C. Tauson vs N. Bartunkova — Prediction & Analysis

TennisEdge's AI model finds no betting value in C. Tauson vs N. Bartunkova (Toronto on 2026-08-04) — the market is efficient, with N. Bartunkova favoured at 1.87.

Model probabilities

Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.

We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?

Elo ratings

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 July 14, 2026.

Recent form

All Aug. 4, 2026 results & picks →

Full analysis

C. Tauson vs N. Bartunkova — WTA Toronto (WTA 1000), hard, 1/32-finals

═══ MODELS vs MARKET ═══
ELO   A 56.0%   |  ML  A 58.9%   |  Sim (model_prob) A 52.3%
Book  1.95 / 1.87  (implied A 49.0%) — CONTEXT ONLY
Betfair ML 1.83 / 2.04 → implied A 54.6%, devigged A 52.7% ← the executable truth
Commission-adjusted A: 1.7885 → breakeven 55.91%

The brief's "value_pick A, edge 3.4%" is an artifact of the stale book quote (1.95).
The exchange prices Tauson at 1.83, i.e. the market already has her where ELO does.
Pre-narrative edge vs the EXECUTABLE price:
  ML alone   58.94% → +3.03%  (probe boundary, min 1.748)
  ELO alone  56.02% → +0.11%  (nil)
  ELO+ML avg 57.48% → +1.57%  (below probe bar)
  Sim        52.33% → −3.58%  (market is RIGHT / A overpriced)
B side is negative on every model (−2.6% to −9.2%). No side clears the full-ladder 5% bar;
only the single most bullish model reaches the 3% probe floor, and the sim contradicts it.

═══ RECENT FORM ═══
C. Tauson (23, rank 18) — last 5:
  07-27 Washington R32  L  Cocciaretto (1793)  4-6 7-6 7-6   GRIND loss, competitive
  07-18 Athens SF       L  Krejcikova (1857)   1-6 6-4 3-6   neutral-positive vs Elite
  07-17 Athens QF       W  Bejlek (1757)       6-2 6-3       DOM over Strong
  07-16 Athens R16      W  Tona (1500)         6-4 6-0       DOM over Avg
  07-14 Athens R32      W  Hibino (1500)       7-5 6-4       DOM over Avg
  2026 context: 0-3 on clay with heavy losses (Rome 1-6 1-3, Strasbourg 1-6 5-7),
  patchy grass. Hard is her surface; Athens SF is the one genuine run in the window.
  Shape: STABLE — the Athens DOMs are the high point, and they came against Avg/Strong
  at a weak-field 250, not tour-level opposition.

N. Bartunkova (20, rank 94) — last 5:
  08-03 Toronto R128    W  Andreescu (1742)    7-5 6-4       DOM over Strong, at THIS venue
  07-23 Prague R16      L  Tararudee (1502)    6-7 6-3 6-7   poor, on clay
  07-21 Prague R32      W  Yuan (1543)         3-6 6-4 6-1   GRIND over Avg
  07-03 Wimbledon R16   L  Krejcikova (1850)   3-6 5-7       competitive loss to Elite
  07-01 Wimbledon R32   W  Siniakova (1766)    2-6 4-6       win over Strong
  Wider window is a genuine breakout: Birmingham FINAL (06-07), beat Keys (1908) in Rome,
  Mertens (1774) and Shnaider in Berlin, took a set off Sabalenka.
  Shape: SHARPENING — DOM over Andreescu yesterday is over-performance vs bracket and
  wipes the Prague clay blip.

═══ PATH & SHARPNESS (Step 1b) ═══
Bartunkova: 1 match at this venue, WON it yesterday in straight sets over the home-crowd
  name. Conditions-adjusted, timing grooved, confidence high. Match-sharp.
Tauson: FIRST match at Toronto, last match anywhere 07-27 (8 days ago) and it was a
  3-set LOSS. Direct entry into a 1000 after a week idle — rust risk, wider margin
  variance in match one. Path favours the DOG, not the side the numbers want.

═══ ELO / RANKING SANITY ═══
Bartunkova ranked 94 yet models make her ~44-48% against the world #18 — a large
ELO-over-rank divergence. Crucially it is NOT farmed: her rating is built on Keys (1908),
Mertens (1774), Siniakova (1766), Andreescu (1742), a Birmingham final. Earned rating,
ranking is simply lagging a 20-year-old's breakout. No discount applies — which means the
ML's 58.9% for Tauson is the number under suspicion here, not Bartunkova's ELO.

═══ FATIGUE (Step 2) ═══
Tauson: 1 match/14d, 136 min, 8 days rest. Fresh — but "fresh" here reads as rusty.
Bartunkova: 3 matches/14d, 360 min, played YESTERDAY, back_to_back 0.
  Yesterday was a TWO-set win → no 3-set-hangover haircut (that −4.5pp applies only to
  3-set wins; a 2-set win is ~neutral). Age 20 → recovers fast, soften one tier.
No extreme spot on either side. Cumulative-minutes gap is context, not an edge —
and this is exactly the spot where the market over-prices the visible "she played
yesterday" story, so it is no argument for Tauson either.

═══ NARRATIVE ═══
Let-down: Bartunkova lost the Birmingham Final 06-07 — 58 days ago, outside the window → None.
Venue momentum: Bartunkova 1-0 at Toronto (live, this week). Tauson no Toronto history in window.
Title recency: none for either.
Surface transition: neither — Tauson's last is hard (Washington); Bartunkova already made
  the clay→hard switch and won her first hard match. No transition penalty.

═══ MARKET CHOICE ═══
alt_markets.handicap: null → AH: not quoted
alt_markets.totals: null; alt_markets.model: null → no margin read available, ml only.
betfair_ml present and plausible (1.83/2.04) → executable ML is the sole priced market.
Best available edge: A ML +3.03% on ML alone, +1.57% blended, −3.58% on the sim.
No refutation check run — PASS decisions get no web access.

GATES
  G1 edge-origin: FAIL — vs the executable 1.83 the pre-narrative edge is +3.0% on the
    lone most-bullish model, +1.6% blended and NEGATIVE on the sim; the 5% full bar is
    unreachable and the 3% probe rests on one model the qualitative read contradicts.
  G2 consensus:   TIER — lone-model (ML) support only; ELO is at breakeven and the sim
    opposes. Would cap at 0.5-1.0u even if G1 cleared. (Stake tier only, never blocks.)
  G3 fatal-risk:  N/A — ML market, G3 does not block ML.

═══ VERDICT: PASS ═══
The apparent value is a stale book quote. On the exchange the market has Tauson at 52.7%
devigged, and the simulation agrees with the market rather than with ELO/ML. That leaves a
3% probe that exists only if you take the single most optimistic model at face value —
and the path read points the opposite way: a rusty #18 with one good week at a 250 in her
last three months, walking into match one at a 1000 against a 20-year-old who beat
Andreescu in straight sets on this court yesterday and made a grass final in June.
Narrative cannot create edge, but it can kill a probe, and it kills this one. The dog side
has no numeric edge at 2.04 (−2.6% to −9.2%), so there is nothing to play on either side.
No bet.