E. Svitolina vs J. Bouzas Maneiro — Prediction & Analysis

TennisEdge's AI model finds no betting value in E. Svitolina vs J. Bouzas Maneiro (Toronto on 2026-08-04) — the market is efficient, with E. Svitolina favoured at 1.19.

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 31, 2026.

Head to head

E. Svitolina leads 3–0 over 3 meetings. Full record.

Recent form

All Aug. 4, 2026 results & picks →

Full analysis

E. SVITOLINA vs J. BOUZAS MANEIRO — WTA Toronto, R2 (1/32), hard

MODELS vs MARKET (player A = Svitolina)
  ELO         86.9%  |  ML 75.6%  |  sim (resimulate) 66.9%
  Betfair ML  1.20 / 5.20  -> devigged A 81.3%
  Pinnacle    1.194 / 5.10 -> no-vig A 81.0% (fetched 2026-08-03 22:55Z, ~fresh)
  Book quote  1.19 / 5.10, value_pick A +5.8% (vs book, not executable)
  Model spread is unusually wide (67 / 76 / 87) and the market sits mid-cluster,
  with Pinnacle and Betfair agreeing to within 0.3pp. That is the signature of an
  efficiently priced match with noisy models, not of a mispricing.

ELO/RANK SANITY
  Svitolina ELO-favourite AND ranked #13 -> agreement, ELO credible (not farmed).
  Bouzas #51, age 23. No divergence to discount on either side.

RECENT FORM (max 5 shown; all fetched matches used for classification)
  Svitolina (31, #13) — hard-court block is modest
   07-31 Washington QF  L  vs A. Eala (1850, Elite)        3-6 4-6  straights
   07-31 Washington R16 W  vs P. Kudermetova (1676, Strong) 7-6 6-4  GRIND
   06-30 Wimbledon R64  L  vs D. Snigur (1648, Avg)        5-7 2-6  poor
   06-25 Bad Homburg QF L  vs Xin. Wang (1500, Weak)       walkover/void
   06-23 Bad Homburg R16 W vs L. Samsonova (1788, Strong)  6-3 3-6 6-2 GRIND
   Peak was CLAY: Rome title 05-16 (d. Rybakina, Swiatek, Gauff) + RG QF, ~80d ago
   — too old to be a let-down spot and not evidence for current hard-court level.
  Bouzas Maneiro (23, #51)
   08-02 Toronto R64    W  vs A. Arseneault (1459, Weak)   6-2 6-4  DOM (vs Weak)
   07-20 Prague R16     L  vs M. Bouzkova (1838, Strong)   5-7 5-7  competitive
   07-13 Athens quali   L  vs Q. Zheng (1500, Weak)        5-7 1-6
   07-03 Wimbledon R32  L  vs J. Pegula (1883, Elite)      1-6 3-6  blowout
   07-01 Wimbledon R64  W  vs D. Yastremska (1684, Strong) 6-3 6-7 6-2 GRIND

SHAPE / PATH (Step 1b)
  Svitolina: STABLE-to-FADING on hard — a tiebreak grind past a 1676 and a
  straight-sets loss to an Elite; the Wimbledon R64 loss to a 1648 is the ugliest
  data point. Path: FIRST match at Toronto (mild rust risk), 4 days since
  Washington. No surface transition (came off hard).
  Bouzas: STABLE. One venue match already (match-sharp, conditions-adjusted,
  clay->hard transition already resolved by the R1 win). But her record vs top-20
  opposition is a run of wide losses (Pegula -8 games, Keys -11, Sabalenka -6,
  Tomljanovic -8) — she stays close to Strong, not to Elite.

FATIGUE (Step 2)
  A: 2 matches/14d, 159 min, last 07-31 (4d rest), back_to_back 0.
  B: 2 matches/14d, 177 min, last 08-02 (2d rest), back_to_back 0.
  No extreme spot on either side; no 3-set win yesterday -> no -4.5pp haircut
  applies. Svitolina is 31 (veteran, slower recovery) but 4 days of rest neutralises
  it. Cumulative-minutes gap is noise per calibration -> roughly EQUAL. Fatigue
  plays no part in this decision.

NARRATIVE
  Let-down: None (Rome final won 05-16, ~80 days — outside the window).
  Venue momentum: Bouzas 1-0 at Toronto this week; Svitolina yet to play here.
  Title recency: TITLE Rome 80d ago (clay) — does not transfer as live hard form.
  Surface transition: neither player materially transitioning (B resolved it 08-02).

SIM VALIDATION & BOUNDED RE-RUNS (Step 4b)
  Baseline resimulate: A win 66.9%, exp. games 20.5, exp. margin 2.1;
  B +5.5 covers 73.3% (market 55.6% -> nominal edge +17.7%). Tempting, and wrong.
  Convergence test FAILS: the sim is 20pp below ELO and 14pp below two independent
  sharp prices. Input inspection shows why — the SPW/RPW are opponent-unadjusted:
   - Bouzas' 50-match window is weak-field heavy (Parma WTA125, Rabat, Athens
     qualifying; opponent ELOs 1334/1441/1459/1503/1505/1595) -> her spw 0.570 is
     inflated.
   - Svitolina's window is elite-loaded (Swiatek 2058, Gauff 2071, Rybakina 2016,
     Kostyuk 1982, Bencic 1840, Eala 1850/1887) plus a bad grass block -> her
     spw 0.586 is depressed.
  Both biases push the SAME way: they overstate B's competitiveness and the +5.5
  cover. Bounded re-runs (each override <=0.05, logged):
   - spw_b 0.570 -> 0.545 (weak-field correction): A win 78.0%, B +5.5 covers 61.1%
   - spw_a 0.586 -> 0.606 (elite-opposition correction): A win 75.9%, B +5.5 62.5%
   - both together:                                    A win 85.1%, B +5.5 52.9%
  The margin distribution is not robust: a defensible 2.5-point input correction
  swings the B +5.5 cover from 73% to 53%. The headline handicap edge is an
  artifact of unadjusted inputs, not information. Corrected win prob (78-85%)
  converges with ELO/ML/market — so the coherent distribution says the market is
  right and there is no ML edge either.

MARKET CHOICE
  ML A (Svitolina): need >84.0% at Betfair 1.20 (eff. 1.19); best coherent estimate
    ~80-82%. NEGATIVE edge. Pass.
  ML B (Bouzas): my estimate ~20% (ELO 13, ML 24, corrected sim 15-22, market 19).
    Betfair 5.20 -> eff. 4.99 = 20.0% breakeven; min odds 1.05/0.20 = 5.25 > 5.20.
    Edge ~0%. Automatic PASS on the min-odds rule.
  Handicap: AH: B +5.5 @ 1.80 (pass) / A -5.5 @ 2.12 (sim 26.7% vs 47.2% implied,
    heavily negative). B +5.5 is |line| >= 4.5 -> HIGH-VARIANCE tier, needs edge
    >= 10% and min odds 1.10/0.61 = 1.80. On corrected inputs the cover is 53-61%
    vs 55.6% implied -> edge -2.6% to +5.5%, fails the wide-AH bar outright.
  Total 20.0: corrected expected games 18.4-19.8 -> Over ~45-48% vs 50.0% implied
    (Under ~52-55% vs 52.1% implied). No edge on either side; and the estimate is
    as input-sensitive as the handicap. Pass.

GATES
  G1 edge-origin: FAIL — no market beats its bar on the pre-narrative numbers:
    ML A negative, ML B ~0% (below min odds 5.25), B +5.5 -2.6/+5.5% vs a 10% bar,
    total ~2%. The only "edge" on the board came from a sim whose inputs are
    demonstrably contaminated by field quality.
  G2 consensus:   stake-tier only — moot; would be capped anyway (lone contaminated
    signal, the sim, against ELO + ML + two sharp prices).
  G3 fatal-risk:  FAIL for the margin bet — backing B +5.5 requires her to stay
    within 5 games of a top-20 player, but her top-20 losses are 1-6 3-6 (Pegula),
    0-6 1-6 (Keys), 4-6 2-6 (Sabalenka), 3-6 1-6 (Tomljanovic). The cover thesis is
    self-negating; units never buy back a negating risk on a margin bet.

VERDICT: PASS
  Betfair and Pinnacle agree within 0.3pp, ELO is above the market and ML below it,
  and the one model screaming value (the sim) collapses under a justified 2.5-point
  input correction. Nothing here is information the market is slow to price:
  Bouzas being match-sharp and Svitolina playing her first match are both small,
  known, and already inside a 5.20. No stake.