Nitpick: When you mentioned the 2-4-6 experiment, I didn't know what it was, so I clicked on the link and read about it. That was unnecessary, because you immediately explained it ... but, somehow, the wording of the narrative didn't signal that the explanation was coming.

Could be just me.

I guess it's an empirical question.

Yes.

A death creates two kidneys. Are there usually two people on a waiting list who need the kidneys and would otherwise die?

Humans aren't lego. Yes, we can transplant but they don't always work and they don't always last indefinitely. We also don't just use them to flip a nice integer 'life saved' up by one. It's ok if the spare organ just increases someone's chances. Or extends a life for a while. Or drastically improves the quality of life for someone who was scraping by with other measures.

If I recall correctly kidneys are actually the easiest organ to transplant - the least likely to cause rejection. With the right donors it gets up into the 90s(%). But translating that into lives saved or 'years added to life' is a little tricky. Especially when we the patients also happen to require transfusions of donor blood throughout the process. We like to say the blood transfusions are 'saving a life'. There are only so many times you can count a life as 'saved' in a given period of time.

That seems reasonable to me: two kidneys save two lives, one liver saves a third life, and so on.

House MD doesn't seem to get that sort of conversion rate from organs to lives saved. Am I generalising from fictional evidence or is your life saving equation absurdly optimistic. Ok, I admit, both.

What he should have done was contingently committed to selling his organs on the black market before committing suicide. Then, there would have been a net benefit to his death, instead of it being zero-sum, and his actions would have been admirable.

Does not follow - the breakup value of your organs is not necessarily greater than your organs working together. Just because someone gets paid doesn't mean that game is positive-sum.

Second, there's 50% less chance of error, since you're connecting only two clamps and not four.

But there are still only two ways to connect the four clamps, since cable color doesn't matter when they're acting purely as cables.

Jump-start a stalled car.

I "get around" this by not carrying around jumper cables, so nobody asks me, which is of course absurd.

Voted up, because, this post I understand.

Nitpick: shouldn't the answer to the disease question be 1/50.95 (instead of 1/50)? One person has the disease, and 49.95 (5% of 999) are false positives. So there are 50.95 total positives.

I'm currently researching the rationality of criminal trials, though my focus is on evidence law. The current adversarial system does have some advantages:

• An adversary system creates incentives to bring information before the court. Parties want to win. In particular, the defendant has a direct interest in avoiding punishment. Defendants can help themselves by pointing out flaws in the prosecution case. In general, an adversary system rewards parties who put arguments and evidence before the court.

• However, this can include bad arguments and emotive but irrelevant evidence. This makes procedural decisions on what will or won't be considered at trial very important. Flawed procedures might allow a rhetorical free-for-all or exclude informative advocacy.

• The prosecution bears the burden of proving guilt beyond reasonable doubt. Discharging this burden means finding witnesses, taking photos of crime scenes, engaging forensic scientists, and so on. The prosecution normally has better investigative tools to do this work. Allocating the burden of proof in this way therefore helps to bring relevant evidence before the court.

Here are some things we could change:

• Currently, evidence can be excluded for technical reasons. For example, evidence gathered in breach of a defendant's rights is normally inadmissible at trial. So, there may be strong evidence of guilt, but the defendant may nonetheless be acquitted.

This is probably about recognizing (legal) rights, and eliminating reasons for prosecutors to breach rights. Much of the Anglo-American criminal law is about being fair to defendants - only convicting when the proper standard of proof is met by the proper means.

An alternative approach would be to allow such evidence at trial, and severely punish investigators who breach rights. This would avoid some technical acquittals where the evidence supports conviction. However, a reluctance to punish high-status investigators means this approach could just result in more breaches of rights.

• Forensic investigations could be conducted by neutral groups. There could be certified investigators. Parties could pass questions and suggested lines of inquiry to these investigators. These investigators could be responsible for leading evidence - presenting a narrative of the relevant events via witnesses and physical evidence. Parties directly involved might then be able to cross-examine the investigators and their witnesses, to test the suggested narrative for flaws.

This would gain some impartiality in forensic investigation. Currently, forensic investigations are normally linked with the prosecution case, and are informed by the prosecution's theory of what took place. With a suspect in mind, confirmation bias can skew the conduct of such investigations. Alternative explanations may be too readily rejected.

Imposing a third party between collection of evidence and arguments at trial could help to mitigate this risk. It could even out the forensic resources available to parties. However, it might decrease the speed and efficiency of investigations.

"Neurotypical" - in context, not being significantly autistic.