I haven’t asked my friends to specify the probability they’ll make a given decision (typical people find this annoying for some reason), but I’ve routinely heard them express high levels of confidence in a choice, only to have made a totally different decision the next day.
From what I’ve noticed, if someone has two options and they can identify the one that they are more likely to choose, then they usually end up choosing that one even if they think they are only slightly more likely to (like Eliezer points out in the sequences), but that if they express high confidence in their choice, this does not translate into very much higher probability that they will actually make that choice than if they only expressed slight confidence. (This is assuming that they seriously considered both options. If you ask a mentally healthy person how likely he is to jump off a bridge tomorrow, he will tell you that he won’t, and he will be right.)
From what I’ve noticed, if someone has two options and they can identify the one that they are more likely to choose, then they usually end up choosing that one even if they think they are only slightly more likely to (like Eliezer points out in the sequences), but that if they express high confidence in their choice, this does not translate into very much higher probability that they will actually make that choice than if they only expressed slight confidence. (This is assuming that they seriously considered both options. If you ask a mentally healthy person how likely he is to jump off a bridge tomorrow, he will tell you that he won’t, and he will be right.)