How to resolve the "safety" "question"

There are two possible approaches for sorting out this so-called "question" once and for all.

1) Compare the M26 taser death rate per full deployment (darts hitting real world subjects, not volunteers) against the X26 taser death rate per deployment (same real world data, no FAKE demos). If the ratio of these two rates is not very close to unity, then that would be clear evidence of an effect from the device itself.

Prediction: X26 will exceed M26. Anyone got complete data?


2) Review the death rate per unit time just BEFORE the taser would have been fired, and AFTER it has actually been fired. Google the taser's "Curious Temporal Asymmetry" for details. Basically we're seeking many hundreds, or even thousands, of reports where the taser was drawn, but then was put away because the victim died of "excited delirium" before the officer had a chance to fire, or during the period of attempted negotiation.

Prediction: 'After' will exceed 'before' per unit time, by a significant margin.


This debate is no longer a philosophical game. It is no longer a semantic game decorated with Latin phrases. These are very simple and straightforward methods of dividing the existing data sets, where ever they may be squirreled away, into the 'effect' and the experimental controls.

Both of these proposals have been repeatedly discussed on this blog.