TennisEdge's AI model finds no betting value in T. Preston vs E. Jones (Toronto on 2026-08-01) — the market is efficient, with T. Preston favoured at 1.73.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | T. Preston | E. Jones |
|---|---|---|
| Elo ratings-based, surface-weighted |
53.4% | 46.6% |
| Machine learning 29-feature gradient-boosted model |
54.6% | 45.4% |
| Serve / return point-by-point simulation |
47.9% | 52.1% |
| Blend what the bot actually prices off |
53.4% | 46.6% |
| Market implied de-vigged from the opening odds |
53.6% | 46.4% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| T. Preston | hard | 1773 | 139–62 | 201 |
| E. Jones | hard | 1735 | 90–45 | 135 |
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.
E. Jones leads 2–0 over 2 meetings. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| May 1, 2026 | W100 Gifu | hard | E. Jones | 2 - 1 |
| Nov. 3, 2024 | ITF W75 Sydney Women | hard | E. Jones | 2 - 0 |
| W | M. Inglis | 0 - 2 | Aug. 1, 2026 |
| W | M. Sawangkaew | 2 - 1 | July 31, 2026 |
| W | P. Hon | 0 - 2 | July 30, 2026 |
| W | H. Dart | 0 - 2 | July 30, 2026 |
| W | H. Kaji | 0 - 2 | July 29, 2026 |
| L | L. Tararudee | 0 - 2 | Aug. 3, 2026 |
| W | V. Lepchenko | 1 - 2 | Aug. 1, 2026 |
| L | M. Sawangkaew | 0 - 2 | July 30, 2026 |
| W | M. Brengle | 2 - 1 | July 30, 2026 |
| W | C. Zhao | 2 - 1 | July 28, 2026 |
All Aug. 1, 2026 results & picks →
T. Preston (A, age 20, #126) vs E. Jones (B, age 18, #147) — WTA hard, Final.
MODELS vs MARKET
ELO A 0.5345 | ML A 0.5459 | model A 0.5345 | market-implied A 0.536 (odds A 1.73 / B 2.00).
Full four-way consensus on Preston as a slight favourite. value_pick B edge = +0.0017 (noise).
RECENT FORM (last 5 shown; classification uses all)
Preston: deep Vancouver run this week, 4 matches, match-sharp.
07-31 SF vs Sawangkaew (1748, Strong) WON 3-set GRIND-win.
07-30 QF vs Dart (1710, Strong) WON straight, DOM over Strong.
07-30 R16 vs Hon (1500, Avg) WON, tiebreak GRIND.
07-29 R32 vs Kaji (1527, Avg) WON straight, DOM.
07-22 Evansville R32 vs Broadus (1500) Lost.
Shape: SHARPENING/STABLE — DOM over Dart 1710, win over Sawangkaew 1748.
Jones: 3 matches at Vancouver, match-sharp but ended on a blowout loss.
07-30 QF vs Sawangkaew (1733, Strong) LOST 3-6 2-6 (blowout).
07-30 R16 vs Brengle (1562, Avg) WON 3-set GRIND.
07-28 R32 vs Zhao (1461, Weak) WON 3-set GRIND — under-performing vs weak.
07-22 Evansville R32 vs Akli (1500) Lost.
06-24 Wimbledon SF vs Gasanova (1571) Lost.
Shape: FADING/UNDER-performing — grinding past weak fields, blown out by Strong.
COMMON OPPONENT: Preston beat Sawangkaew (SF, 5-7 6-3 6-3); Jones lost to Sawangkaew 3-6 2-6 (QF). Favours Preston — but only supports the market's existing favourite pricing, no disagreement.
FATIGUE (Step 2)
Preston: won a 3-set match YESTERDAY (07-31) -> 3-set-hangover haircut -4.5pp, but age 20 (<=22, fast recovery) softens one tier -> ~-2 to -3pp effective. Pushes her toward Jones but stays inside the consensus band.
Jones: last match 07-30, not back-to-back; roughly neutral. Market over-prices visible fatigue; the haircut narrows but does not flip value to Jones.
NARRATIVE
Both match-sharp on hard, no surface transition. No first-Final let-down spot in window. No motivation edge shown in data.
PRICING
alt_markets: none at analysis time. No exchange (betfair_ml null) -> ML book-priced, bar = edge >= 8% vs book. Neither side clears it: models put Jones ~0.455-0.465 vs book-implied ~0.50 (Jones OVER-priced by market), and Preston edge ~0 vs 1.73.
GATES
G1 edge-origin: FAIL — no numeric model-vs-market disagreement pre-narrative; all four estimates within ~1pt.
G2 consensus: stake-tier only (n/a, no bet).
G3 fatal-risk: n/a (ML, no bet).
VERDICT: PASS. Full model-market consensus on a slight favourite, no exchange liquidity, no exploitable disagreement. Qualitative (common-opp, shape) merely confirms the market. Textbook pass.