Sixty Per Cent Over Fifty Games Means Anything From 46 to 74

Streaks are what randomness looks like

A player winning six in a row and then losing five feels like a form curve and is exactly what an unchanging fifty-per-cent chance produces over enough games. Runs of that length appear regularly in genuinely random sequences, and a sequence with no runs at all would be the suspicious one. This matters because the natural reaction is to explain the streak — a change of tactics, a bad night, a new patch — and the explanation then feeds back into decisions. Before attributing a run to a cause, ask how often a run that long would occur by chance alone across the number of games actually played.

The opponents are not in the number

A win rate treats every game as equivalent, and no competitive schedule is. Sixty per cent against the strongest half of a league and sixty per cent against the weakest are not the same performance, and a player who can choose opponents can raise the figure without improving. This is why ranking systems that adjust for the strength of the opposition exist, and why they routinely disagree with a raw win rate about who is better. Where such a rating is available, it answers the question people are actually asking; where it is not, the raw rate has to be read together with who was played and in what conditions.

Frequently asked questions

How many games before a win rate means something?

It depends on how large a difference you want to detect. Distinguishing a 60 % player from a 50 % one takes far fewer games than distinguishing 55 % from 50 %, because the margin has to shrink below the gap being tested. As a rough guide, the interval halves each time the sample quadruples: fifty games give about fourteen points, two hundred give seven, eight hundred give three and a half. Decide first what difference would change a decision, then work out how many games make the interval narrower than that.

Should draws count as half a win?

That is the usual convention and it is a choice worth stating rather than assuming. Counting a draw as half suits a sport where a draw is genuinely halfway between the outcomes, and it distorts one where a draw is closer to a loss for the stronger side or where the points system does not treat it that way. The alternative — reporting wins, draws and losses separately — is longer and never ambiguous. Whatever is chosen, the same convention has to be used on both sides of any comparison, because a table mixing the two is comparing different quantities under one heading.