01-06-2013, 08:30 AM
Walter,
what you have published is very close to the conclusions I reached from the link I posted. Looks like the CS notice board is moving into the tried and tested peer group review methodology.
For me any lab results quoting odds of billions to one for anything has to be questionable. Consider the chances of human error in labelling samples incorrectly, or the well known but not controlable issues of degradation or contamination of sample collected from crime scene, potential cross contamination from suspect DNA sample handled in same lab, let me add two more -unfortunate typos on a querty keyboard where the reader reads the right result but characters are not perfect, identical twins that have not met for decades. Hidden in that long report I linked to is the comment that ideally suspect and crime scene DNA should not be handled in the same room.
I want to maintain brevity in this posting but traditional police methodology still has a key role. From page 68 " although any reported identification frequency represents an estimate to be used in conjunction with other evidence linking a defendent to a crime".
While this appears to take the thread off topic I still maintain that Boris Ivanov might be able to put up a successful defense in court. Based upon ..............Ivanov trained with the Houdini chess engine and tried to imitate its style. He succeeded - no other leading chess player has had the same training method. My (pseudo)client has been prejudged by journalists and discussion fora making a fair trial impossible.
The odds against errors in reports made by journalists are considerably shorter than for reports made by DNA testing labs
Harry Houdini was a master of escapology; how appropriate that his name was used for this chess engine. If frustrated by the prowess of Houdini just remember that you are better at kick boxing than the computer -
<!-- m --><a class="postlink" href="http://www.nytimes.com/learning/general/onthisday/bday/0324.html">http://www.nytimes.com/learning/general ... /0324.html</a><!-- m -->
PS who saw my deliberate typo on first read through.
what you have published is very close to the conclusions I reached from the link I posted. Looks like the CS notice board is moving into the tried and tested peer group review methodology.
For me any lab results quoting odds of billions to one for anything has to be questionable. Consider the chances of human error in labelling samples incorrectly, or the well known but not controlable issues of degradation or contamination of sample collected from crime scene, potential cross contamination from suspect DNA sample handled in same lab, let me add two more -unfortunate typos on a querty keyboard where the reader reads the right result but characters are not perfect, identical twins that have not met for decades. Hidden in that long report I linked to is the comment that ideally suspect and crime scene DNA should not be handled in the same room.
I want to maintain brevity in this posting but traditional police methodology still has a key role. From page 68 " although any reported identification frequency represents an estimate to be used in conjunction with other evidence linking a defendent to a crime".
While this appears to take the thread off topic I still maintain that Boris Ivanov might be able to put up a successful defense in court. Based upon ..............Ivanov trained with the Houdini chess engine and tried to imitate its style. He succeeded - no other leading chess player has had the same training method. My (pseudo)client has been prejudged by journalists and discussion fora making a fair trial impossible.
The odds against errors in reports made by journalists are considerably shorter than for reports made by DNA testing labs
Harry Houdini was a master of escapology; how appropriate that his name was used for this chess engine. If frustrated by the prowess of Houdini just remember that you are better at kick boxing than the computer -
<!-- m --><a class="postlink" href="http://www.nytimes.com/learning/general/onthisday/bday/0324.html">http://www.nytimes.com/learning/general ... /0324.html</a><!-- m -->
PS who saw my deliberate typo on first read through.