Topics Probability

Question In a chapter on regression to the mean (Thinking Fast and Slow) Daniel Kahneman resorts to "luck" as an explanation for why one professional golfer shoots a lower score in a round than his/her rivals given that the talent pool is reasonably even. While a "lucky" (or unlucky) bounce can impact one's score, I find luck as a concept a poor explanation for performance. What is the philosophical status of luck, and are there different flavors of luck depending upon the philosophy? Is luck to chance as evidence is to data?

Accepted:
December 19, 2013

Comments

Games typically involve a blend of things that a player can control and things s/he can't. A golfer can work on her backswing; she can't do anything about the moment-by-moment shifts in the wind and the fine-grained condition of the greens. Things like the winds and the lay of the greens or the outcome of a dice-roll are what we might call externalities. It's not that they have no explanations and it's certainly not that they have no bearing on who wins and who loses. But the players don't deserve any blame or credit for how they turned out. In that sense, they're matters of luck. Depending on the game, skilled players may have ways of compensating for them to some extent, but they can produce advantages and disadvantages that are outside the players' control.

With that in mind, I don't take Kahneman's appeal to "luck" to be an explanation. An explanation would call for specifics about conditions and causes, and the mere appeal to luck doesn't provide any of those. I take the appeal to luck to be a way of saying that any detailed explanation will not be primarily in terms of well-chosen actions and displays of skill.

So yes: "luck" as such isn't an explanation. It's a way of alluding to a lack of an explanation. But note: it's not a way of saying that there's no explanation. It's plausible that if we had a sufficiently detailed account of exactly what happened in the golf game, we'd see exactly why Jones ended up with the lowest score. It's a way of saying that there's no explanation of a certain sort—in this case, in terms of things that fall under the control of the players.

As for the philosophical status of luck, that's a big topic. We tend to use the word "luck" when there's something at stake, and when what's at stake is outside our control—whether or not the lack of control is a matter of chance events. There's a large literature on what's called moral luck, and it might be a place to start looking. The Stanford Encyclopedia of Philosophy has a survey article on the topic HERE