Most criminal defense lawyers know that science in Court is not like CSI. Fingerprints are rarely used anymore. The FBI has admitted that the evidence used to convict thousands of people over several decades (bullet lead analysis, bite mark analysis) has proven to be completely unreliable junk science. Now, what has previously been seen to be near perfect science, DNA, is turning out to be questionable as well.
In this report from NPR, the Texas Forensic Science Commission has concluded that there is something wrong in how state labs were analyzing DNA evidence, specifically when the labs were analyzing samples from several different people. In one specific case, it changed the probabilities from the testified 1 million to one, to 1 in 40. Do you think that just might have made a difference at the trial?
The problem our criminal justice system has is that these findings are presented at trial like facts that have no room for error. Lab techs and prosecutors use this evidence as holy grails that point to nothing other than a person’s guilt. And juries and judges buy it, even though the so called “science” has been proven to be fallible over and over again. Thousands of innocent people have gone to prison over things that were considered “settled science” at the time.
The bottom line is that when someone gets on the stand and gives probabilities, those numbers should be considered with extreme caution. Someone’s innocence depends on it.