I am not involved in the Sykes case, but i am familiar with the prosecution expert's report. Mr. Lentini was quite specific about the errors in that report and he was correct.

The Sykes case is a poster child for the new wave of junk science in fire investigation. It began with a botched investigation. The prosecution then brought in a "high-powered" expert to rehabilitate the case by applying the voodoo of fire dynamics. The new expert relied on methods published by the NRC which were incorrectly summarized from a technical paper written years earlier. That paper described three empirical equations designed to approximate the heat release rate required to bring a room to flashover. The three equations give wildly different answers using the parameters of real (as opposed to test) rooms. The expert averaged the three wildly different answers to arrive at his result. This procedure would be unacceptable even if he had used the correctly selected three equations -- which he did not.

The problem is compounded by the fact that the NRC got it wrong when they selected the equations from the original paper. They one equation which was designed only for a compartment with very specific parameters -- unlike those which would be found in a typical residence.

Fire dynamics is a very useful research tool, but it is not yet ready for general plug-and-play application in fire investigation.

Once an investigation is thoroughly botched as it was in the Sykes case, trying to rehabilitate it with an overlayer of technobabble is like beating a dead horse.

There were also errors in the computer (CFAST) work, but even if the program had been working correctly, it would merely have generated a nice curve passing through a shotgun pattern of actual laboratory test results.

None of this has any bearing on Ms. Sykes actual innocence or guilt. All that can be derived from the existing data is the fact that one cannot derive the origin and cause of the fire from those data.