Conditional by Design: Sex Differences in Chess Middlegame Decision Style Emerge at the Expert Tier

Studies of sex differences in strategic decision-making have produced inconsistent results across domains. I suggest that part of this inconsistency comes from the question being asked: most studies ask whether differences exist rather than when they appear. I used a title-matched sample of 2,846 FIDE-titled chess players (2,216 women, 630 men) and 210,316 rated online games to measure ten engine-independent indicators of middlegame decision style, and tested for sex differences across skill tiers and time controls. At the elite tier (Elo >= 2500), male and female players differed in their multivariate style profile (Hotelling's T² = 19.70, F(4, 497) = 4.90, p = .0007). The largest single contributor was a greater male tendency to exchange pieces (Holm-adjusted p = .0095). The indicators were organized along an attacking-positional dimension that is stable across time controls in the full sample (Tucker's congruence >= .989), though within the elite tier alone the rapid-time-control comparisons are less stable. The difference was small (|d| about 0.2 to 0.3), the individual distributions overlapped heavily, the effect concentrated at the expert tier, and three independent tests found no modulation by time pressure. Similar patterns appear in five other behavioral domains. I read these results as evidence that sex/gender differences in strategic style are conditional and expertise-dependent rather than stable traits.