picture on the wall in Rome as one of the SCO's greatest 'achievements' and the Police Unions who were out there 'protesting' the acquittal, the Polizia Scientifica who'd used bogus DNA work in court and actually pretended they could get precise forensic evidence from the crime scene they'd already trashed.

That's a lot of people in Italy who need for Raffaele and Amanda to be considered guilty or their own guilt in these matters might just be investigated, or at the very least they must accept their profound humiliation in being party to the travesty. One thing the Court of Cassation of Italy has shown is that they simply do not care what the ECHR has to say about their many and notorious violations and the Republic of Italy shells out hundreds of millions of euros because of it and it's been going on for decades and has never gotten fixed.

See: http://www.youtube.com/watch?v=8JL6nIkaYLs

 

[Kauffer]

Apart from the fact that is a lot of evidence in the room: there's the whole autopsy report, a lamp of Knox, and DNA of Sollecito; the simple point is that your argument is completely irrational. The alleged "no evidence" in the room is nothilg like evidence of innocence, it's not like something that can counter-balance something else. It is simply false that you must prove she was in the murder room. It is false both legally and by common sense. And the room is not "more important" to the crime than other places which are equally necessarily involved in the crime, in which the murderer necessarily performed activities. So that: if "absence of physical evidence" is evidence of absence, something that must be won and overcome, than a reverse equivalent argument could be brought in favor of Guede: how do you explain the absence of Guede's DNA or any other proof of him in the small bathroom, where the murderer must have switched on the light, washed himself/herself, his hands and feet, touched the taps and walked barefoot? And also: if physical evidence on the crime scene is an important indicator, how is it that you don't consider the room where there was the breaking a window as part of the crime scene? Why do you ignore the presence of luminol stains 176 and 177 with mixed DNA traces from the victim and with Knox, why don't you "count" them as physical evidence findings?

No, no, no. Everything depends on her being in the room. If she wasn't in the room, then the luminol hits have no probative value; if she wasn't in the room then the 'fake' burglary didn't happen - there would be no need for it. It would be logically incompatible and completely unnecessary. There is nothing in the autopsy report findings confirming the presence of Ms Knox in the room; the presence of the lamp is not proof of how it got there. It could have got there in a number of ways. The PG argument hinted at by Nencini is that it was used as the detection tool for evidence that was then cleaned; yet there is no evidence of cleaning. Why should I not expect to see at least a similar quantity and quality of evidence for Ms Knox and Mr Sollecito's presence in the room as for Guede? Where are the prints in blood in that room? It's a small space; there were, according to you, four people in the space, yet no evidence of Ms Knox and a few cells of Mr Sollecito? How is this possible?

Please, make your case. Explain the mechanics of the murder in that room in such a way that it is clear why there are only a few cells of evidence for the presence of two people acting in concert? And no evidence at all for one of them?