Some people seem born lucky. Everything they touch turns to gold. Others are dogged by misfortune.

It’s not just people who might be lucky or unlucky – it can be single acts. When the ball hits a post in soccer, the commentators often say the striker was unlucky. We sometimes argue whether an act was lucky or not. I might say your pool shot was lucky. “Not luck; skill”, you might reply.

Is any of this talk sensible? Is there really such a thing as luck? Do some people have more of it than others (just as some people are better at pool than others)? I think there is a perfectly reasonable way of making sense of talk about luck. But there is no such thing as luck. It isn’t a property, like mass, or an object. Rather, to talk about luck is to talk about how things might easily have gone.

This view entails that no-one has luck. We can’t truly say of someone they’re lucky, meaning they are the kind of person to whom lucky things can be expected to happen.

It has sometimes been suggested that luck exists only if a certain interpretation of quantum mechanics is true: if causality is not “deterministic”. If physical determinism is true then every event that occurs is entirely predictable (in principle), by someone who knows enough about the universe and its laws.

If indeterministic physics is true, then such predictability is not possible: no one, no matter how much they know, can predict every event that happens, even in principle.

I don’t know which interpretation of quantum mechanics is true, but it seems unlikely to me that we need to settle that debate to decide whether some things are lucky. It seems obvious to me that the person who was hit by lightning (on a clear day, if you like) was unlucky, and the person who wins the lotto is lucky.

Here’s how I understand luck. I think something is lucky (or unlucky) for a person if two things are true of it: it matters (somehow) to them, and it might easily not have happened. The second condition needs some explanation.

To say that something might easily not have happened is to say that, given how things were at the time just before, the event might well not have occurred. We might think of this in terms of replaying the event. If I set up the pool table again and ask you to retake the shot, we can discover whether your shot was luck or skill. We will need to do it a few times: you might get lucky twice, but you very unlikely to be lucky ten times in a row.

If every time you try (roughly) the same shot, you sink it I will have to concede: that’s skill, not luck. But if you can’t do it again, you were lucky the first time. Similarly, someone was unlucky to be hit by lightning if it is true that were they to be in similar conditions again, they (probably) would not be hit by lightning. If, on the other hand, lightning is so prevalent around here that any time anyone goes out they get hit, then they weren’t unlucky.

If this is right, there can’t be lucky or unlucky people. At least, there can’t be people who have the property of having lucky events happen to them. Whether I am lucky in doing something depends on how skillful I am at doing things like that. If I’m really good at it, then I am less lucky at succeeding than if I am bad at it.

So, roughly, the more often something happens to someone, the less luck is involved. Of course someone can be lucky or unlucky twice: lightning can strike twice. But the person who is lucky twice, or more, is not a lucky person: their past luck doesn’t give us any reason to expect luck in their future.

There is one way in which we can say that someone is lucky or unlucky. Rather than compare an event to what we would expect to happen, given roughly the same circumstances, we might compare a person’s circumstances or their traits to what is statistically normal for a group. Using this kind of measure, we can say that someone born severely handicapped is unlucky and someone born into wealth is lucky.

What is the relevant group for this kind of comparison? I don’t think there is a single right answer here: it will depend on the context and our aims. For some purposes, a narrower group might be relevant, and for some, a broader. This entails that the same person might be said to be both lucky and unlucky.

Think of the contemporary Australian who loses her job, through no fault of her own. We might say she is unlucky, comparing her to other contemporary Australians. But compared to humanity as a whole, she might be lucky if she remains able to feed and house herself.

This same kind of context sensitivity and relativism is characteristic of luck in events as well. The same event can be lucky and unlucky for a person. Think of someone who misses her flight and takes another one, which then crashes. She is unlucky to be involved in a plane crash, given that she might easily have been on the earlier flight. But if she is the only survivor, she might be lucky, given that everyone else died.

That’s why we can find ourselves saying of someone who has broken three ribs and both legs that they are lucky.

Neil Levy is the author of Hard Luck: How Luck Undermines Free Will and Moral Responsibility.