So . . . one of the topics that I am most interested in is outcomes, and wrote a post about it a couple of months ago. If you can't be bothered to follow the link, the essence of the post is that we confuse best decisions (choosing the best option prior to learning the outcome) and right decision (the best decision in light of the eventual outcome). The example is use is the decision of whether to trade three lottery tickets for one; the best decision is to decline the trade, but if the single ticket is the eventual winner, the right decision would have been to do the trade.
One comment I got about this post was that it sounded like "hindsight is 20/20." I would agree insofar as the right decision goes - we determine whether we made the right decision through hindsight. The best decision remains unchanged - it is based on probabilities, not outcomes. And the fact remains that unless we are Marty McFly or Doc Brown, we never have access to the eventual outcomes. So ultimately what we really want are best decisions, even if that means accepting that some of them won't be right decisions.
This was illustrated when I appeared on Jeopardy! a few years back. Prior to taping the show, the contestant wranglers give an hour-long spiel about all aspects of the show - rules, gameplay, and some rudimentary strategy. One suggestion that was give was to avoid jumping around the gameboard looking for daily doubles, and instead pick the clues from lowest dollar amount to the highest. Now if you could somehow suss out where they were your probability of winning would be much higher. But looking for one or two clues out of 24 is a fool's errand, because the odds of actually finding it are low. Should you jump around the board and find it, you would have made the right decision, but the strategy itself would never the best decision (interestingly, the contestant wranglers said that only one contestant in the show's history had a talent for knowing where they were, and he was very successful; a more likely explanation is that he was lucky and that luck led to his success).
But the hindsight-is-20/20 argument also assumes a different point, that we would in fact change our decision given the opportunity. There's an old joke: two guys are watching a boxing match at a bar, and they make a bet on the outcome. One of the boxers makes a big mistake and gets knocked out. The winner collects his money, but then has a pang of guilt and admits that the match was actually re-broadcast, and he had seen it already, knew the outcome, and bet on the one he already knew won. The loser then admits to having seen the match before as well. When asked why he bet on the loser, he says that he didn't think he would make the same mistake twice.
If we had access to time travel (oh, if only we did . . .), would we change our behaviour? One of Kurt Vonnegut's novels, Timequake, centers around an incident by which everyone re-lives the previous ten years of their life. The catch is that we are passive observers reliving everything, and are unable to change anything. We have to watch ourselves make the same mistakes over again. Painful, huh? But if given free will, would we change anything, or hope that this time it works out?
And if you really believe in randomness, as I do, then you should continue to make the best decisions (rather than what you 'know' to be the right ones) in such a circumstance, because the outcome is still the result of a random mechanism. Because you change your action, the result may very well change. Look at Bill Murray in Groundhog Day - he changes a few things, and the fates of everyone in the town are also changed. Marty McFly meeting his parents negates his existence (but, luckily, he doesn't disappear immediately, and has time to entertain us with his attempts to restore his being). And swapping three lottery tickets for one may very well affect the outcome.
Okay, I'm going to stop now, because my brain is tired. Hopefully this kind of makes sense.
No comments:
Post a Comment