TennisEdge's AI model finds no betting value in N. Mejia vs M. Landaluce (Montreal on 2026-08-02) — the market is efficient, with M. Landaluce favoured at 1.29.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | N. Mejia | M. Landaluce |
|---|---|---|
| Elo ratings-based, surface-weighted |
25.2% | 74.8% |
| Machine learning 29-feature gradient-boosted model |
29.4% | 70.5% |
| Serve / return point-by-point simulation |
52.1% | 47.9% |
| Blend what the bot actually prices off |
25.2% | 74.8% |
| Market implied de-vigged from the opening odds |
26.4% | 73.6% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| N. Mejia | hard | 1591 | 88–100 | 188 |
| M. Landaluce | hard | 1731 | 73–47 | 120 |
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. 3, 2026.
N. Mejia leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| Aug. 3, 2026 | Montreal | hard | N. Mejia | 2 - 1 |
| W | M. Landaluce | 2 - 1 | Aug. 3, 2026 |
| W | M. Trungelliti | 1 - 2 | Aug. 1, 2026 |
| L | A. Hernandez | 0 - 2 | July 26, 2026 |
| W | M. Soto | 2 - 0 | July 12, 2026 |
| W | J. P. Varillas | 2 - 1 | July 11, 2026 |
| L | N. Mejia | 2 - 1 | Aug. 3, 2026 |
| L | M. Zheng | 2 - 0 | July 29, 2026 |
| L | S. Kwon | 3 - 0 | June 29, 2026 |
| L | J-L. Struff | 2 - 1 | June 22, 2026 |
| L | T. Atmane | 2 - 1 | June 15, 2026 |
All Aug. 2, 2026 results & picks →
MEJIA (A) vs LANDALUCE (B) — Montreal R1 (hard)
MODELS vs MARKET
ELO: B 0.748 | ML: B 0.706 | Sim: B 0.748 | Pinnacle no-vig: B 0.688
Betfair ML: A 3.2 / B 1.41 (B implied 0.709). Book A 3.6 / B 1.29.
Consensus: all three models + market agree B favored ~0.71-0.75. Efficient.
RECENT FORM (last 5 each)
Mejia (26, #168): W Montreal quali-Final 08-01 (3-set GRIND, won yesterday);
L Los Cabos SF 07-26 (5-7 2-6); W Bogota-Final 07-12 clay (6-1 7-5, DOM);
W Bogota SF (3set); W Bogota QF (DOM). PATH: match-sharp hard-court qualifier,
good recent form incl. Bogota title. SHAPE: STABLE/positive.
Landaluce (20, #94): L Los Cabos R32 07-29 (6-4 6-3 straights); L Wimb 06-29
(Kwon, straights); L Mallorca 06-22 (3set); L Halle 06-15 (3set); W Halle-Final
06-14. SHAPE: FADING (lost last 2 straights), only 1 match/14d = some rust.
FATIGUE
Mejia: WON a 3-set match yesterday (08-01) -> -4.5pp hangover haircut (age 26,
baseline). Sharp-path partly offsets but toll is real; drops his ~0.25-0.29 lower.
Landaluce: 1 match/14d, last 07-29 (4d ago) — rested, not fatigued.
MARKET CHOICE
ML B: eff odds 1.3895 (0.95 comm) -> implied 0.720. Model best 0.748 (elo/sim),
ml 0.706. Edge ~+2.8% (elo) / negative (ml). BELOW 5% bar -> pass.
ML A: eff odds 3.09 -> implied 0.324. Model max 0.2945. Market prices Mejia ABOVE
the models -> negative edge. No numeric edge to stake.
Totals: model over 0.528 @21.5; market line 22.0 -> over ~0.50 vs implied 0.520.
Negligible -> pass.
Handicap A+3.5 @1.92: BLOCKED by self-contradiction guard — Mejia is fatigue-
haircut for the win, cannot also claim his cover. B-3.5 @2.02: no model cover
data; scrappy match-sharp qualifier = tight-scoreline risk (G3 margin veto). pass.
AH shadow: A +3.5 @ 1.92 (cover thesis blocked by self-contradiction guard).
refutation_check: not run — PASS decision, no web access permitted.
VERDICT: PASS. Full model+market consensus on B with only ~2.8% best-case edge
(below 5% bar); dog Mejia is already priced above the models AND carries a 3-set
hangover, so the sharp-qualifier story cannot lift him to a numeric edge. No
bettable market. Predicted winner: Landaluce.