Mission Statement - De-Spinning the Pro-Taser Propaganda

Yeah right, 'Excited Delirium' my ass...


The primary purpose of this blog is to provide an outlet for my observations and analysis about tasers, taser "associated" deaths, and the behaviour exhibited by the management, employees and minions of Taser International. In general, everything is linked back to external sources, often via previous posts on the same topic, so that readers can fact-check to their heart's content. This blog was started in late-2007 when Canadians were enraged by the taser death of Robert Dziekanski and four others in a short three month period. The cocky attitude exhibited by the Taser International spokespuppet, and his preposterous proposal that Mr. Dziekanski coincidentally died of "excited delirium" at the time of his taser-death, led me to choose the blog name I did and provides my motivation. I have zero financial ties to this issue.

Thursday, April 30, 2009

Presentation by Dorin Panescu

Estimation of TASER Current Flow and Effects on Human Body

Available here [LINK] (833KB .pdf)

This report is an example of why many well-run high-tech companies avoid hiring too many PhDs.

The modeling of the human body presented in this report is primitive to the extreme. The ratio of the complexity of a real human body to this model must be at least 100,000-to-1. Medical school would last an hour if the human body was as trivially simple as this crude model.

At best, this sort of crude model might be good for some rough order of magnitude approximations. Providing two-digit results implies approximately 1% accuracy ('significant figures') which is completely laughable.

And where the hell is the tolerance analysis? You know? ...where you use actual population Bell Curves to calculate the actual population risk ratios? Remember your first year statistics? Monte Carlo analysis? Oh Hellllloooo?

Even given all of the above, on Slide 23, it appears to indicate that the safety margin for 'capture' can be as low as 1.7 times. I assume that 'capture' means roughly the same thing as 'affecting the heart'.

And this is with a very crude model that is really only good for rough estimating. When I see a safety margin for affecting the heart of just 1.7x, given the utter crudeness of the modeling effort, then this causes alarms bells to go off. I'll bet that the tolerance of the accuracy of this crude model firmly overlaps with the real-world population Bell Curve. Quote me: "Low end single digits."

In my opinion, this counts as a huge I Told You So.

PS: Slide 23: "**TASER CEW M26 margins are wider..." I Told You So on that as well.

No comments: