Safety Margin: "15-to-1" and all that...

Taser claims that the safety margin for 'the taser' is "15-to-1". And that this means that the risk is exceedingly, extremely, amazingly low.

First of all, anyone that uses the all-in-one phrase 'the taser' when discussing safety issues with the M26 and X26 tasers should be forced to sit in the corner for five minutes. I've seen plenty of technical information that indicates that the M26 and X26 tasers would probably have different safety margins. The 2003-era X26 taser's waveform has characteristics (low frequency, high duty cycle) that are the exact opposite of what Taser claimed was safe on the older 1999-era M26 (high frequency, low duty cycle). Very strange that.


So where does the purported "15-to-1" safety margin come from?

The only numbers I recall seeing that are consistently in a 15-to-1 ratio are Taser's claimed "average" current of around 2 mA (1.9 mA or 2.1 mA) for the M26 and X26 tasers as compared to the safety limit of IEC 479-1 at 30 mA. 30 mA to 2 mA equals 15-to-1.

So, let's ignore the fact that the M26 taser is also rated at 162 mA RMS and that the X26 taser is also rated at 151 mA RMS. We'll ignore the RMS question for now. And not bother asking why the RMS value was deleted from later spec sheets.

We'll also ignore those pesky skeptics that have measured output currents 84.5 times higher than the "average" current specified by Taser.

And we'll ignore for the time being those skeptics that don't accept the anticipated circuit resistance presented by the victim as assumed by Taser.

We'll also ignore the fact that the 'applicable safety standard' probably isn't applicable because it was designed to protect lives in the event of an accidental electrocution, and there may be other issues with using it to set limits for intentional application of electric current. Especially when used against people that (for example) failed to pay their transit fare.

Even accepting all this, what about the 30 mA?

The 30 mA limit covers 95% of the (normal) population. The reaction of the lowest 5% of the population is undefined. It may be reasonable to assume a Bell Curve response tailing off towards lower and lower currents. So, by the normal application of high school statistics, there will be a significant fraction (5%) of the normal population where the purported 15-to-1 safety margin does not exist.

This issue exists even after giving Taser the benefit of the doubt on all the issues mentioned above and using their numbers.


Also, we are constantly hearing that perhaps the victim died because they were on drugs, or they were drunk. There's an implication (not explicitly stated) that perhaps these people brought their fate upon themselves. That somehow maybe the tasering that triggered some abnormal reaction isn't really to blame...

The argument gets a bit fuzzy at this point because Taser knows exactly where it leads:

If drunks and drug addicts make up much of the population that is likely to be tasered, then the taser should have been designed for this population. Some might even claim that it would be criminal negligence to design a device obviously going to be used on drunks and drug addicts and not take into account any differences in that population's sensitivity if such differences exist.

At this point, you'll see a man wearing a bowtie riding a bicycle backwards at a very high rate of knots. Taser will have nothing to do with the argument that drunks and drug addicts are more susceptible. But their fanboys keep implying it, which is fine for Taser.

Perhaps that's why it is such a "good idea" to taser fare-cheats on the transit system. At least you'll increase the number of people being tasered that are not drunk nor on drugs. Yeah, good idea - taser some normal citizens going about their business to wash the statistics towards a more-normal population mixture. Get some truant children in there too. Increase the denominator to flatten out the overall statistics a bit. Maybe it would be a good idea to taser some healthy police officers during their taser training for a further dilution of the statistics away from the real world applications.


"15-to-1" safety margin does not constitute a proper safety analysis.

A proper safety analysis can be distinguished by the final result which is the numerical value of the estimated death rate. And the calculated result should approximately match the actual field results.


NASA's Space Shuttle program went through a similar circumstance in the 1980s. The top NASA management thought that the risk of a Space Shuttle accident on any mission was about 10,000-to-1. But some NASA engineers thought it was more like 100-to-1. But NASA management went with the preposterous 10,000-to-1 value. So they started to put teachers into the space program because they thought it was so safe. After Challenger exploded (killing a teacher), they had to go back and not only fix the design, but also admit that the risk was never anywhere near 10,000-to-1. Even now, they all know the risk is more like 100-to-1. Everyone goes in with their eyes open.

Sounds like a familiar story-in-progress, doesn't it?