Police found the naked body of Diana Sylvester near her Christmas tree.

For more than three decades, Sylvester's slaying went unsolved. Then, in 2004, a search of California's DNA database of criminal offenders yielded an apparent breakthrough: Badly deteriorated DNA from the assailant's sperm was linked to John Puckett, an obese, wheelchair-bound 70-year-old with a history of rape.

The DNA "match" was based on fewer than half of the genetic markers typically used to connect someone to a crime, and there was no other physical evidence.

Puckett insisted he was innocent, saying that although DNA at the crime scene happened to match his, it belonged to someone else.

At Puckett's trial earlier this year, the prosecutor told the jury that the chance of such a coincidence was 1 in 1.1 million.

Jurors were not told, however, the statistic that leading scientists consider the most significant: the probability that the database search had hit upon an innocent person. In Puckett's case, it was 1 in 3.

The case is emblematic of a national problem, the Los Angeles Times has found.

Prosecutors and crime labs across the country routinely use numbers that exaggerate the significance of DNA matches in "cold hit" cases, in which a suspect is identified through a database search.

Jurors are often told that the odds of a coincidental match are hundreds of thousands of times more remote than they actually are, according to a review of scientific literature and interviews with leading authorities in the field.

Two national scientific committees, including the FBI's DNA advisory board, have recommended portraying the odds more conservatively. But interviews with expert witnesses and DNA analysts from crime labs across the country show that few if any have adopted that approach.

The FBI lab, which oversees the nation's offender databases, has disregarded the recommendation of its own advisory board, bureau officials acknowledged. So far, the courts have ruled in law enforcement's favor on this issue.

As a result, some experts fear, a technology best known for freeing the innocent could be causing its own miscarriages of justice.

For years, police used DNA only to compare a crime-scene sample to a single person they had other reasons to suspect. In court, prosecutors could legitimately cite the remote odds that someone selected at random off the street would share the same genetic profile.

But in cold hit cases, the investigation starts with a DNA match found by searching thousands, or even millions, of genetic profiles in an offender database. Each individual comparison increases the chance of a match to an innocent person.

Nevertheless, police labs and prosecutors almost always calculate the odds as if the suspect had been selected randomly from the general population in a single try.

The problem will only grow as the nation's criminal DNA databases expand. They already contain 6 million profiles.

The danger of implicating an innocent person through a cold hit becomes particularly acute with crimes that are decades old, like Sylvester's killing. Witnesses have died, evidence has been lost and memories have faded. All of this puts greater weight on the DNA evidence, which also may have deteriorated.

Prosecutors say they don't mention the database problem to juries because it could unfairly reveal that a defendant has a criminal record. They also say that using more conservative statistics would not make a difference in many cases, because the odds would still overwhelmingly favor the prosecution.

Now the California Supreme Court is weighing the statistical and other concerns raised by cold hit cases -- the highest court yet to consider them. People vs. Nelson, which involves the 1976 murder of a college student in Sacramento, is scheduled for oral arguments Thursday.

It could influence how such cases are tried across the country and undermine some convictions in California.

...

It was three days before Christmas, 1972, in an apartment building near San Francisco's Golden Gate Park. Helen Nigodoff, the building's landlady, heard thumping and other noises from the apartment above hers.

Nigodoff went upstairs to check on Sylvester. The door to the apartment was open. She saw a bearded man attempting to cover his face with his hand. "Go away!" he screamed, Nigodoff later told investigators. "We're making love!"

Nigodoff returned to her apartment and called police, who found Sylvester's body. There were no signs of forced entry or a struggle.

Initially, police focused on a street artist named Robert Baker, then 32, who had recently escaped from a mental institution. He had been identified as a suspect in a rape that occurred two weeks before Sylvester's killing and four blocks from her apartment.

The victim in that earlier case and a friend identified Baker as the rapist, but for reasons that are now unclear, authorities decided they did not have enough evidence to prosecute him.

Baker also was identified as the man who, four days after Sylvester was killed, harassed a woman and a young girl and followed them to their home a few doors from Sylvester's apartment, police records show.

He was not charged in that incident -- but he was booked for the rape four blocks from Sylvester's apartment.

Police thought he might also be Sylvester's killer.

In Baker's van, police found a parking ticket with drops of blood on it. The blood type was O, the most common, the same as Sylvester's. DNA testing was not available then.

The crime-scene fingerprints did not match Baker's, and he was never charged with Sylvester's killing. He died in 1978.

The investigation into Sylvester's death went cold.

In 2004, San Francisco police received state funding to test DNA evidence from unsolved homicides.

A crime-lab analyst retrieved from storage a swab that had been taken from Sylvester's mouth back in 1972. It contained sperm presumably left behind by her assailant. Though the sample was severely degraded, the analyst was able to develop a partial DNA profile.

Most profiles examine 13 genetic markers, usually enough to distinguish between two people. The sperm sample produced only 5 1/2 and was mixed with traces of DNA from another person, probably Sylvester, the analyst determined.

It was too little information to search California's DNA database; for that, at least seven markers were needed. To bridge the gap, the analyst used evidence from additional markers that were too faint to be considered conclusive. When this expanded profile was compared with the 338,000 profiles of convicted offenders in the state's database, investigators got a cold hit. The profile matched Puckett's.

His genetic information was in the database because he had pleaded guilty to two rapes and a sexual assault in the Bay Area in 1977, five years after Sylvester's killing. After his release in 1985, he had committed misdemeanor battery, but since 1988 his record had been clean.

On Oct. 12, 2005, San Francisco Police Inspector Joseph Toomey knocked on the door of the Stockton, Calif., mobile home where Puckett lived with his wife.

Puckett hobbled to the door clutching a bag of urine, explaining that he had just had triple-bypass heart surgery, Toomey later testified. During a 2 1/2-hour interview, Toomey showed Puckett photos of Sylvester and the crime scene.

"You never had sex with that girl?" Toomey asked.

"No," Puckett said.

"You've never been in the house we showed you?"

"No."

Puckett volunteered to give Toomey a fresh sample of his DNA. When it matched the same markers as the earlier test, Puckett was arrested and charged with murder.

The case was assigned to the San Francisco public defender's office, which had recently hired a DNA specialist named Bicka Barlow.

Barlow had earned a master's degree in genetics from Cornell University before going to law school. A veteran of dozens of DNA cases, she considered Puckett's the weakest match she had ever seen.

Typically, prosecutors rely on FBI statistics to estimate the rarity of a particular DNA profile in the general population. This calculation is known as the Random Match Probability.

The chance that two unrelated people will share the same 13 markers can be as remote as 1 in a quadrillion -- a number with 15 zeros. Because the match in Puckett's case involved only 5 1/2 genetic locations, the chance it was coincidental was higher but still remote: 1 in 1.1 million.

But Barlow thought this figure vastly exaggerated the strength of the evidence. It did not take into account how Puckett had been identified: through a search of a large database.

The general-population figures used by prosecutors portray the odds of matching crime-scene DNA to a single, randomly selected person.

But because database searches involve hundreds of thousands or millions of comparisons, experts say using the general-population statistic can be misleading.

Barlow argued during pretrial hearings that the jury should be told about the recommendation of two leading panels of scientific experts, one convened by the National Research Council and the other by the FBI.

Both committees settled upon a statistical remedy to adjust for the many individual comparisons made during a database search. It has been widely but not universally embraced by scientists.

In every cold hit case, the panels advised, police and prosecutors should multiply the Random Match Probability (1 in 1.1 million in Puckett's case) by the number of profiles in the database (338,000). That's the same as dividing 1.1 million by 338,000.

For Puckett, the result was dramatic: a 1-in-3 chance that the search would link an innocent person to the crime.

David Merin, a deputy San Francisco district attorney, argued that the statistical adjustment recommended by the scientific panels would be "confusing and misleading."

Superior Court Judge Jerome Benson ruled in favor of Merin. In doing so, he followed a California appellate court ruling on the same issue. The jury would never hear the 1-in-3 statistic.

At 2:30 p.m. on March 6, after 48 hours of deliberation, the jurors filed into the courtroom with a verdict: guilty of murder in the first degree.

Puckett, now 74, sat motionless in his wheelchair. He later was sentenced to life in prison, with the possibility of parole after seven years. His attorneys said they would appeal.

Interviewed outside court after the verdict, jurors said they had struggled to weigh the different statistics. In the end, however, jurors said they found the 1-in-1.1-million general-population statistic Merin had emphasized to have been the most "credible" and "conservative." It was what allowed them to reach a unanimous verdict.

"I don't think we'd be here if it wasn't for the DNA," said Joe Deluca, a 35-year-old martial arts instructor.

Asked whether the jury might have reached a different verdict if it had been given the 1-in-3 number, Deluca didn't hesitate. "Of course it would have changed things," he said. "It would have changed a lot of things."