A terroristic problem

[ QUOTE ]
The poll is flawed because it presents your only option as shooting the guy in the head. If you have a gun, and have a shot at him, and you have the skill to shoot him in the head at will, you could (a) shoot him in the leg to keep him from making it to his target, and/or (b) shoot him in the arm/hand to keep him from detonating. Thus if there's any doubt, there's no reason to take the person's life.

[/ QUOTE ]

If you shoot him in the legs, he can still detonate the bomb. The only place to shoot him to guarantee that he won't be able to detonate the bomb is in the head.

That said, I do not support shooting people unless the police are 100% sure that they are a threat.

[MMMMMM]

[ QUOTE ]
Could you please point out the point the OP illustrates with the question?

[/ QUOTE ]

That mathematical principles have some value and should not be completely ignored in such scenarios.

[ptmusic]

[ QUOTE ]
Do you guys not realize he RAN AWAY FROM THE POLICE in a subway station wearing heavy clothing in the summer.

[/ QUOTE ]

There are countless explanations for this behavior, only one of which is that he is carrying a bomb that he intends to use right then and there.

-ptmusic

[ptmusic]

If the purpose of this is to make a life or death decision based on an "easy calculation" (how absurd is that) to save the most lives, one needs to consider the INFINITE increase from 0 deaths to 1 death (you shoot and kill him). The increase from 1 death to 5 or more is not NEARLY as big an increase.

Therefore, you better be pretty damn certain that you will be saving lives before you make that infinite increase from 0 to 1.

-ptmusic

[MMMMMM]

[ QUOTE ]
If the purpose of this is to make a life or death decision based on an "easy calculation" (how absurd is that) to save the most lives, one needs to consider the INFINITE increase from 0 deaths to 1 death (you shoot and kill him). The increase from 1 death to 5 or more is not NEARLY as big an increase.

Therefore, you better be pretty damn certain that you will be saving lives before you make that infinite increase from 0 to 1.

[/ QUOTE ]

But that is not an infinite increase, numerically speaking; it is an increase of exactly 1. And a jump in deaths from 1 to 6 would be a numerical increase of exactly 5.

An increase from 1 to 6, in comparison with an increase from 0 to 1, is five times as large an increase in actual numbers. And I think we're comparing saving actual numbers of lives, not abstract percentages, aren't we ;-) ?

MMMMMM,

I don't particularly care if having a policy like this is proven to "save lives". Giving the police the power to shoot a suspected suicide bomber is quite dangerous, and would result in a lot of innocent deaths.

[MMMMMM]

[ QUOTE ]
I don't particularly care if having a policy like this is proven to "save lives". Giving the police the power to shoot a suspected suicide bomber is quite dangerous, and would result in a lot of innocent deaths.


[/ QUOTE ]

I agree that there are concerns and considerations which go beyond the raw numbers.

[ACPlayer]

Do you tbink they have values other than as a systematic way of documenting assumptions and then testing the assumptions? Which is exactly what I tried to do with the question, in pointing out its inherent flaw.

The field here came up with essentially two opposite answers to the question. So the answer is not obvious but is clearly subjective.

It seems to me that the science of human nature has not evolved to the point where we can come up with a mathematical representation or come up with a unique correct answer. Perhaps it will in the future, I hope not.

[MMMMMM]

I think the question has an objective answer, given only the assumptions or premises as posed in the question (one of the premises is to assume that the goal is to save lives).

Of course, the assumptions given are quite limited, and the real world scenario is more complicated than that.

[SumZero]

[ QUOTE ]
[ QUOTE ]
Could you please point out the point the OP illustrates with the question?

[/ QUOTE ]
That mathematical principles have some value and should not be completely ignored in such scenarios.

[/ QUOTE ]

Yeah, but the setup for the math in the initial question is flawed. I'd give the benefit of the doubt here based on what I know about the London shooting to the police.

The reason the math is flawed is it is really a conditional probability question that states GIVEN there is a guy not responding to police acting some what suspiciously (for whatever definition of some what suspiciously you have) THEN what is the probability he is a terrorist that is about to explode a bomb?

If the conditional part happens once a century then clearly you don't need to be so sure. If it happens 1000 times a day then you need to be very sure before you shoot to kill. The guy in question could clearly not speak english, be deaf, be listening to earbuds, be lost in his own world not paying attention, etc. So I don't think the GIVEN part is that unlikely (and the poll question doesn't give this necessary piece of information), and as a result ~20% is too low [imagine it "the circumstance" happens 1000 times with innocent people for each 5 times it happens with guilty people and that you have a 20% rate of correctly identifying people. Then the probability that this is really a terrorist given you are 20% sure it is a terrorist is only 1/201 or less than half a percent].