I’m pretty familiar with this case having read a couple of books about it and followed both of the websites listed and a few others.
I would say there is 0 chance of Knox or Sollecito being involved in the murder at all. They have no motive. The prosecution completely failed to show that they had ever participated in “sex games” together or with different partners in the past. There is no evidence of either of them at the crime scene at the time of the crime. None of their past behaviors make them likely candidates to commit this type of crime. Even the circumstantial evidence Judge Massei used to convict them is stretched and convoluted: that Knox bathed twice in 24 hours, that they turned their cell phone off together apparently before going to bed and on again together when they got up briefly early in the morning, that Sollecito didn’t use his computer for a few hours. There are many many reasons besides taking part in a homicide by which these behaviors might occur. I doubt that these two people are in any way involved with the crime.
Rudy Guede I’d say 60% he’s probably guilty of the murder. All the physical evidence, bloody hand print, bloody foot prints, mitochondrial DNA in the victim, in the victim’s purse, belong to Rudy Guede. He was certainly at the crime scene. He certainly had some kind of interaction with Meredith Kercher that night. He has exactly the type of criminal profile that would lead him into this sort of circumstance. He is probably the murderer.
The thing that gives me pause is Guede’s persistent denial that he murdered Kercher. This guy is a smash and grab burglar. He does not think ahead. He does not plan his escapes. He does not enter the crime scene thoughtfully with the idea of maximum profit. All his crimes are crimes of desperation, crudely executed. I doubt he meant to murder Meredith Kercher. He was there to rob. Rape presented itself to him as an opportunity. He attempted it. She resisted. He murdered her accidentally. With criminals of this type you would expect the perpetrator to be remorseful and confess. They nearly always do. Rudy does not. That bothers me.
If we believe Rudy, however, we enter the next realm of probability. Could Meredith Kercher be unlucky enough to be burglarized by Guede and then have a couple of other bandit murderers stop by , too? Doesn’t seem likely to me. Was she MI5 or dealing drugs, herself? Did anybody check? I’m not happy with it.
I suspect you’ll agree with me within 20%. There are no scientists or law enforcement officials outside of Italy that give any credence to the prosecution’s analysis of this crime. As of yesterday, even the Italian government is backing away from these police and prosecutors. Knox and Sollecito will be freed on the review of the DNA analysis. Knox will be given time served on her so-called “slander” charge and they’ll send her home to Washington state before her 24th birthday in July.
It might help to read the linked essay in my previous comment. When one assigns probability 0 or 1 to something, one cannot update based on evidence. That is, if one really believes that, there should be no amount of evidence that will convince you. That’s problematic if one is trying to be a rationalist. And if there is some amount of evidence that can convince you then you don’t really assign probability 0 or 1 to the claim.
One either GOES to the restroom or one DOESN’T go to the restroom. It’s not a 1% possibility thing. Something is or it isn’t. Not a good idea? Based on the fact that once in the courtroom it’s nothing but a big pokergame. But there is 0 evidence.
Probabilities are measures of how certain one is about something. Thus, for example if one is going to estimate how likely a coin is to come heads one will say .5. This doesn’t change the fact that the coin either will come up heads or come up tails. But I don’t know which one it will be. Similarly, if something in the past may or may not have occurred, then you can assign numeric estimates to how likely it is to have occurred.
I’m pretty familiar with this case having read a couple of books about it and followed both of the websites listed and a few others.
I would say there is 0 chance of Knox or Sollecito being involved in the murder at all. They have no motive. The prosecution completely failed to show that they had ever participated in “sex games” together or with different partners in the past. There is no evidence of either of them at the crime scene at the time of the crime. None of their past behaviors make them likely candidates to commit this type of crime. Even the circumstantial evidence Judge Massei used to convict them is stretched and convoluted: that Knox bathed twice in 24 hours, that they turned their cell phone off together apparently before going to bed and on again together when they got up briefly early in the morning, that Sollecito didn’t use his computer for a few hours. There are many many reasons besides taking part in a homicide by which these behaviors might occur. I doubt that these two people are in any way involved with the crime.
Rudy Guede I’d say 60% he’s probably guilty of the murder. All the physical evidence, bloody hand print, bloody foot prints, mitochondrial DNA in the victim, in the victim’s purse, belong to Rudy Guede. He was certainly at the crime scene. He certainly had some kind of interaction with Meredith Kercher that night. He has exactly the type of criminal profile that would lead him into this sort of circumstance. He is probably the murderer.
The thing that gives me pause is Guede’s persistent denial that he murdered Kercher. This guy is a smash and grab burglar. He does not think ahead. He does not plan his escapes. He does not enter the crime scene thoughtfully with the idea of maximum profit. All his crimes are crimes of desperation, crudely executed. I doubt he meant to murder Meredith Kercher. He was there to rob. Rape presented itself to him as an opportunity. He attempted it. She resisted. He murdered her accidentally. With criminals of this type you would expect the perpetrator to be remorseful and confess. They nearly always do. Rudy does not. That bothers me.
If we believe Rudy, however, we enter the next realm of probability. Could Meredith Kercher be unlucky enough to be burglarized by Guede and then have a couple of other bandit murderers stop by , too? Doesn’t seem likely to me. Was she MI5 or dealing drugs, herself? Did anybody check? I’m not happy with it.
I suspect you’ll agree with me within 20%. There are no scientists or law enforcement officials outside of Italy that give any credence to the prosecution’s analysis of this crime. As of yesterday, even the Italian government is backing away from these police and prosecutors. Knox and Sollecito will be freed on the review of the DNA analysis. Knox will be given time served on her so-called “slander” charge and they’ll send her home to Washington state before her 24th birthday in July.
Assigning probability zero to something is not in general a good idea.
Why? What negative result follows when you assign probability zero to something? I haven’t consciously experienced any so far.
It might help to read the linked essay in my previous comment. When one assigns probability 0 or 1 to something, one cannot update based on evidence. That is, if one really believes that, there should be no amount of evidence that will convince you. That’s problematic if one is trying to be a rationalist. And if there is some amount of evidence that can convince you then you don’t really assign probability 0 or 1 to the claim.
One either GOES to the restroom or one DOESN’T go to the restroom. It’s not a 1% possibility thing. Something is or it isn’t. Not a good idea? Based on the fact that once in the courtroom it’s nothing but a big pokergame. But there is 0 evidence.
Probabilities are measures of how certain one is about something. Thus, for example if one is going to estimate how likely a coin is to come heads one will say .5. This doesn’t change the fact that the coin either will come up heads or come up tails. But I don’t know which one it will be. Similarly, if something in the past may or may not have occurred, then you can assign numeric estimates to how likely it is to have occurred.