Subscribe via RSS Feed

Scoring the Evidence Against Willingham

[ 77 ] June 13, 2011 |

So it seems likely that Rick Perry will run for president.

To distract myself, I’ve been reading Bill James’ new book Popular Crime. The chapter on Lizzie Borden, in an attempt to distinguish meaningful evidence from less or non-meaningful evidence, proposes a scale in which evidence is given weight, with a score of 100 meaning that someone has been proven guilty beyond a reasonable doubt. Something that’s close to being dispositive (a DNA match in a sex murder being probably the strongest evidence) would be 80; reliable evidence of previous threats against a victim might be 25, etc. The limitations of the method are obvious, but it’s an interesting starting point for analyzing the state’s evidence. (Incidentally, James scores the evidence against Borden at about 20.)

To pick a completely random example, let’s use this method to assess the evidence against Cameron Todd Wilingham:

  • “Scientific evidence” of arson: Literally worthless.   The damaged house is perfectly compatible with an accidental fire, and the “science” purporting to demonstrate otherwise is about as useful as astrology.   Zero points.
  • The testimony of Jimmy Webb: Webb, a mentally ill jailhouse snitch, testified that Willingham confessed to the murder. Webb testified that Willingham (who was executed because he  refused to cop a plea) decided out of the blue to confess to a stranger while passing by his cell.    This story 1)is extraordinarily implausible on its face, 2)contained assertions that were inconsistent with the evidence, 3)was told by an extremely unreliable witness with strong incentive to lie, and 4)was later recanted by Webb.    Basically, this is about as useless as such evidence could be.   Three points.
  • Willingham’s “character”: Amazingly enough, one of Willingham’s killers has essentially conceded that the two pieces of “evidence” above — the only evidence that the fire was arson in the first place — are of no value.    Essentially, the prosecution’s case is now “Willingham was a really bad guy so he must have done it.”   But — the arguments of another of Willingham’s killers notwithstanding — being a bad person or bad spouse aren’t capital offenses, and they’re relevant only insofar as they show someone capable of having burned his children alive.    So how relevant is the character evidence in this case?  Not very.  His penny-ante criminal history is neither here nor there in terms of a tendency to commit a crime of this magnitude.   His adultery and history of domestic violence would be of some relevance if he had been accused of killing his wife, but as evidence that he killed his children (who he apparently never harmed), it’s very weak tea.   It would carry some weight if there was actually evidence of arson, but in and of itself it’s almost nothing.   I’ll score it at 15 points, and I think that’s generous to the prosecution.
  • Alleged Failure to Act In Accordance With Arbitrary Standards of How Victims Should Act. Except in extreme cases, I’m very dubious about this evidence in any case; it’s impossible to predict how someone will react to an unfathomable tragedy within reason, and Willingham’s actions seem within a reasonable range.   Moreover, even the allegedly damaging witness testimony in this respect is unreliable; witnesses seemed to think his actions were worse after he was arrested than people reported at the time.   5 points, and again if anything I think that’s too high.
  • “Expert” Testimony of his Sociopathy: It was delivered by James Grigson, so essentially all it’s evidence of is that the state’s check cleared.    Let’s just say that among the evidence Grigson used to provide evidence that Willingham was a sociopath was the fact that he had Led Zeppelin and Iron Maiden posters as a teenager.  1 point.

I think that’s basically it.   So my final score is 24.    It would be irresponsible for a DA to seek an indictment based on a case this weak, and the fact that someone was executed based on it is an outrageous miscarriage of justice  that in a rational universe would permanently disqualify anyone involved from public office.  If, hypothetically, the governor who signed off on this execution and then suppressed an inquiry into what happened was a credible candidate for president, it would be a national disgrace.


Comments (77)

Trackback URL | Comments RSS Feed

  1. Vance Maverick says:

    Naive, non-tendentious question: how do real DAs assess this stuff? Are there places (regions, jurisdictions) where there’s a scoresheet even as formal as this?

    • Gregory says:

      Naive, non-tendentious question: how do real DAs assess this stuff?

      “How will prosecuting this case help my re-election chances?”

      (Or perhaps more generously, “How will not prosecuting this case hurt my re-election chances?)

    • Steve-O says:


      Speaking as someone who served as a criminal prosecutor for many years I would say the best answer is “seat of your pants.” Also, I’ve never known a prosecutor or office that uses any type of score sheet. Of course that’s tempered by the fact that I never practiced law outside of WA, but I have discussed this issue informally with a number of PA’s from other states (although interetingly enough no one from Texas).

  2. richard says:

    I haven’t read James’ book. I loved the original Baseball Abstracts but the only James’ wiritng I’ve read on a subject other than baseball statistics was his several articles that there was insufficient evidence to believe that Pete Rose had bet on baseball (of course written prior to Rose admitting he bet on baseball) and his crime analysis there was extremely weak. (Although I certainly agree with your analyis of the Willingham case). Has James ever reevaluated his erroneous Rose analysis especially in light of the new metric he has developed in Popular Crimes?

    • Scott Lemieux says:

      He’s certainly admitted that he was wrong about Rose.

    • Boudleaux says:

      Well, I don’t suppose it needs to be pointed out that saying that evidence known at a certain juncture is insufficient to support a conviction is not the same as saying that the accused is innocent.

      • richard says:

        But I read his analysis at the time and it was just silly. And admitting he was wrong, in light of Rose’s confession, is, of course, not the same thing as admitting the analysis was wrong.

        • richard says:

          answering my own question, I’ve learned that James still defends his position that the Dowd Report did not contain adequate evidence of Rose’s guilt (although no longer taking the position that Dowd and Fay Vincent lacked integrity and honesty) and that he takes a “perverse pride” in being the last person on Earth to be convinced of Rose’s guilt.

          • John F says:

            James’ analysis of the Dowd report was quite frankly terrible- and as someone who’d bought all but the 1st two Abstracts when they originally came out terribly disappointing. What I always found odd was that James was certainly no fan of Pete Rose (the player) – I mean he had his lapses in objectivity (especially when it came to the Royals)- but his defenses of Rose were almost religious in their refusal to even contemplate Rose’s guilt.

            • richard says:

              That was my impression as well. The Dowd report was based on the testimony of the co-conspirator but more importantly on the betting slips. As I recall, James argued that the betting slips couldn’t be believed because one of them referred to a game on a day where there was actually not a game that day – the obvious explanation was that the date was written down wrong on that one occasion but James too the position that this one mistake proved that all the betting slips were created after the fact to smear Rose.

              And its a little bizarre for him to now claim perverse pride in getting it wrong. Despite this, I’ll probably end up buying the Popular Crimes book

            • timb says:

              to be fair, as a Reds fan who believed it from the beginning and was glad to see it (it was the only thing that was going to get him fired as the terrible manager he was), believing it at the time meant believing that Rose was simply the dumbest man alive and people who did not see him on a daily basis probably didn’t know he was the dumbest man alive.

              Thankfully, he has shown since then that his sense of dignity and decorum is as lacking as is brain power

        • JB2 says:

          I remember the Rose essay – just bizarre. James was like a dog with a bone. Every great writer or thinker will sometimes attach themselves to a bad idea – like Robert Christgau’s “A+” grade for the 2nd New York Dolls record.

      • Malaclypse says:

        While theoretically true, in this case 1) there was almost no evidence that a crime actually even happened, and 2) I’m willing to assume that people don’t, in general, go to elaborate lengths to kill their children without a reason that could be discovered.

        • Malaclypse says:

          And now I realize your comment was about Pete Rose, not arson. Apologies for the lack of coffee and reading comprehension.

          • richard says:

            Yeah, there was certainly evidence that Rose committed a crime. I fully agree with you that there was almost no competent evidence to believe that a crime had been committed in the case of the Willingham fire. It looks like a tragic case of accidental fire.

    • David Nieporent says:

      Has James ever reevaluated his erroneous Rose analysis especially in light of the new metric he has developed in Popular Crimes?

      The problem was that James wasn’t engaging in Rose analysis. He’s by nature a contrarian, and since everyone was piling on Rose, he became (consciously or unconsciously) a Rose advocate. So rather than saying, “What’s the best possible reading of this evidence?” he said, “Can I explain away this evidence if I try?” He could, of course, but it was convincing only to those who wanted to be convinced.

      • Scott Lemieux says:

        I think this is right. He made a very convincing case that Dowd was acting as a prosecutor rather than a neutral arbiter, and was correct that a lot of the Dowd Report established that Rose was a scumbag while being neither here nor there in terms of whether he gambled on baseball. But the article failed because James did exactly what he accused Dowd of doing — essentially, of interpreting every piece of evidence in a manner most consistent with a belief that Rose was innocent even when Dowd’s assessment was more plausible.

  3. Flowers says:

    “the governor who signed off on this execution”

    Texas governors don’t sign off on executions, nor do they have unilateral authority to grant pardons. It’s part of the Texas Constitution. (His tampering with the forensic commission is shady, but that board wouldn’t have been able to even start the process for a pardon.)

    Texas has an extensive appeals process for capital cases. The fact that the judges are elected in partisan elections probably has more to do with the appeals and trial process being worthless than does the governor who has no say in the process.

    Being an actual lawyer, with an actual license to practice law in Texas, I think your analysis of the judicial situation leaves a lot to be desired.

    • Scott Lemieux says:

      If Perry had recommended that the board commute Willingham’s sentence, or replaced the board that voted to uphold the death sentence handed out to a man based on virtually no relevant evidence, your point would be meaningful. As it is, it’s completely irrelevant. Perry obviously bears substantial responsibility for Willingham’s execution — he picked the idiots who rubber-stamped it, didn’t use his de facto powers to stop it, and then did whatever he could to cover up the crime.

      Which, of course, doesn’t mean that many other elements of the Texas criminal justice system didn’t fail here, although since nothing in my post suggests otherwise your point is a silly non-sequitur.

    • Lindsay Beyerstein says:

      Texas governors do have the power to grant stays of execution. Willingham asked Perry for a stay of execution and Perry refused.

      • richard says:

        As I recall, a Texas governor can only grant one stay of execution. Even if he had granted that one stay, nothing would have changed because Willingham had exhausted all appeals and all requests to the parole board for commutation. I’m not so sure a governor can remove the parole board after it makes an adverse decision and then have the new board decide the issue again. But he certainly could have asked the board to commute the decision so bears some responsibility.

  4. The process by which “capital-qualified” juries are choses produces panels that are biased in favor of conviction.

    • Dudley Sharp says:


      Trial cases in the US result in conviction about 80% of the time, quite remarkable when you understand the burden of proof is the prosecutors.

      One would hope that the conviction percentage is much higher with capital cases, where prosecutors must have a much stronger case than in others, because of what’s at stake.

      I hope it is at least 95%. It should be.

  5. Adam says:

    Please let me know when American voters start to care about the execution of a potentially-innocent man. I’ll go tell the ghost of Karla Faye Tucker.

    • Scott Lemieux says:

      Whether one agrees with Tucker’s execution or not, she was almost certainly guilty.

      • richard says:

        She admitted her guilt in the course of requesting that her sentence be commuted. The only questions were whether her crimes (or any crimes) justified the death penalty or whether her subsequent rehabilitation during 14 years of prison justified commutation (the latter argument formed the basis of her argument to the Texas authorities)

        • Scott Lemieux says:

          As I’ve said before, Bush (given his erroneous belief in favor of the death penalty) acted correctly in supporting Tucker’s execution. Essentially, the case for Tucker was “attractive white women should be exempt from execution.” Which makes about as much sense as “abortion is murder but women aren’t rational enough to be held at all responsible.”

      • Dudley Sharp says:

        She was 100% guilty.

  6. Lindsay Beyerstein says:

    Tampering with the forensic commission is even worse than failing to commute a death sentence (where legally possible).

    Failure to commute a death sentence could, in theory, be a good-faith decision based on the governor’s own flawed interpretation of the facts.

    Whereas, tampering with the forensic commission suggests a deliberate attempt to cover up a miscarriage of justice.

    I remember then Texas Gov. George W. Bush mocking Karla Faye Tucker’s plea for clemency. Was she misinformed about the power of the governor when she begged for her life?

    • Dudley Sharp says:

      Some common sense is needed.

      It was impossible to cover up anything in the Willingham case. It was laready a public case being investiagted by a public commission.

      By replacing the mebers of the commission, Perry brought even more scrutiny to both the case and to Parry.

      Very obvious.

      No cover up was remotely possible and di not occur, as it couldn’t.

      • Anonymous says:

        You forgot to follow the rest of the ‘common sense’ chain you started, that by removing membership that was unbiased toward Willingham and adding membership that was biased against Willingham, Perry ruined your stupid biased argument before you even made it.

  7. Linnaeus says:

    A spectre is haunting Texas.

  8. DivGuy says:

    1) Perry, if he runs, is the frontrunner for the nomination.

    2) The state of the economy suggests a not-incompetent Republican should have at least a 30% chance at the presidency, and it could go higher.

    3) This is George W Bush: The Next Generation, and he could fucking be our president in eighteen months.

    • BigHank53 says:

      It’s not too late to move your retirement funds into canned goods and ammunition.

    • Scott says:

      Trust me. He can’t win. He has zero charisma, he’s completely unable to disguise either his arrogance or his sense of entitlement, and he’s widely disliked in his own state. And the “Texas Miracle” is about to come crashing down around his ears.

      • Anonymous says:

        He has zero charisma, he’s completely unable to disguise either his arrogance or his sense of entitlement, and he’s widely disliked in his own state.

        All of which was true of Bush in 1999, with the possible exception of being disliked in his own state.

        • Malaclypse says:

          gah – above was me.

        • mpowell says:

          No, this is wrong. Bush has plenty of legitimate charisma as far as a certain set of people are concerned. Perry does not. People will claim he does, but they will be lying. They were legitimately not lying about Bush when they claimed that a decent portion of the public liked the man.

          And nothing compares to the problems with the Texas economy. It’s easy for Perry to boast about things now, but get into a debate where people can start talking about the enormous smoking hole in the state budget and Perry is going to be ruined. And that’s before you even get to the general election where Obama will be hitting him with the patently idiotic/evil steps the legislature is taking to attempt to fix their budget.

          It’s not impossible for him to win, but Perry is a terrible candidate.

  9. CapnMidnight says:

    Alas, I was hoping for the evidence against Josh Willingahm. Of course, the evidence is right before me every time I listen to an A’s game, but I looked forward to a sabre-tronic deconstruction. I will have to content myself with being angry about the death penalty. Again.

  10. Jeff R. says:

    I think you’re mixing up Mr. Death (Leuchter) and Dr. Death, James P. Grigson who was referred to in the New Yorker article. I’m not sure it changes the final calculus very much.

    And, can you give negative points for the arson “experts”?

  11. mpowell says:

    Referring back to Bill James proposal, I haven’t read his book, but I’m not sure an additive process is the best way to go about this.

    You can have a guy with an undeniable motive and opportunity but if the cause of death is clearly not murder, he’s not guilty. I could imagine breaking up the evidence into different categories like evidence of a crime taking place, motive of the suspect and opportunity for the suspect and taking something like the geometric mean of the additive results in these 3 categories.

    I think James’ proposal is actually quite a decent idea but it requires more tweaking before it could really be useful, in my opinion. It’s not particularly interesting to note that his method correctly rejects the claim that Willingham’s guilt was established at a reasonable level. Because that’s trivial.

    • Scott Lemieux says:

      I think that’s right; really, the factors are interactive. Willingham’s background would take on some relevance if combined with real evidence but without it worth pretty much nothing.

      One thing I do agree with James about — actually the point he was responding to — is that “means/motive/opportunity” is a useful way to focus on suspects but not a useful way to think about guilt or innocence. And the Willingham case is an excellent example: he had means and opportunity and if you squint really hard maybe motive, but there’s no evidence that he actually did anything.

  12. Peter Muldoon says:

    I was going to make essentially the same point mpowell did, because this post really bothered me. There is zero evidence of arson. Therefore, there was no crime, and the rest of the evidence doesn’t matter.

    All of the other evidence purports to show that IF a crime was committed, Willingham was the one who committed it. As you scored the evidence of arson to be zero, the whole case should scored zero.

    I don’t see how this method of determining guilt has much use. The usefulness of evidence is often predicated on whether other evident is true.

    Of course it’s outrageous that we would execute Willingham when we have 24% confidence that he’s guilty; it’s also outrageous that you believe there is a 24% chance that he is guilty of a crime which you stipulate was never committed!

  13. Dudley Sharp says:

    The fire most certainly could have been arson and the evidence is certainly stronger for arson than not.

    The problem with an arson fire is when there is a flashover, which most think happened with this fire.

    A flashover may burn away any evidence of arson.

    In other words, what many experts have said is, if arson, the flashover made that evidence dissapear, making the fire indistinguishable from an accidental fire.

    Other, non forensic evidence is strong for guilt and arson.

    Some experts still insist that there was solid evidence of arson.

    Please review:

    8) “Cameron Todd Willingham: Another Media Meltdown”, A Collection of Articles

  14. […] column from Coates. After bringing much-needed national attention to Rick Perry’s role in the murder of Cameron Tood Willingham, the punchline says it all: Whatever one thinks of the death penalty, the accounts of those who […]

  15. […] “person didn’t act like a victim is allegedly supposed to act” routine was also a major part of the case against Cameron Todd Willingham, which should tell you what you need to know about its value. […]

    • stevo67 says:

      Darlie Routieere is another Tx defendant convicted of capitol murder primarily for not acting in accordance with public standards of victimization. She was gravely injured in the attack that cost the lives of her two small sons – though the prosecution claims the wounds were self-inflicted. What really ruined her case was that she gave a bizarre birthday party for one of her two deceased sons that made the Dallas-Ft Worth TV news – but what was not shown was the somber memorial service before the party. After the broadcast, the police stopped inquiring about any other leads and focused solely on her. After the Defense Atty made the mistake of asking and receiving change of venue to a rural community – the conviction was almost automatic.

  16. […] apparently the dumber version of Michelle Bachmann, accessory to murder Rick Perry, will indeed be seeking the Republican nomination.    Let’s hope I’m […]

  17. […] 2004 for allegedly killing his three children by burning down his family’s house. This execution was an appalling miscarriage of justice. The case against him was a complete shambles, with the evidence that the fire was deliberately set […]

  18. […] it is in fact possible for someone to get executed in Rick Perry’s Texas on the basis of no reliable evidence whatsoever, and not only will Perry let it go forward but he’ll quash any subsequent […]

  19. […] where it is impossible to prove innocence to absolute certainty but we know that the state has no evidence that a crime was committed, let alone that Willingham committed it. But it is still utterly […]

  20. […] Todd Willingham.   A governor and his death panel can proudly sign off on the execution of a man who is clearly legally innocent and almost certainly factually innocent.   And not only will you not pay any political price for it, you can do this and be a credible […]

  21. […] Perhaps in a decade or two a Texas governor will pardon Cameron Todd Willingham… […]

  22. […] it happens, this also came up during the Willingham case. And while Casey Anthony is a different issue in that it’s very possible that she actually […]

  23. […] of executing an almost certainly innocent man — which, Lane’s hand-waving aside, is not merely hypothetical — really is […]

  24. […] spend nearly 30 years in prison for a murder he didn’t commit, based on evidence weaker than the “evidence” that convicted Cameron Todd Willingham. The state had no physical evidence, a credible alternative theory based on people who (unlike […]

  25. […] they looked at Texas and felt that its executions of innocent people don’t happen frequently […]

  26. […] not sure what the alternative to arbitrarily selecting a small subset of people who may or may not be guilty of heinous crimes and killing them with methods that may or may not be humane […]

  27. […] he wouldn’t cop a plea decided to spontaneously confess to a prisoner in another cell — doesn’t have a lot of room to get less credible. It underscores that the state of Texas always had […]

  28. […] So how humane is Texas’s death penalty? I would like to ask Cameron Todd Willingham. Except that he was killed by the state of Texas. Even though the state had essentially no evidence that he committed the crime for which he was accused. […]

  29. […] still alive knows to an absolute certainty whether Willingham committed arson. But we do know that the case against him was so weak that no responsible prosecutor should have even indicted him. And it’s equally possible that he […]

  30. […] in a state where you can be executed even when the state has no actual evidence that you’ve committed any crime at all, not executing actually killers because they’re severely mentally ill is probably going to be […]

  31. […] As improbable as it might seem that Texas might be preparing to execute an innocent man, they very likely are: […]

Leave a Reply

You must be logged in to post a comment.