TennisEdge's AI model finds no betting value in D. Altmaier vs R. Collignon (Kitzbuhel on 2026-07-21) — the market is efficient, with R. Collignon favoured at 1.58.
═══ D. Altmaier vs R. Collignon — Kitzbuhel (clay), 1/16-finals ═══
MODELS vs MARKET
ELO: A 33.8% | ML: A 33.3% | Sim: A 32.4% → consensus B ~67%
Book: A 2.40 / B 1.58 (impl A 39.7%) | Betfair ML: A 1.52 / B 1.58
⚠ Exchange market malformed: 1/1.52 + 1/1.58 = 129% overround — A-side price
unusable/stale (book 2.40 vs exchange 1.52 contradiction). Executable side = B @1.58 only.
value_pick B, model edge ~4% pre-commission (0.676 − 0.633)
RECENT FORM (last 5 shown)
Altmaier: L Darderi(1916,Elite) 4-6 4-6 | W Gaston(1654,Strong) 6-2 7-5 DOM |
L Molcan(1554) | L Bergs(1616) | W Kovacevic(1422,Weak) GRIND
Deeper: Halle scalps of Medvedev(1842) GRIND + Hurkacz(1819). Clay-solid, mixed results.
Collignon: L Tsitsipas(1852,Elite) Gstaad FINAL yesterday GRIND | W Cerundolo(1867,Elite) GRIND |
W Vacherot(1778,Strong) GRIND | W Sonego(1628) GRIND 7-6 7-6 | W Skatov(1622) GRIND
5-1 hot streak vs Strong/Elite — but every single match a GRIND, zero DOM wins.
FATIGUE
Altmaier: 2 matches/14d, 216 min, last played Jul 15 → 5 days rest. Fresh.
Collignon: 5 matches/7d, 15 sets, 885 min court time, 2x back-to-back, played the
Gstaad FINAL yesterday (Jul 19), travels Gstaad→Kitzbuhel same week.
Court-time diff 669 min → Collignon MASSIVELY disadvantaged.
NARRATIVE
Let-down: Collignon lost Gstaad Final 1d ago → NOTE (deep-run hangover + no rest).
Title recency: Collignon won Halle Final 36d ago (grass) — hot summer, but the bill
for this stretch is coming due.
Venue: no Kitzbuhel history for either player in window.
VERDICT — PASS
Model edge on B is thin: commission-adj 1.58 → eff 1.551 (break-even 64.5%); model
67.6% → ~3% edge, below min_odds bar (1.05/0.676 = 1.553 ≈ the price itself).
Qualitative read KILLS the B side rather than validating it: 885 min in 7 days,
final yesterday, every win a grind — my true B prob ≤ 60%, i.e. negative edge at 1.58.
The fatigue story points to Altmaier, but there is no executable A price: exchange
shows 1.52 (would need 66%+ win prob vs model 33%), book 2.40 is a quote the exchange
contradicts — stale, not value. No handicap/totals lines offered. Nothing stakeable.