Police suspected that William Richards had killed his wife, Pamela, the night her body was found.
There was no sign of an intruder, and police said the crime scene appeared staged. But Richards denied killing Pamela, and authorities had trouble obtaining a conviction.
After two juries hung, a third heard new evidence: A forensic odontologist testified that a "bite mark" on Pamela's hand was consistent with Richards' unusual dentition, a pattern the prosecution expert said was found in only about 2% of the population. That jury convicted, and a judge sentenced Richards to 25 years to life.
Ten years after the conviction, the prosecution odontologist recanted his testimony. Relying on new computer technology that made it possible to view the "bite mark" more clearly, the odontologist ruled out Richards as its source.
The California Supreme Court is now weighing his case, wrangling over what to do when forensic evidence is later discredited. A broad ruling could affect scores of criminal convictions. A narrow one would offer little hope to those convicted at least in part by so-called junk science.
"A lot of forensic science is not well validated to begin with, and that especially includes bite mark evidence," said Hastings Law School professor David L. Faigman, an expert on scientific evidence. He said the court could rule for Richards but limit the kinds of cases that would be affected. If discredited science alone became grounds for a retrial, he said, "then you are potentially opening up the floodgates."
The case against Richards, 63, was largely circumstantial. Richards said he discovered his wife's body after returning home from work shortly before midnight Aug. 10, 1993. But the sandy, sagebrush-pocked land around their remote, high desert home in San Bernardino County contained no unknown tire tracks or footprints. Nothing had been stolen.
Although Pamela was found naked from the waist down, blood splatter on her pants, found nearby, suggested she had been fully dressed when her head was crushed with a cinder block steppingstone. The blow had been so brutal that it left one of her eyes dangling from its socket. An examination failed to detect signs of rape.
Both Pamela, 40, and her husband openly had affairs, and the prosecution contended that Pamela was about to leave her husband for her lover. There were written plans in place for the couple to divide their assets.
Pamela worked as a waitress, and Richards was a mechanical engineer. They had been married for more than 20 years, and their rural home consisted of a trailer, a shack and a shed. A generator provided power.
San Bernardino County sheriff's deputies viewed Richards' behavior that night as suspicious. He told them his wife was "stone cold" dead, though a deputy said the body appeared fresh.
He showed police evidence, including her bloody slacks, across the property and suggested how the crime must have been committed. Police said it was too dark for Richards to have spied the pants and other evidence unless he knew where they had been left.
Richards had clocked out from work that night, and according to the police time line, he had only eight minutes to kill his wife. Investigators said she was strangled and bludgeoned.
Writing from prison, Richards persuaded the California Innocence Project to take his case. San Bernardino County Superior Court Judge Brian McCarville held a hearing on Richards' case in 2009.
The hearing discredited the old bite mark testimony and unveiled new evidence. DNA from an unknown male was found on the steppingstone used to smash Pamela's head, at spots a law enforcement investigator said the killer would have touched. Another test ruled out Richards as the source of a 2-centimeter hair found under one of Pamela's fingernails.
The defense also demonstrated that a tiny tuft of blue thread that prosecutors said was found under Pamela's fingernail was not visible in early photographs of her hand. Richards had been wearing a blue shirt the night of the murder. Police blamed the poor quality of the photo for the discrepancy.
Prosecutors said the DNA could have come from someone sneezing on the steppingstone, and the hair could have been picked up anywhere. None of the new evidence was enough to overturn the conviction, they argued.
McCarville, however, ruled that the developments undermined the prosecution's case and pointed "unerringly to innocence." He overturned the conviction.
A three-judge appellate panel disagreed and reinstated it. New evidence, they ruled, is not enough to reopen a case unless it clearly points to innocence, and this evidence did not meet that high standard.
Richards' lawyers have asked the California Supreme Court to rule that the dental expert's original testimony amounted to false evidence. A demonstration of false evidence is enough to get a retrial if it significantly affected the verdict.
The California Supreme Court appeared divided during oral argument in September. Justice Carol A. Corrigan suggested that expert testimony is opinion and cannot be considered true or false. Several justices appeared to be siding with her.
Justice Goodwin Liu, however, countered that it was not the opinion that constituted a falsehood but the fact that the expert represented it as being backed by science.
Deputy Dist. Atty. Stephanie Zeitlin cautioned the jurists about the effect of their decision. If inaccurate expert testimony could reopen criminal convictions, the finality of criminal trials would be "eviscerated," she said.
"The people are entitled to finality; the victims are entitled to finality," Zeitlin argued. "It should not be subject to a mere battle of the experts."
But bad science has been implicated in more than half of all wrongful convictions, according to Jan Stiglitz, co-director of the California Innocence Project. And even if the court favored Richards, he said, it would remain difficult for inmates to win new hearings.
"I don't think we really have to worry about the floodgates," said Stiglitz, a professor at San Diego-based California Western School of Law.
If Richards prevails, San Bernardino County will retry him, said retired Assistant Dist. Atty. Michael W. Risley, who prosecuted Richards. "There is not a doubt in my mind that anybody else other than William Richards murdered Pamela," Risley said.
Asked whether a jury hearing the new evidence would convict, Risley replied: "Good question."