MODEL PREDICTION · 2026-07-25
CLAY

A. Bublik vs Q. Halysprediction

Kitzbuhel
✗ Missed
BUBLIKWIN PROBABILITYHALYS
73%
model prob.
@1.45
odds · 69% impl.
H2H 0–1 Bublik🎾Serve 72%📈Form 6/10 · 3✓
CONDITIONS OF THE MATCHin the modelcontext
Surface
Clay

Slow court, high bounce: longer points, rewards whoever holds up from the baseline.

Temperature
27°C

Warm: the ball flies a little more and fitness counts.

Humidity
37%

Dry air: the ball travels normally.

Wind
6 km/h

Light wind: no noticeable effect.

Altitude
760 m

Moderate altitude: the ball flies a little more than at sea level.

Surface feeds the model (surface specialization is one of its factors). Weather and altitude are context we publish for you — they do NOT move the probability.

OUR MODEL'S REASONING

Ranking: #11 vs #90 (better ranked)

Head-to-head: 0-1 against

Recent form: 6/10 in recent matches

On a streak: 3 wins in a row

Match-sharp: 4 matches in the last 2 weeks

WATCH FOR

!Unfavorable head-to-head record (0-1)

Calibrated model probability (~65% out-of-sample accuracy). Not a guarantee: the model ≈ the market on average, so the odds already capture almost all the edge. 18+ · gamble responsibly.
@1.36
fair odds
+6.2%
expected value
HOW EACH FACTOR MATTERS
Level (Elo/ranking)▸ Bublik●●●
Bublik's Elo 2009 vs 1901 and #11 vs #90 ranking translate into a 62% vs 40% baseline model edge favoring him.
Head-to-head▸ Halys●●
Halys won the only prior meeting, in 2026 at ATP level, giving him a psychological and tactical marker against Bublik.
Form▸ Halys●●
Halys is on a 4-match streak and beat Bublik (Elo 2009) plus Vacherot (1974); Bublik's 3-win streak follows two losses in the last 10.
Rest▸ Bublik●●
Both had 1 day off, but Halys played 7 matches in 14 days versus Bublik's 4, a heavier physical load that could show late in sets.
Serve/return▸ Bublik
Bublik's 72% serve-points-won edges Halys's 69%, while return numbers (32% vs 33%) are essentially even, giving Bublik a slight serving edge.
Altitude▸ Bublik
At 760 m with warm (28°C), dry (32% humidity) air, the ball moves faster, a small boost to the better server, Bublik at 72% vs 69%.
CLASS GAP

The clearest signal in this match is the level difference: Bublik sits at Elo 2009 and world #11, well above Halys's 1901 Elo and #90 ranking. That gap is reflected directly in the baseline model, which gives Bublik 62% to Halys's 40% before any adjustments — a substantial structural edge built on sustained results over time, not a single data point.

RECENT MOMENTUM

Form and head-to-head tell a more complicated story. Halys enters on a 4-match win streak that includes a victory over Bublik himself (Elo 2009) and another over Vacherot (1974), while Bublik's own 3-match streak comes after a rockier stretch (5 wins in his last 10). The single head-to-head meeting also went to Halys, in 2026 — recent enough to matter as a psychological data point, even if it doesn't offset the larger class gap.

SCHEDULE LOAD

Both players are coming off Kitzbuhel semifinal runs with just one day of rest, so neither has a clean recovery advantage in that narrow sense. But the broader workload numbers diverge sharply: Halys has played 7 matches in the last 14 days against Bublik's 4. That nearly double match count is a tangible fatigue risk for Halys as the match progresses, even if today's short-term rest is identical.

SERVE AND CONDITIONS

On serve, Bublik's 72% points won is a few points clear of Halys's 69%, while return numbers are essentially a wash (32% vs 33%). The 760 m altitude combined with warm, dry, low-wind conditions (28°C, 32% humidity, 6 km/h wind) tends to speed up the ball slightly, which generally benefits the better server — a small additional tilt toward Bublik rather than a decisive one.

VALUE READ

The model rates Bublik at 73%, noticeably above the market's implied 67% (odds of 1.49), producing a 9.2% expected value. That gap stems mainly from the level and altitude factors outlined above, while the head-to-head loss and Halys's current form are legitimate reasons for some market caution.

This is a genuine but modest edge, not a guarantee: Bublik remains the more probable winner on the numbers, but 27% for Halys is far from negligible given his recent streak and the direct head-to-head result. Treat the positive EV as a data-supported lean rather than a sure thing.

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.

Analyze today's matches →