Is Chess Luck or Skill?

The question turns up constantly, usually as an argument between someone who plays chess and someone who plays poker. It has a clean answer on the mechanics and a messier one on outcomes, and most arguments happen because the two sides are answering different questions.

What "perfect information" means

Game theorists put chess in a specific category: a two-player, zero-sum, perfect-information game with no chance element. Unpacking that:

• Perfect information — both players can see everything relevant at all times. No hidden cards, no fog of war, no concealed pieces.
• No chance element — nothing is randomised. The starting position is fixed, and every change to it is a deliberate choice by a player.
• Zero-sum — your gain is exactly your opponent's loss.

Backgammon has perfect information but dice. Poker has a random deal and hidden cards. Go and draughts sit in the same category as chess. Among widely played games, that category is fairly small.

There is a stronger theoretical point too. Because chess is finite and has perfect information, it is solvable in principle — a proof exists that with perfect play from both sides the game has a determined result, whether that is a win for White or a draw. Nobody knows which, because the number of positions is far too large to compute. But the answer exists and does not depend on chance.

So where does luck come in?

Here is where the honest answer gets less tidy. Chess results contain chance even though chess rules do not.

Pairing. In a tournament you do not choose your opponents. Drawing a favourable pairing in the last round is luck by any ordinary meaning of the word.

Preparation collisions. If your opponent happens to play the exact opening line you spent last week studying, you get an enormous advantage that had nothing to do with skill in that game.

Human condition. Whether your opponent slept badly, is distracted, or is ill affects the result and is outside anyone's control.

Blunder timing. Both players make mistakes. Whether a given mistake happens in a position where it costs a pawn or costs the game is partly a matter of where in the game it lands.

None of this is randomness in the game. It is variance in the circumstances around the game — the same kind of chance that affects a chess result also affects a job interview or an exam.

The test that settles it: does skill persist?

The useful question is not "is there any chance at all" but "how quickly does skill dominate". This is measurable, and chess sits at an extreme.

That middle column is the mechanical answer. The right-hand column is the practical one, and it is where chess separates sharply from the games it gets compared to. In poker, variance means a strong player needs a large sample before their edge shows. In chess, a 400-point rating gap shows up in a single game with near-certainty.

This is why the Elo rating system works as well as it does. It was designed for chess precisely because results are consistent enough for a numerical model to predict them.

The strongest argument for "some luck exists"

Worth stating fairly rather than dismissing. Chess is too complex for any human to calculate exhaustively, so every player is operating on incomplete analysis. When you choose a move because it "feels right", you are making a judgement that might be correct or might not, and you frequently will not know which until several moves later.

You could reasonably call that luck: two moves look equally good to you, you pick one, and it happens to be the one that works. From your subjective position the outcome was uncertain even though the position was objectively determined.

The counter is that this is uncertainty from ignorance, not randomness. A stronger player facing the same choice would not be guessing. As skill increases, that uncertainty shrinks — which is the opposite of how genuine randomness behaves. Dice do not become less random as you improve.

Why the answer has legal consequences

This is not purely a bar argument. In several jurisdictions the legal distinction between a game of skill and a game of chance determines whether an activity can be played for money, how it is taxed, and whether it falls under gambling regulation.

Chess sits comfortably on the skill side of that line almost everywhere, which is why chess tournaments with prize funds are not treated as gambling. Games like poker occupy far more contested ground, and the classification has been litigated repeatedly in different countries with different outcomes.

Bottom line

Chess contains no randomness. Whether it contains "luck" depends on whether you mean the rules or the result — the rules have none, and the results have the ordinary variance that any human contest has. What chess does not have is the kind of structural variance that lets a weak player beat a strong one, and that is the difference that actually matters.

If you want to test the point directly, play our computer opponent at its highest level. You will not get lucky. You will need to get better — which is rather the whole argument.

Questions readers ask

Chess is not alone at the skill end of the spectrum. Classic arcade games sit there too: a snake game with wrap-around walls has zero randomness — every crash is pilot error, which is exactly what makes beating your own high score satisfying.