Y. Wu vs S. Kozlov — prediction

The Elo differential of 135 points (1871 vs 1736) is the single strongest signal in this match, reinforced by Wu's ATP ranking of 101 and a positive trend of 17 spots over recent months. This gap reflects a real, if moderate, quality edge built over a large sample — the model notes 153 tracked matches for Wu, giving the rating some statistical weight even in the softer Challenger pool.

Still, an Elo edge of this size is far from decisive at this level; Challenger fields are volatile, and a 135-point gap translates to a competitive edge, not a lock.

Form is the clearest counter-signal to the ranking gap. Wu has dropped three straight matches and sits at 4-6 over his last ten, while Kozlov is riding a two-match win streak and a 7-3 record over the same span. This divergence suggests Kozlov is playing with more rhythm and confidence entering this match, which can matter as much as raw rating in a single best-of-three or best-of-five encounter.

Momentum doesn't overturn the rating gap on its own, but it narrows the practical gap between the two players beyond what the Elo numbers alone would suggest.

Wu arrives fresher: 10 days since his last match and only one outing in the past two weeks, compared to Kozlov's two matches in 14 days on just 2 days' rest. This rest advantage could matter most if the match extends into a decisive set, when physical freshness often decides close points.

This factor works in Wu's favor and helps offset some of the concern raised by his negative recent form.

Wu's 64% rate on service points is a real strength, and it's the most concrete individual metric available for either player. However, the match conditions — 25°C, 80% humidity, and 24 km/h wind — create an environment where wind can disrupt service rhythm and first-strike precision, a factor that could shave a few points off that serve number without a clear offsetting benefit given the lack of any comparable serve data for Kozlov.

No return-game statistics exist for Kozlov, so it isn't possible to quantify how much Wu's serve strength will actually translate into service breaks or holds against this particular opponent.

The model gives Wu a 69% win probability, but the market prices him at an implied 81% (odds of 1.24), producing a -15% expected value. That gap means the market is considerably more confident in Wu than the model's rating-based estimate — a sign that backing the favorite here offers no value, regardless of the ranking and rest advantages discussed above.

This is also a soft, less-analyzed Elo market (Challenger tier), so the model's edge — or lack of it — should be treated as an estimate rather than a proven signal. Given the negative EV and the favorite's negative recent form, there is no rational basis to treat Wu as an attractive bet at these odds.