I recently bought a new home and one of the items that came up in the inspection report was the risk that

The power switch for the garbage disposal in the sink could be accidentally turned on by a person standing at the sink while their hand was in the disposal.

That is to say, the switch is right next to the sink.

I thought about this, and realized that I preferred that situation in the “risky” state.

While the likelihood that I will flip that switch on if I have my hand in the garbage disposal for some reason are exceedingly low, I’m willing to bet that they’re still lower than the likelihood that someone would accidentally flip the wrong switch and turn on the disposal during one of the rare occasions when I have my hand in it.

If I were writing the report, the risk I’d have identified (had it existed) would have been something like:

The power switch for the garbage disposal is located such that a person at the sink cannot reasonably ensure that another person won’t accidentally turn it on while the person at the sink has their hand in the disposal.

Who’s right? We’re both using the same data (the location of the switch in relation to the sink) to evaluate the same incident (the disposal being accidentally turned on) while in a vulnerable state (hand stuck in the disposal).

The threat is different (I turn it on myself versus someone else turns it on), but I would argue that the risk is truly lower with the switch by me for three reasons:
1) I am highly motivated to not accidentally hit that switch while fishing in the disposal;
3) I, by definition, only have 50% as many hands which could potentially turn the switch on if one is in the disposal; and
3) I’m effectively protecting the switch from someone deliberately turning it on in order to to mangle my hand and thus, say, put an end to my blathering on about risk management in a weblog.

This was a trivial case involving clearly defined impacts, assets, and very low likelihoods, and yet it still was able to produce diametrically opposing positions on what the appropriate risk stance should be. Is it any wonder that risk management of non-trivial problems is hard?

Great post. But I was expecting research from you! We need to find the frequency with which “garbage disposal/hand” accidents happen per year.

My guess is lack of frequency (billions of disposal uses per year, less than one accident per year) creates a low risk situation for both are situations, even though the impact is really, really high :)

You, by definition have 50% fewer hands the first screw-up. After that, it’s totally safe.