Since 1991 approximately 40 men and women across the United States have been exonerated from wrongful arson convictions. In over two-thirds of those cases, the unreliable, unscientific, or clearly exaggerated testimony of expert fire investigators played a lead role in the original convictions.
This website contains data and materials relevant to each of the common, reoccurring unreliable forensic techniques at work in each case. Additional information for each factor and the cases they have influenced is being added to the website as documents become available.
Fire investigator and accelerant-detecting canine guidelines caution against relying on the subjective behavior of an accelerant-detecting canine, especially when laboratory samples of the item in question are returned as negative or inconclusive for ignitable liquid. Yet many courts have accepted the testimony of accelerant detecting canine handlers even when fire debris samples taken from the fire scene are negative.
Negative corpus is the process of determining what caused a fire by eliminating each possible cause, one by one, until only one possible cause remains. The one remaining possibility is then determined to be the cause of the fire, regardless of a complete lack of physical evidence to support that conclusion.
Although the fundamental methodology of fire-pattern and fire-dynamics analysis provided in NFPA 921 and other texts appears to provide some guidance to assist the investigator in determining a fire’s area of origin, the reliability and validity of this method has not been rigorously measured. Of special concern is the accuracy of area of origin determination in a ventilation controlled compartment fire.
Complex fire conditions can create fire patterns and burn damage that can be easily interpreted by the fire investigator as “multiple areas of origin”. For instance, the same full room involvement conditions that can lead to an incorrect single-area-of-origin determination can be easily misinterpreted as multiple areas of origin.
Until the first edition of NFPA 921: Guide for Fire and Explosion Investigation was published in 1992, there were a number of widely accepted myths and misconceptions regarding how a fire investigator could examine fire patterns and burn damage at a fire scene and deduce the presence of an ignitable liquid.
Elimination of an electrical appliance or electrical conductor as a heat or ignition source by visual examination is a common and troubling feature of many fire-scene examinations. A standardized methodology for conducting such an examination does not exist and NFPA 921 provides little guidance in how to go about identifying electrical appliances that may have caused a fire from a similarly damaged appliance that did not. The error rate of this type of critical forensic examination has never been measured and is completely unknown.