IT WAS, BY ALL ACCOUNTS, the high court’s low point. A teenage girl named Deanna Ogg had been raped, bludgeoned, and stabbed to death on a late September afternoon in 1986 near the tiny town of New Caney, north of Houston. Roy Criner, a 21-year-old logger, was arrested after three friends said that, within hours of the time of Ogg’s death, Criner had bragged about picking up a hitchhiker, threatening her with a screwdriver, and forcing her to have sex. No other evidence tied him to the crime, but Criner was convicted and given 99 years for aggravated sexual assault. In 1997 newly available DNA tests showed that the sperm found in Ogg was not Criner’s. To be certain, the Montgomery County district attorney did a second test in the state’s lab and got the same results. Criner’s attorneys moved for a new trial, and in January 1998 the trial court agreed he deserved it.
Four months later, the Texas Court of Criminal Appeals, the highest criminal court in the state, went against law, science, and, it seemed, all common sense when it wrote, “The new evidence does not establish innocence,” and overruled the trial court. Sharon Keller, who had been on the CCA only a little more than three years but was rapidly becoming the court’s philosophical leader, cited the incriminating statements to the three friends as “overwhelming, direct evidence” of Criner’s guilt. New evidence ofinnocence, she argued, had to be so clear and convincing that no reasonable jury would have convicted Criner had it known about it. DNA, she said, was not enough. Keller noted that perhaps Criner had worn a condom or failed to ejaculate. There was also testimony, she wrote, that the victim had said that she “loved sex,” so perhaps she had had sex with someone and then met her demise at the hands of the logger. These theories had not been alleged at trial, nor was there evidence that Ogg had had sex with anyone else within 48 hours of her death, and court watchers wondered why an appellate judge was