The 2007 case of Kentucky versus Robert Yell resulted in some of the worst case law I have ever seen. It proves the old adage that bad cases make bad law. The KY Supremes blessed unconfirmed canine alerts, shiny alligators and irregular patterns among other myths.

It is thus my pleasure to congratulate Paul Bieber and Greg Gorbett for overcoming this bad law and persuading the Court to grant Mr. Yell a new trial.

Important findings:

Not revealing exculpatory evidence:

“Detective David West defended his conclusions that holes in the floor meant the use of an accelerant by stating his belief that it remains "possible" that an accelerant poured on a carpet with padding on a wood floor may result in a hole burned in the floor. He conducted his own experiment on the point which did not result in a hole but did result in charred wood. This would have been important had his testimony at trial been that arson was possible. Unfortunately, his testimony at trial was that the holes in the floor meant the fire was intentionally set and no reasonable doubt about this conclusion was admitted.”

Pattern analysis fails in fully involved rooms:

“The Court concludes that full room involvement did, in fact, occur in all of these areas. Under modern fire investigation methodology, full room involvement invalidates pattern analysis to determine points of origin.”

Unconfirmed canine alerts should not be admitted:

“The testimony presented at Mr. Yell's trial concerning the fire having had multiple points of origin and the canine "hits" having indicated the presence of an ignitable liquid, was unreliable and baseless and should not have been admitted into evidence.”

We have a problem:

“There may be something about the state of fire science and the nature of arson investigation that tends to permit errors.”

The case for bifurcation:

“Placing the duties of investigation and prosecution upon the same persons expccted to give an objective scientific analysis may be common practice in arson cases and this may be part of the problem. The more reliable expert testimony in criminal cases is presented by persons not directly involved in the general investigation and prosecution of: case. Ordinarily, police officers are not allowed to testify as to their personal conclusions on matters affecting the: defendant 's guilt They arc simply to present objective facts so that the jury can reach its own conclusions based upon those facts.”

“Juries are entitled to expect that the conclusions presented by witnesses qualified as "experts" arc entirely scientifically based and not merely the personal feelings of a prosecuting officer.”

“Scientific evidence should be evaluated separately from the other evidence in the case. This may help avoid the implication of “confirmation bias” and the inadvertent mixing of folklore, personal opinion and other non-scientific factors.”


The awful 2007 Supreme Court decision and the December 28, 2016 order granting a new trial are available here

[app.box.com]

John Lentini, CFI, D-ABC
Fire Investigation Consultant
Florida Keys
[www.firescientist.com]