In your example, the literal truth of Alvy’s claim strongly depends on all the conjuncts being true, but the insightfulness or truth-like-ness etc. of the claim does not. If 90% of his guesses are right, that’s still pretty good. I think this pattern is more general: unless a story falls apart when you take any of its components away (e.g. because it’s a math proof), conjunction bias is mitigated by the fact that a conjunction can be what humans call “true” without being literally true.
Plus the more specific the claim, the more points he gets for insightfulness. Guessing that someone is left-wing doesn’t get many points at all, but if Allison’s father really does have Ben Shawn drawings then Alvy gets a lot.
In your example, the literal truth of Alvy’s claim strongly depends on all the conjuncts being true, but the insightfulness or truth-like-ness etc. of the claim does not. If 90% of his guesses are right, that’s still pretty good. I think this pattern is more general: unless a story falls apart when you take any of its components away (e.g. because it’s a math proof), conjunction bias is mitigated by the fact that a conjunction can be what humans call “true” without being literally true.
Plus the more specific the claim, the more points he gets for insightfulness. Guessing that someone is left-wing doesn’t get many points at all, but if Allison’s father really does have Ben Shawn drawings then Alvy gets a lot.
Nice catch. Worth a short top-level post if you have time to kill.