if it’s impossible to win (in that case, just lose less; a semantic difference)
if “winning” is defined as something else than achieving what you truly value
That’s all of them, I think.
ETA: more in the context of this post, a good reason to lose at some subgoal is if winning at the subgoal can be done only at the cost of losing too much elsewhere.
Another is failure of knowledge. It’s possible simply not to know something you need to succeed, at the time you need it. No one can know everything they might possibly need to. It is not irrational, if you did not know that you would need to know beforehand.
I exclude bad luck from this list, since winning might as well be defined over counterfactual worlds. If you lose in your real world, you can still figure out how well you’d do in the counterfactuals.
I am wondering, what are the good reasons for a rationalist to lose?
bad luck
if it’s impossible to win (in that case, just lose less; a semantic difference)
if “winning” is defined as something else than achieving what you truly value
That’s all of them, I think.
ETA: more in the context of this post, a good reason to lose at some subgoal is if winning at the subgoal can be done only at the cost of losing too much elsewhere.
Another is failure of knowledge. It’s possible simply not to know something you need to succeed, at the time you need it. No one can know everything they might possibly need to. It is not irrational, if you did not know that you would need to know beforehand.
I exclude bad luck from this list, since winning might as well be defined over counterfactual worlds. If you lose in your real world, you can still figure out how well you’d do in the counterfactuals.
Well-chosen risks turning out badly?
I’ll give you odds of 2:1 against that this coin will come up heads...