Death Row Software

By Deane Barker on April 27, 2005

Software models death row outcomes: Two reseachers created a software program to examine the demographics of death row inmates. The results were highly predictive…which is bad, if you read the second paragraph below.

The neural network, which learns by constantly scanning the data for patterns, was given 1,000 cases from 1973 to 2000 where the outcome was known. Once trained on that information, it was fed another 300 cases but without the outcome included. That’s when its prediction proved highly accurate.

What some observers find alarming about the outcome is that the 19 points of data supplied on each death-row inmate contained no details of the case. Only facts such as age, race, sex, and marital status were included, along with the date and type of offense.



  1. What is “, along with the date and type of offense.” if not “details of the case”?

    I’m guessing they are trying to prove racial or other prejudice towards death row inmates.

    I’m on the edge as far as this debate is concerned. First I don’t think people should be convicted, especially on murder cases, without hard evidence. And for life or death sentences their should be DNA along with other information before conviction is sealed.

    I think our juror process now is a joke. If I killed someone and were on trial. I would seriously consider letting the judge determine the case, depending on judge of course ;-)

  2. The elephant in the room is pervasive corruption of the entire legal apparatus… if death sentences are this skewed consider how bad it is for cases that get far less attention.

  3. It wouldn’t take a rocket scientist to figure out a large percentage of them without knowing anything about the case except what State the trial is in. “The clues are already there, Harper says. For instance, the state where someone is convicted is key. California has the largest number of inmates on death row and has executed just two people in the last 38 years, whereas Texas has been “an execution machine,” he says.” I think I could predicedt most of the cases in California without even using a calculator let alone writing a program.

    I not sure I agree with Brian that DNA should be present for for life or death sentences. What about admission of guilt, multiple eye witnesses, etc. Does the mastermind of a heinous crime deserve death? What if there were a third person who convinced or paid Terry Nichols and Timothy McVeigh? Would that person be deserving of death? Surely there would be none of their DNA at the federal building. In fact, I don’t think there was any belonging to Nichols or McVeigh.

    It is still possible to plant DNA and therefore there could still be and innocent person on death row.

  4. hjmler I agree with you. However in the current legal system I think there are far too many convictions for life/death sentences that are overturned later with proper evidence (DNA in most cases). I guess it is too extreme to require DNA as in the case you stated for the Oklahoma bombing. My general thought it our legal system is not as justice oriented as it is prosecution oriented.

Comments are closed. If you have something you really want to say, tweet @gadgetopia.