On the morning of Oct. 8, 1988, Lloyd Martz's ex-wife found him slumped in the corner of his bedroom. The 49-year-old car leasing agent had been stabbed and beaten with a fire poker.
During his autopsy, a medical examiner noticed bite marks on his shoulders. Inside his Mercedes convertible, investigators found strands of brown hair.
Forensic experts concluded that the hairs and the bite marks belonged to Michael Annen, a 24-year-old Publix clerk. Martz had picked him up the night before his murder, when Annen was hitchhiking.
Prosecutors made Annen a deal: Plead guilty and avoid the death penalty. In September 1990, Annen accepted the offer and was sentenced to life in prison. After his hearing, he confessed to the murder, police records state.
But next February, Annen, now 51, will be back in court alongside Pinellas-Pasco Public Defender Bob Dillinger and the Innocence Project of Florida. They will explain that Annen was convicted on flawed science.
The case is the latest in a wave of nationwide appeals questioning the validity of forensic evidence long relied on by the criminal justice system.
"It's incredibly difficult to get courts and the forensic community to acknowledge error," said Chris Fabricant of the Innocence Project of New York, "to acknowledge that a technique that has been used for a century is no longer valid."
• • •
The first recorded use of hair evidence in a criminal case dates back to 1855 on a cotton plantation in Mississippi.
An overseer was found dead. Witnesses found a noose nearby that contained a few stray hairs. The strands "corresponded exactly in color and length" with the victim's, according to an article published in a Virginia legal journal last year.
Through the decades, police and prosecutors used microscopic hair analysis to identify suspects.
It works like this: lab examiners mounted hairs found at a crime scene to slides and, with a microscope, compared those strands to hairs from a suspect. If the strands are "microscopically similar," they concluded that the hair came from or "likely" came from the suspect.
But hair analysis has now been widely challenged. In 2005, Congress directed the National Academy of Sciences, a nonprofit of the country's leading researchers, to study all forensic sciences, including hair analysis. Their conclusion: This technique can't identify one person to the exclusion of all others.
After three men from Washington D.C. convicted in part by hair evidence were exonerated by DNA between 2009 and 2012, the FBI launched a review into these kinds of cases.
To date, the agency, with help from the Department of Justice, the Innocence Project, and the National Association of Criminal Defense Lawyers, have reviewed more than 2,000 cases.
Nearly 170 of them are from Florida, an FBI spokesperson said.
Among the cases is Annen's. Court records show that in 2014, Pinellas-Pasco State Attorney Bernie McCabe
got a letter from the Department of Justice. The lab report in Annen's case, it read, "exceeded the limits of the science."
The FBI's review triggered appeals nationwide. A Wisconsin man's 1990 rape conviction was overturned this year after DNA testing proved the hair found at the crime scene wasn't his.
In North Carolina, another man who wrongfully served 25 years in prison for rape will get a $9.5 million settlement.
In Massachusetts, prosecutors recently dismissed the case against George Perrot, who spent nearly 30 years in prison for sexual assault.
"These cases are both exhilarating and devastating," said Fabricant, who represented Perrot. "They're exhilarating because, for obvious reasons, freeing the innocent, there can be no higher purpose as a lawyer."
"They're devastating because I recognize and all of the other attorneys at the Innocence Project recognize that there are thousands of George Perrots out there."
• • •
Dr. Richard Souviron remembers the Annen case.
On Jan. 17, 1989, a prosecutor and a Pinellas Sheriff's detective flew to his Coral Gables office. They showed Souviron photos of the bite marks on Martz's body and a wax bite impression of Annen's teeth, records state.
"With a reasonable degree of forensic dental probability," a detective wrote in a report, "the casts match the bite marks."
Reached by phone this week, Souviron said he told investigators that he needed more time with the photos and the impressions. But that never happened because Annen pleaded guilty.
In 2013, a friend of Annen's e-mailed Souviron the autopsy photos. He replied: "I would be hard pressed to even call them bites."
"I changed my mind," Souviron told the Times. "I've had, what, 20 years more experience or 25 years more experience. At that time, I'd done only maybe a dozen cases."
Just like hair analysis, recent studies show that bite marks can't be used for identification.
In its 2009 report, the National Academy of Sciences pointed to flaws in this kind of analysis. Although it could "sometimes reliably exclude suspects," researchers wrote, the process can't be used to make a match.
"Unfortunately, bite marks on the skin will change over time and can be distorted by the elasticity of the skin, the unevenness of the surface bite, and swelling and healing," the report reads.
Souviron gained fame during Ted Bundy's 1979 trial, when he testified that Bundy's crooked teeth matched the indentations of a victim's bite mark.
Although it can't definitively identify a suspect, bite mark comparison can exclude people, he said. He recalled a case in South Florida where he looked at the marks on a victim and told police the suspect would be missing a lower front tooth.
The man they had recently arrested had a full set of teeth. Police released him.
"They found somebody that ended up being the right guy," Souviron said.
• • •
But there is one kind of science that defense attorneys and prosecutors agree is the forensic holy grail.
"DNA is absolutely the most reliable," said Keith Findley, senior adviser at the Wisconsin Innocence Project. "It is the model, it is the standard, and none of the others come close."
It's what prosecutors will present during Annen's hearing next year. His defense attorneys declined to comment since the case is pending.
Most of the evidence in the case, including the hairs and the fire poker, have been destroyed, which is typical in decades-old cases. But fabric clippings from a pair of gloves remained.
At the time of Martz's murder, one glove was found in his condo and the other in his car. Assistant State Attorney Richard Ripplinger, the original prosecutor in the case, submitted the fabric for testing, which detected Annen's DNA.
"We were just fortunate," Ripplinger said, "that they still had those cuttings."
Times senior researcher Caryn Baird contributed to this report. Contact Laura C. Morel at firstname.lastname@example.org. Follow @lauracmorel.