R. Collignon vs S. Tsitsipas — prediction

Collignon's higher Elo rating (2037 vs 1932) and better ranking (#43 vs #87) are the clearest structural edge in this matchup, reinforced by a positive ranking trend (+21) against Tsitsipas's decline (-5). This gap explains most of the model's lean toward Collignon despite the raw baseline number (47% vs 52%) actually tilting slightly toward Tsitsipas before ranking and Elo are factored in.

Recent form adds another layer favoring Collignon: both players are riding 4-match win streaks, but the quality of wins differs. Collignon beat Vacherot (Elo 1990) and Cerundolo (Elo 1960), both higher-rated opponents, while Tsitsipas's best recent scalp was Buse (Elo 1913). This suggests Collignon's current form is being tested against tougher competition.

The serve numbers cut against the favorite. Tsitsipas wins 73% of his service points compared to Collignon's 68%, a 5-point gap that matters over best-of-three or five sets since it reduces break chances for Collignon. Tsitsipas also edges the return column slightly (33% vs 32%), meaning he's marginally better at both ends of the point.

The 1,050m altitude in Gstaad thins the air and speeds up the ball, a dynamic that historically rewards the better server. Since Tsitsipas already holds the stronger serve number here, altitude likely amplifies his advantage on that specific dimension rather than helping Collignon close the gap.

Rest is a non-factor: both players are one day removed from their last match and have played 4 matches in the past two weeks, so neither carries a scheduling disadvantage into this one.

Weather (22°C, 49% humidity, 12 km/h wind, described as mild and dry) doesn't point to a clear beneficiary. There's no data suggesting either player's game is especially wind-sensitive or humidity-dependent, so this factor is treated as neutral.

The model sets Collignon at 60% to win, well above the market's implied 46% (odds of 2.18), producing a stated EV of +30.9%. That gap is driven mainly by the Elo and ranking disparity, which the model weighs heavily, while the serve/return numbers and altitude dynamic actually lean toward Tsitsipas.

This is a case where being the favorite doesn't automatically mean value is real — the model and market disagree by a wide margin, and that disagreement should be treated with some caution given Tsitsipas's edge in the serve and return columns. If you trust the model's ranking/Elo weighting, the price looks generous; if you weight the serve data more heavily, the gap looks less convincing. Either way, this isn't a lock — it's a probabilistic lean with a specific mechanical trade-off working against it.