A. Bublik vs Q. Halys — prediction
›Ranking: #11 vs #95 (better ranked)
›Recent form: 5/10 in recent matches
›Match-sharp: 3 matches in the last 2 weeks
The most reliable signal here is the sheer level difference: Bublik's Elo of 1997 against Halys' 1840, matched by a ranking gulf of #11 to #95, translates into a 63% versus 39% baseline win rate. That gap is not marginal — it reflects a sustained difference in results over the season, not a single hot streak.
Combined with his 72% service points won against Halys' 64%, Bublik's game is built around a weapon that travels well against lower-ranked opposition. Halys' 34% return rate is competent but not enough on its own to flip the class differential.
Gstaad sits at 1050 meters, and the thinner air at that altitude tends to speed up the ball and add value to the bigger server — a dynamic that favors Bublik's 72% service rate more than Halys' 64%. Weather is a secondary factor here: warm and dry at 24°C with only 9 km/h of wind, conditions are calm enough that neither player's precision should be meaningfully compromised.
Rest is where the match picture sharpens further. Bublik arrives with nine days off, while Halys played as recently as yesterday and has logged two matches in the last two weeks against Bublik's three. In best-of-three or best-of-five play, that kind of turnaround can show up in Halys' movement and endurance if the match extends.
Recent form is essentially a wash: both players sit at 5-10 across their last ten matches. Halys carries a one-match win streak into Gstaad, while Bublik is coming off a loss, but neither trend is pronounced enough to override the underlying level gap.
Quality wins are comparable in caliber — Bublik has beaten Tiafoe (Elo 2031) and Kokkinakis (Elo 1948), while Halys has wins over Humbert (Elo 1948) and Arnaldi (Elo 1902). Neither résumé changes the fundamental picture established by Elo and ranking.
The model sets Bublik at 75% to win, while the market prices him at 72% (odds of 1.39), producing a modest +4.8% expected value. That is a real but small edge, and it sits within the normal noise band of a calibrated model that performs close to market efficiency on average.
Being the favorite here is not the same as being a value bet in a large sense — this is a case where model and market largely agree, with a slight tilt toward Bublik. The rest and altitude factors provide plausible extra support beyond the raw numbers, but this remains a moderate, not decisive, edge.
Impact and analysis from real match data (Elo, form, head-to-head, rest, surface vs baseline, weather, altitude). The model ≈ the market on average; the odds already capture almost all the edge. 18+ · gamble responsibly.