TennisEdge's AI model finds no betting value in A. Paszun vs W. Falkowska (Warsaw on 2026-08-03) — the market is efficient, with W. Falkowska favoured at 1.04.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | A. Paszun | W. Falkowska |
|---|---|---|
| Elo ratings-based, surface-weighted |
22.2% | 77.8% |
| Machine learning 29-feature gradient-boosted model |
28.8% | 71.2% |
| Blend what the bot actually prices off |
21.8% | 78.2% |
| Market implied de-vigged from the opening odds |
12.8% | 87.2% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| A. Paszun | hard | 1278 | 10–11 | 21 |
| W. Falkowska | clay | 1486 | 93–74 | 167 |
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.
W. Falkowska leads 2–0 over 2 meetings. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| Aug. 3, 2026 | Warsaw | hard | W. Falkowska | 2 - 0 |
| June 18, 2026 | W35 Bolszewo (Poland) | clay | W. Falkowska | 2 - 1 |
| L | W. Falkowska | 0 - 2 | Aug. 3, 2026 |
| L | D. Gorska | 2 - 0 | July 29, 2026 |
| L | W. Ewald | 2 - 1 | June 25, 2026 |
| W | O. Gniewkowska | 1 - 2 | June 23, 2026 |
| L | W. Falkowska | 1 - 2 | June 18, 2026 |
| W | A. Paszun | 0 - 2 | Aug. 3, 2026 |
| L | C. Noel | 2 - 1 | July 29, 2026 |
| L | H. Vandewinkel | 2 - 1 | July 15, 2026 |
| W | D. Dilek | 1 - 2 | July 14, 2026 |
| W | V. Panshina | 0 - 2 | July 13, 2026 |
All Aug. 3, 2026 results & picks →
═══ A. Paszun vs W. Falkowska — Warsaw (hard) ═══
MODELS vs MARKET
ELO(A) 22.2% | ML(A) 28.8% | blend 21.8% | market implied(A) 12.8%
Betfair: A 6.8 / B 1.03 (exchange). value_pick A, edge 9%.
Models say Paszun ~22-29%; market prices ~13%. Numeric gap present.
ELO/RANKING SANITY
Paszun ranked 1060; Falkowska ranked 504. Paszun ELO-fav-ish (22%) but her
recent WINS are over 1066/1224/1250-ELO fields — farmed rating, inflated by
weak opposition. Discount the model edge; lean PASS unless live form backs it.
RECENT FORM (last 5 each)
Paszun (A): L Gorska(1500) 6-4 7-6 clay 07-29 | L Ewald 07-25 | W Gniewkowska(1224)
| L Falkowska(H2H) 07-18 | W Ozerova(1066) 6-0 6-0. FADING — losing to 1450-1500,
wins only vs sub-1250. Weak, level-dropping player.
Falkowska (B): L Noel 4-6 6-3 7-6 07-29 | L Vandewinkel(quali) 07-15 | W Dilek 07-14
| Istanbul quali run | W35 Amstelveen TITLE 07-05. SHARPENING/active, title winner,
match-sharp. Clearly the stronger side.
H2H
Falkowska beat Paszun 06-18 (3-6 7.8-6.6 2-6) on clay — competitive but Falkowska won.
Since then Falkowska won a title; Paszun faded. Gap has widened, not closed.
FATIGUE
Both last played 07-29 (Falkowska 121min Dublin, Paszun 104min). Both ~4d rest.
Roughly equal; no extreme spot. Falkowska mild surface/travel transition (Dublin→
Warsaw hard) — not enough to move a 1.03 price. Not an edge.
GATES
G1 edge-origin: PASS(numeric) — but edge sourced from farmed-ELO overrating the
weak underdog; qualitative validation ABSENT and CONTRADICTORY.
G2 consensus: stake-tier — models agree but on a discredited (weak-field) rating.
G3 fatal-risk: n/a (ML, not margin).
MARKET CHOICE
ml only. alt_markets: handicap/totals null. AH: not quoted.
betfair_ml present (A 6.8). No cover/total read available.
VERDICT: PASS. The model edge on Paszun rests on weak-field-farmed ELO. Live form
and H2H say Falkowska is clearly the stronger, sharper player (title winner, match-
sharp, beat Paszun) while Paszun is fading and losing to 1450-1500-level opponents.
A BET requires model-market disagreement AND independent qualitative validation;
validation is absent/negative here, and thin ITF exchange liquidity compounds the
risk. No edge to stake.