TennisEdge's AI model finds no betting value in S. Tsitsipas vs K. Rice (Montreal on 2026-08-01) — the market is efficient, with S. Tsitsipas favoured at 1.07.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | S. Tsitsipas | K. Rice |
|---|---|---|
| Elo ratings-based, surface-weighted |
82.9% | 17.1% |
| Machine learning 29-feature gradient-boosted model |
93.8% | 6.2% |
| Serve / return point-by-point simulation |
52.6% | 47.4% |
| Blend what the bot actually prices off |
82.9% | 17.1% |
| Market implied de-vigged from the opening odds |
85.8% | 14.2% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| S. Tsitsipas | hard | 1815 | 133–83 | 216 |
| K. Rice | hard | 1528 | 17–23 | 40 |
Our own ratings, computed from match results since 2018. Players with fewer than 5 rated matches are omitted rather than shown at a provisional starting value. Last updated Aug. 1, 2026.
S. Tsitsipas leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| Aug. 1, 2026 | Montreal | hard | S. Tsitsipas | 2 - 0 |
| W | K. Rice | 2 - 0 | Aug. 1, 2026 |
| L | A. De Minaur | 2 - 1 | July 29, 2026 |
| W | R. Collignon | 2 - 1 | July 19, 2026 |
| W | A. Shevchenko | 1 - 2 | July 18, 2026 |
| W | A. Rinderknech | 1 - 2 | July 17, 2026 |
| L | S. Tsitsipas | 2 - 0 | Aug. 1, 2026 |
| L | M. Lajal | 0 - 2 | July 31, 2026 |
| W | D. Milavsky | 1 - 2 | July 30, 2026 |
| W | P. Llamas Ruiz | 2 - 1 | July 29, 2026 |
| W | H. Wendelken | 1 - 2 | July 29, 2026 |
All Aug. 1, 2026 results & picks →
S. Tsitsipas (75, 27y) vs K. Rice (989, 20y) — Montreal, hard, qualifying-level.
MODELS vs MARKET
ELO/model A 82.9% (Rice 17.1%) | ML A 93.8% (Rice 6.2%) | book implied A 85.8%.
Betfair ML: A 1.02 (~98%) / B 10.0 (~10%). Book A 1.07 / B 6.45.
value_pick B edge 2.9% is a BOOK-only artifact off ELO's Rice 17.1%.
Models split violently on the dog: ELO 17% vs ML 6%. ELO is the lone outlier
vs both ML and a sharp (if thin) exchange that prices Rice at ~10%.
RECENT FORM
Tsitsipas: WON Gstaad title on clay 19 Jul (beat Collignon 1865 in F, Buse 1823),
then hard at Washington 29 Jul — competitive 3-set LOSS to De Minaur (1982, elite)
6-4 3-6 6-3. Match-tough, on-surface, elite-tested. SHARPENING/STABLE.
Rice: ITF/Challenger grinder, opp ELOs 1140-1727. Yesterday (31 Jul) SMASHED in
SF by Lajal (1727) 3-6 2-6. Winning ugly vs weak fields (7-6s vs 1365/1409).
FADING against even mid-level opposition.
ELO/RANKING SANITY
75 vs 989 is a chasm. ELO grants Rice 17% but his rating is farmed grinding
ITF/Challenger fields; Tsitsipas's ELO is a level-dropped former #3 (depressed).
The ELO OVERSTATES Rice. ML's 6% and the exchange's ~10% are the credible reads.
Backing Rice on the ELO number is the textbook untrustworthy-underdog-ELO trap.
FATIGUE
Rice: 704 court-min/14d, 6 matches/14d, 4/7d, back_to_back 3, played yesterday
(2-set loss, no 3-set hangover). Age 20 softens, but a heavy grind and beaten
badly 24h ago. Tsitsipas: 479 min/3 matches, last out 29 Jul — fresher, rested.
Fatigue leans AGAINST Rice, further undercutting any dog case.
MARKET CHOICE
alt_markets: none at analysis time (handicap/totals null). ML only.
Rice at Betfair 10.0: eff odds 1+(9.0*0.95)=9.55; edge only if you trust ELO's
17.1% (0.171-1/9.55=+6.6%) — but true prob is ML/exchange 6-10%, so edge is
illusory. min_odds on a realistic 10% prob = 10.5 already exceeds 10.0.
Tsitsipas at 1.02: min_odds 1.05/0.938=1.12 > 1.02 — no value laying the fav.
AH: not quoted.
GATES
G1 edge-origin: FAIL — staked "edge" exists only off an untrustworthy (farmed)
underdog ELO; ML+exchange+qualitative all refute it. No credible numeric edge.
G2 consensus: stake-tier only — moot given G1 fail (lone-model ELO outlier).
G3 fatal-risk: n/a (ML), but class gap + Rice fatigue + fresh beatdown = no dog case.
VERDICT: PASS. Only the ELO disputes the market and the qualitative read (class
gap, Rice fatigue, Rice fading vs mid-level fields, Tsitsipas match-tough on hard)
sides with ML/exchange, not the ELO. Longshot has no real value; favourite unbackable at 1.02.