From: Hachiroku on
On Tue, 23 Feb 2010 08:16:41 -0500, C. E. White wrote:

> I don't think his math works (a 1 in 8000 chance per year does not
> necessarily equate to a 1 in 800 chance in a decade). I am not sure
> about the other math either. Assuming that every death in a Toyota as a
> result of a UA incident is reported as such is a stretch. Of course it
> is also true they might be over reported. But if the driver died, how
> would we know it was a UA incident?

Something that no one seems to be doing: get the data from the ECU. The
ECU will store the conditions the car was operating under prior to the
accident and can be retrieved, showing speed, throttle position, attempts
at braking, etc.

So far the people claiming accidents caused by UA have not mentioned the
ECU. I'm willing to bet once the data is pulled from the units the amount
of claims will drop substantially.