Several weeks ago I wrote that CSI labs are not at all like they’re portrayed on TV. Revelations about shoddy practices are causing hundreds of convictions to be re-examined or overturned. Unfortunately, the same is true of arson investigations.

A string of exposés of wrongful arson convictions in Massachusetts, Illinois, Texas and elsewhere has revealed investigators using outmoded practices and relying more on intuition than science. In many cases, investigators have concluded that arson occurred when the real cause was probably accidental.

The grandfather of these cases was the murder-by-arson conviction of Cameron Todd Willingham in Texas – the case became the subject of exposés in The New Yorker magazine and most recently, PBS-TV’s Frontline. Despite mounting evidence that the fire was accidental, Texas courts and the U.S Supreme Court declined to review the murder conviction and then-governor Rick Perry refused to commute. On February 17, 2004 Willingham was executed by lethal injection.

The problem isn’t that arson investigators don’t have the science to differentiate accidental fires from those that are purposely-set. In fact, 1992 the National Fire Protection Association published a white paper on the latest scientific techniques; and over the next decade those guidelines became widely adopted. (Not surprisingly, in the last couple of decades the number of arson cases has dropped precipitously while the findings of accidental fires have jumped just as steeply.)

Unfortunately, even though today’s arson investigations may be scientifically based, those of previous decades often were not. And so an unknown number of people who may have been wrongfully convicted of arson — or even murder — sit in jail or, in some states, on death row.