MODESTO — Stanislaus County prosecutors were so persuasive, their evidence so overwhelming and their witnesses so convincing that it took jurors only a couple of hours to find George Souliotes guilty of triple murder and arson.
That was in May 2000.
Now, after 13 years of appeals, federal judges have agreed that much of the scientific evidence used to convict the Modesto man was inaccurate and his original attorneys did a lousy job defending him.
Those judges have determined he is "actually innocent."
Even the state attorney general's office now concedes there's no proof the fire was an arson and there's no chemical evidence linking Souliotes to the blaze.
How could prosecutors, witnesses and jurors have gotten things so wrong? Or is legal wrangling simply causing doubt where none should exist?
Souliotes, 72, has been locked behind bars since the day of that 1997 Modesto rental house fire, which killed his tenant Michelle Jones and her two young children.
Federal judges insist that Souliotes be set free by May 12, unless prosecutors agree to promptly retry him without using the faulty arson and chemical evidence they previously had relied on. Stanislaus County District Attorney Birgit Fladager said she will decide within a week what to do.
Souliotes, who always has insisted he didn't set the fire, waits in Salinas Valley State Prison.
While it's rare for murder convictions to be overturned, those involving suspected arson have come under increased scrutiny because of advances in fire science.
Burn patterns and residues
Much of what once was considered decisive evidence of arson has been scientifically invalidated, said David A. Moran, a professor at the University of Michigan Law School.
Back in the late 1980s, for example, Moran said arson was blamed for about 25 percent of house fires. But now only about 6 percent of house fires are considered arson.
Moran said that's primarily because fire investigators used to believe certain burn patterns and residues were proof that an accelerant like a flammable liquid had been used. He said it's now known those things can be caused by a naturally occurring "flashover phenomenon."
"Unless you have a very good chemist trained in fire science at the scene, people mistake burned petroleum-based products for accelerants," Moran said. "That's a common problem."
That may have been what happened in the Souliotes fire.
Two Modesto fire captains believed they had found evidence an arsonist had used a flammable liquid to start the blaze in the Ronald Avenue home.
The prosecution's primary fire investigator testified at Souliotes' trial that it "was a crude arson job and it's as obvious as an elephant. Liquids were poured throughout this house and ignited."
Last year, the attorney general's office, which represents the prosecution in appellate cases, acknowledged that wasn't true.
In a statement of "undisputed facts," the state officials conceded Modesto fire investigators had relied on factors that "are now known not to be indicators of arson using a flammable liquid."
"The floor damage and burn patterns located inside the house and garage that the original investigators attributed to arson using a flammable liquid appear to be the result of flashover burning and fall down from the roof collapsing, and are not indicators of arson using a flammable liquid," the state acknowledged.
The state conceded there's no evidence as to whether "the fire was accidental or the result of arson."
Modern science also has disproved the prosecution's contention that chemical residue found in the burned-out remains of the home matched residue found on Souliotes' shoes. It doesn't.
But it took until 2005 for a new testing method to be used. That test proved the two chemicals were different and did not originate from a common source. That finding significantly guts the prosecution's case.
Navigating a legal maze
Since the chemical evidence essentially was scientifically debunked eight years ago, why is Souliotes still behind bars?
Despite knowing the prosecution's key fire science evidence was wrong, Souliotes' volunteer attorneys were required to maneuver through a legal maze before they could get a judge to consider his case. At one point, they missed a deadline by a couple days, which caused years of delays.
Souliotes' lead attorney, Jimmy S. McBirney, said it wasn't enough for his client to prove he actually is innocent. He also had to show he didn't get a fair trial.
Federal judges now agree Souliotes was poorly represented at his second murder trial (the first ended in a hung jury favoring conviction 11-1). His two attorneys didn't call witnesses to defend him or to refute the prosecution's evidence.
"No reasonable explanation can be imagined" for the defense not calling witnesses, U.S. Magistrate Judge Michael J. Seng wrote in his ruling last month. "Without the testimony of the witnesses, the jury was not provided critical information that could create reasonable doubt."
Seng noted how prosecutors presented 50 witnesses, but he called the district attorney's case vulnerable: "Its main eyewitness suffered significant credibility concerns. The fire investigators hastily came to conclusions based on inaccurate and incomplete review of the physical evidence."
Had Souliotes received an adequate defense, Seng said he may not have been convicted. Now that the key evidence about arson and his shoes have been debunked, it will be much harder to prove Souliotes killed the Jones family.
So why is it taking so long for Souliotes to be set free?
"Some prosecutors are reluctant to admit mistakes. They cling to the idea that wrongful convictions don't happen, at least not in their jurisdiction," said Moran, whose university helps maintain The National Registry of Exonerations.
That registry lists more than 1,100 people who had been convicted of serious felonies but have been cleared of all charges since 1989.
Currently, there are no Stanislaus County exonerations listed, but Souliotes' name could be added next month.
Bee staff writer J.N. Sbranti can be reached at firstname.lastname@example.org or (209) 578-2196.