Cameron Todd Willingham was just 23 when he was convicted of setting a fire that killed his three little girls - 2-year-old Amber and 1-year-old twins, Kameron and Karmon.
Willingham told police he tried to save his girls, who all died in the 1991 fire at the family's Corsicana, Texas home, but fire investigators say clues at the scene told them he'd actually set the fire. He was convicted of arson homicide.
But today, some leading fire investigators around the country say the old method of determining whether or not a fire was arson is outdated and unreliable. Pour patterns on the floor are no longer considered proof an accelerant was used, they say. There is a newly understood phenomenon called "flashover" that can cause such patterns without an accelerant ever being introduced.
These investigators say for years arson determinations have been based more on folklore, than fact. Hunches handed down for generations.
So is it possible fire investigators in the Willingham case, who believe they found pour patterns on the floor and three points of origin for the fire, got it wrong in the Willingham case? That would mean on February 17, 2004, the state of Texas may have executed an innocent man.
Fire investigator and forensic scientist John Lentini studied the Willingham case and determined it was not arson that killed those little girls. He calls the original investigation B.S. - Bad Science.
Lentini told me, ''There's maybe 75,000 suspicious fires a year. That's 75,000 chances to get it wrong.'' We met Lentini at a Maryland lab so he could show us why he believes the relatively new arson science debunks the myths he says have been handed down for years. I was amazed at what I saw. I don't know if you've ever seen the show "Myth Busters" on the Discovery Channel, but I felt like I was in the middle of a "Myth Busters" episode.
One so-called "indicator" for those who originally investigated the Willingham case is something called "crazed glass." For years, crazed glass -- which has webbing or tiny cracks inside it -- was believed to be a sure sign of a very hot fire, one that likely involved an accelerant. But Lentini and the others showed us how "crazed glass" is actually caused by rapid cooling, not rapid heating. We sprayed a piece of hot glass with water, the same way a fireman would spray a window with his hose, and that is when the glass began to crack inside. It didn't crack at all when we heated it up.
The man who prosecuted Willingham calls the new findings "silly" and says he has no doubt Willingham was guilty, based on fire evidence and Willingham's history of drinking and domestic abuse. The original fire investigators also stand by their findings. The Texas Governor's office would not comment for this report.
These new forensics are now used as the gold standard of arson investigation around the world. It may have come too late in the case of Cameron Todd Willingham (the governor of Texas reviewed the new findings just 15 minutes before Willingham's execution and chose to go ahead with it) but they could save hundreds of others behind bars for arson who claim they're innocent. Problem is, the International Association of Arson Investigators, doesn't see a need to reopen or revisit all of the arson convictions on the books.
If that was your loved one behind bars, wouldn't you want the new ways of looking at evidence heard?