Being a gamer and an athlete, I’ve long been interested in luck. Put roughly, a person can be said to be lucky when chance favors them in a way that is statically significant and, very importantly, this cannot be attributed to other factors. For example, if someone consistently rolled better than chance dictated in games involving dice and this was not due to weighted dice or a special skill in rolling better, then they could be regarded as lucky. Bad luck works in reverse: chance goes against the person.
Believe in luck is generally considered a superstition. Naturally, people do believe in it and often think that they have evidence for it. However, luck can often be explained away in two ways.
First, people who are seen as lucky or unlucky are actually creating their own luck. To be specific, a person who has a reputation for being consistently being lucky is most likely taking steps that increase his chances of success. For example, I have a reputation for being lucky in war games that use dice to resolve combat (like in Risk, Axis & Allies and BattleTech). However, I am not really lucky-I am careful to set things up so that the situation will be favorable to me. I will, for example, make sure that I get as many attacks as possible. It is then not a matter of luck that I hit often, it is a matter of normal probability: if your units fire more times and are in favorable circumstances, then they will hit more often.
Second, people tend to remember vivid cases much more than normal cases. For example, the time V’Rell put an arrow into the Wyrm Lord’s left eye and slew him (because I rolled really well in a D&D game) is well remembered. The other thirty arrows fired in the battle are forgotten. As another example, the time an athlete slips while trying to make a free throw is remembered, but the times she did not fail to stand out in our memory. So, what people think of as luck is merely the result of selective memory.
So, what would real luck be? Real luck would involve having a left over statistical anomaly after all other factors are eliminated-such as skill and the vagaries of memory. This anomaly would have to be statistically significant so that the results could not be attributed to pure chance. After all, a person could roll lower than average in one game while still staying nicely within what is expected in terms of pure chance. Testing for luck would involve conducting controlled experiments and using a statistical analysis of the results.
Suppose that it is found that some people are lucky and others unlucky. What then might be the nature of luck?
Historically, people have suggested many things. A common explanation for luck involves the gods or other divine (or infernal) beings. For example, for the ancient Greeks the Fates supposedly determined the, well, fate of things. The gods of old were supposed to grant good luck or inflict bad luck. For monotheists, God would take care of luck. Of course, this explanation does involve some rather serious metaphysics-which is problematic.
A second common explanation involves psychic factors-some people are supposed to be able to influence dice rolls and other random events. This is still a rather odd explanation, but does not seem to be impossible-especially when one takes into account the notions in quantum physics about observations. Put far too simply, random events are determined when an observer observes them (as per Schrödinger’s Cat). If that is acceptable science, then it is not much of a stretch to allow that an observer can affect probability by her observations. In short, someone could be lucky or unlucky in a quantum way.
While I tend to intellectually reject the notion of luck, years of gaming and sports have inclined me to suspect it exists. The best evidence I have is for bad luck-one friend of mine has consistently horrible luck through all dice games. So, bad luck seems real.