DENVER — Three people convicted of murder have been released from prison because their cases were tainted by a now discredited theory that bullets found at a crime scene could be linked to bullets found in possession of suspects.
Nearly five years after the FBI abandoned its so-called comparative bullet lead analysis, the FBI has yet to complete its review of nearly 2,500 cases where law enforcement used such evidence to investigate a case.
So far, the agency has found 187 cases where so-called comparative bullet lead analysis evidence was not only used in the investigation, but came into play at trial where FBI experts provided testimony. It has notified prosecutors in those cases where testimony from its experts "exceeds the limits of the science and cannot be supported by the FBI," one agency letter says.
At least three convictions - that of a Colorado man who served 12 years in prison for a double slaying, a Florida man who served 10 years after being convicted of killing his wife, and an Oregon man convicted of a triple slaying - have recently been overturned.
All three men are now free.
Comparative bullet lead analysis was based on the theory that lead bullets pick up trace elements such as copper, antimony, arsenic, bismuth and silver during manufacturing. When the soft metal is shaped into bullets and packaged, bullets in the same box would contain similar amounts of the trace elements, the theory went.
FBI lab technicians compared bullet fragments from a crime scene with bullets possessed by suspects. If the trace elements closely matched, prosecutors - backed by FBI testimony - would argue the suspects' guilt.
Defense attorneys say the analysis appeared to be a miracle of science: It required a small nuclear reactor, once housed at an FBI lab at the Hoover Building in Washington, D.C., and relied on the expertise of only a handful of qualified FBI agents.
FBI experts wowed jurors by explaining how gamma rays, energy released from bombarding a bullet with neutrons, could be measured to make a match.
"Sure, you have this whiz-bang, whipper-dipper machine that looks at all the elements of the universe, but it doesn't mean anything," said attorney Dave Wymore, a former director of the Colorado public defenders office who fought successfully to exclude such evidence in a triple-murder case and won an acquittal in 1999.
The FBI began the tests in the mid-1960s. It quit in 2005, after the National Research Council of the National Academy of Science concluded that while its methods of measuring trace elements were sound, its conclusions were flawed. Millions of other bullets could contain trace elements in identical quantities, the council said. That rendered the FBI's box-by-box conclusions meaningless.
FBI lab spokeswoman Ann Todd said the agency has reviewed about 2,000 cases and is waiting on information from prosecutors who may have used the bullet evidence at trial in the remaining 500 cases. In cases where an FBI expert testified, the agency is reviewing trial transcripts.
Cases where a defendant pleaded guilty aren't being reviewed.
In Colorado, attorneys for Tim Kennedy, convicted in 1997 for the double slaying in Colorado Springs, received an FBI letter in May 2008. The judge cited the discredited evidence, among other factors, in reversing Kennedy's conviction in April last year. He was freed in May.
Kennedy's attorneys, John Dicke and Kathleen Carlson, said testimony by FBI bullet expert Ernest Roger Peele added an air of credibility to what they called a weak case.
At first, investigators thought that the 1991 slayings of Steve Staskiewicz and his girlfriend, Jennifer Carpenter, were a contract "hit" to silence Carpenter, who was a victim and a witness in a sexual assault and kidnapping case.
When investigators couldn't tie Kennedy to two people suspected of ordering the hit, prosecutors developed the theory that Kennedy was desperate for money to buy drugs and allegedly killed his friends inside their trailer. A gun owned by Kennedy was used in the slayings. Kennedy testified that he had lent Staskiewicz three of his guns.
The defense conceded that the couple was killed by a gun and bullets belonging to Kennedy but argued he wasn't the killer. But jurors heard hours of testimony from the FBI's Peele, who said bullets recovered from the victims and bullets found in Kennedy's apartment were so identical that if they were "dropped on the floor, I couldn't tell which specimen they came from."
"The FBI guy was the determining factor," Dicke said. "It was the air of overwhelming credibility. Even the judge was impressed."
Peele, who conducted such tests for at least 15 years and testified in more than 150 cases, declined to comment when reached by The Associated Press.
El Paso County prosecutors are appealing the judge's ruling. Dan Zook, who prosecuted the case, declined to comment, though at a May hearing he argued that other evidence against Kennedy was overwhelming, including lying to the police about why his gun was the murder weapon.
Kennedy remains free on bail, while prosecutors appeal the reversal of his conviction.
In Florida, Jimmy Ates was released from prison in December 2008 after serving 10 years for the 1991 murder of his wife, Norma Jean Ates, in Baker. While other evidence helped win his release, the use of comparative bullet lead analysis "was really the lynchpin to get that case back into court," said Seth Miller of the Innocence Project of Florida.
A letter from the FBI was the determining factor in allowing Ates to reopen his case after he had exhausted his appeals. He is to be retried next year.
The FBI has agreed to notify attorneys with the Innocence Project, along with prosecutors, of any suspect cases. Attorneys at the Chicago law firm of Winston & Strawn are helping ensure the defendants learn of the discredited evidence.
"Somebody needs to be there to review those cases," said Miller, whose organization has received 24 letters from the FBI. "The reality is if you're an indigent person in prison, you may not even get that letter if it goes to the prosecutor."
His group isn't taking action in 12 cases where it determined prosecutors have overwhelming evidence of guilt.
"We're sort of taking one piece away and determining what it means," Miller said. "It's not always an easy calculus to figure out."
James Colaw, the division chief of the Clay County state attorney's office who will retry Ates, cited the risk of manipulations of the FBI's sweeping review.
"The analysis should be: Can we say that that evidence contributed to that conviction, and if that evidence was not used, would the outcome have been the same?" Colaw said.
Philip Scott Cannon of Salem, Ore., was freed most recently, on Dec. 18 of last year, after serving more than 10 years in prison for the 1998 fatal shootings of Celesta Graves, 24, and Suzan Osborne and Jason Kinser, both 26.
In that case, prosecutors hold out little hope that Cannon would be retried because exhibits from the trial could not be found.
Copyright 2013 Associated Press. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.
"Polk County (Oregon) totally screwed this thing up," Graves' sister, Jennifer Murdock of Salem, told the (Salem) Statesman Journal last month. "It's basically a bunch of crap that the families have to go through all this again."