TennisEdge's AI model finds no betting value in T. Maria vs I. Marton (Toronto on 2026-08-01) — the market is efficient, with T. Maria favoured at 1.02.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | T. Maria | I. Marton |
|---|---|---|
| Elo ratings-based, surface-weighted |
87.8% | 12.2% |
| Machine learning 29-feature gradient-boosted model |
89.7% | 10.3% |
| Serve / return point-by-point simulation |
43.7% | 56.3% |
| Blend what the bot actually prices off |
87.8% | 12.2% |
| Market implied de-vigged from the opening odds |
88.7% | 11.3% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
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. 4, 2026.
T. Maria leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| Aug. 1, 2026 | Toronto | hard | T. Maria | 2 - 0 |
| L | C. McNally | 1 - 2 | Aug. 4, 2026 |
| W | I. Marton | 2 - 0 | Aug. 1, 2026 |
| L | R. Zarazua | 0 - 2 | July 29, 2026 |
| W | T. Gibson | 0 - 2 | July 27, 2026 |
| W | K. Volynets | 0 - 2 | July 12, 2026 |
All Aug. 1, 2026 results & picks →
T. Maria (A) vs I. Marton (B) - Toronto, hard court
===========================================================
MODELS vs MARKET
ELO: A 87.8% | ML: A 89.7% | model A 87.8%
Book: A 1.02 / B 8.0 | Exchange (betfair_ml): A 1.02 / B 21.0
value_pick B edge 0.88 pct - computed vs SOFT book quote 8.0, not the
executable exchange price 21.0. Exchange implies B ~4.8 pct.
RECENT FORM (last 5)
A T. Maria (rank 54): Newport Final winner + Eastbourne Final (lost to Keys),
beat Ostapenko/Zakharova on grass - legit top-60 tour level. Recent dip:
lost Memphis qualifying 4-6 4-6 to Zarazua (1500). Transition grass->hard.
B I. Marton (rank 1373, age 18): W35 Saskatoon QF loss 0-6 3-6 to J.Lu (1461);
W75 Granby loss 3-6 3-6; string of W15 losses incl. 6-0 6-0. Pure ITF-junior
level, mostly losing to 1350-1500 opponents. No credible tour-level threat.
FATIGUE (context only, not an edge)
A: 2 matches/7d, last 07-29, 133min - mild, no extreme spot.
B: last match 07-24 (8d), some rust. Neither moves a 98/2 line.
NARRATIVE
Enormous class gap: WTA top-60 vs ITF W15/W35 junior. Marton ELO near-default
(opponents all logged 1500) - farmed/undefined rating case. Exchange 21.0
(~4.8 pct) is the sharp estimate; ELO 12 pct overrates a junior with no
reliable rating and a losing ITF record.
GATES
G1 edge-origin: FAIL - the only edge (B) rests on an unreliable near-default
junior ELO (12 pct) contradicted by Marton losing ITF form; exchange prices
B at ~4.8 pct and is the executable truth. No numeric edge survives.
G2 consensus: n/a (stake-tier only) - models + market agree A wins ~90 pct+.
G3 fatal-risk: n/a - no margin bet.
PRICING
A at exchange 1.02: eff 1.019, implied 98.1 pct vs model 88 pct -> neg edge;
min_odds 1.167, price far below -> auto-PASS.
B at exchange 21.0: eff 20.0, implied 5.0 pct; ELO 12 pct farmed/unreliable and
refuted by form -> no credible edge. Manufactured-value longshot trap.
AH: not quoted. alt_markets: handicap/totals none at analysis time.
VERDICT: PASS. Full model-market consensus on the winner; the lone B value is
manufactured by a soft book quote + unreliable junior ELO the qualitative read
refutes. No side clears its bar.