MODEL PREDICTION · 2026-07-20

R. Collignon vs D. Altmaierprediction

✗ Missed
COLLIGNONWIN PROBABILITYALTMAIER
56%
model prob.
@1.60
odds · 63% impl.
🌡21° · 52% hum760 m altitudeRest 1d vs 5d🎾Serve 67%📈Form 6/10
THE MODEL'S REASONING

Ranking: #42 vs #60 (better ranked)

Recent form: 6/10 in recent matches

Match-sharp: 5 matches in the last 2 weeks

Model 56% vs market 63% → the model sees it as less likely than the odds

WATCH FOR

!Played a long match (3 sets) very recently — possible fatigue

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.
@1.79
fair odds
−10.6%
expected value
HOW EACH FACTOR MATTERS
Level (Elo/ranking)▸ Collignon●●●
Elo gap of 157 points (2021 vs 1864) and ranking #42 vs #60 align with the 51% vs 42% baseline model edge for Collignon.
Rest▸ Altmaier●●●
Collignon has 1 day of rest after 5 matches in 14 days (including a Gstaad final yesterday) vs Altmaier's 5 days off and only 2 matches — fatigue risk is real.
Form= Even●●
Both are on a 1-match losing streak; Collignon's 6/10 record edges Altmaier's 5/10, but Altmaier's quality wins (Medvedev 2048, Hurkacz 1958) outrank Collignon's (Vacherot 1990, Cerundolo 1960).
Serve/return= Even
Serve percentages are close (67% Collignon vs 65% Altmaier) and Altmaier's better return (34% vs 32%) largely offsets the serve gap.
Altitude▸ Collignon
At 760m the thinner air modestly speeds serves, a small nudge toward Collignon's slightly higher 67% serve-point rate.
Value▸ Altmaier●●●
Model gives Collignon 56% vs a market-implied 65% (odds 1.54), producing a -13.9% EV — the price overstates his edge.
LEVEL AND RANKING

Collignon's higher Elo (2021 vs 1864) and better ranking (#42 vs #60) are the clearest structural advantages in this match, and they line up with the baseline model split of 51% to 42% before adjustments. This is a real quality gap, not noise — it reflects sustained results over time rather than a single hot streak.

Still, the model's final call of 56% is only marginally above a coin flip, meaning the ranking and Elo edge gets diluted once other factors (mainly rest and recent workload) are folded in.

FATIGUE AND SCHEDULE

The single biggest swing factor against Collignon is his schedule: one day of rest after five matches in the last two weeks, including a final in Gstaad completed just a day before this match. Altmaier, by contrast, arrives with five days of rest and only two matches played in the same span — a much fresher physical state.

This kind of turnaround, especially coming off a deep tournament run, tends to show up in the back half of matches through slower movement and shorter concentration on serve. It's a tangible drag on Collignon's favorite status that the ranking gap alone doesn't capture.

FORM AND SERVE PATTERNS

Both players enter on a one-match losing streak, so momentum is roughly a wash. Collignon's 6-4 record over his last ten edges Altmaier's 5-5, but the quality of wins tilts the other way — Altmaier beat a higher-Elo opponent in Medvedev (2048) and Hurkacz (1958), while Collignon's best scalps (Vacherot 1990, Cerundolo 1960) are a notch below.

On serve, the numbers are close enough to mostly cancel out: Collignon serves at 67% against Altmaier's 65%, but Altmaier is the better returner (34% vs 32%). Neither player should expect to dominate the other's service games based on these figures alone.

VALUE READ

Collignon is the favorite by both the model (56%) and the market (65% implied by 1.54 odds), but the market is pricing him well above what the model supports — a gap that produces a -13.9% expected value. Being the favorite here does not equate to a betting edge; if anything, the price looks inflated relative to the model's read of the match.

Given the fatigue evidence — congested schedule and a deep run ending just one day prior — there's a reasonable case that the market hasn't fully priced in Collignon's physical state. On balance, this looks like a match to watch rather than back: the favorite tag is legitimate on paper, but the numbers don't support paying this price for it.

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 →