W. Falkowska vs N. Noha Akugue — Prediction & Analysis

TennisEdge's AI model finds no betting value in W. Falkowska vs N. Noha Akugue (Warsaw on 2026-08-05) — the market is efficient, with N. Noha Akugue favoured at 1.40.

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 9, 2026.

Recent form

All Aug. 5, 2026 results & picks →

Full analysis

W. Falkowska (A) vs N. Noha Akugue (B) — Warsaw, hard, 1/8-finals

MODELS vs MARKET
  Book: A 2.75 / B 1.40 (overround 1.078) -> devigged A 33.7% / B 66.3%
  ELO:  A 23.5%  (B 76.5%) -> leans B, edge vs book quote (1/1.40 = 71.4%) = +5.1%
  ML:   A 42.71%           -> leans A, edge vs book quote (1/2.75 = 36.4%) = +6.3%
  Model composite: A 23.5%
  The two models straddle the market and point at OPPOSITE sides — a 19pp
  ELO-vs-ML split. There is no single coherent numeric edge to stake.

RANKING / ELO SANITY
  A: rank 504, age 26. B: rank 296, age 22. ELO agrees with ranking direction
  (Akugue higher ranked AND higher ELO) -> ELO is credible here.
  The ML number is the suspect one: Falkowska's inputs are built on ITF W35/W50
  clay fields (opponent ELOs 1055-1607, mostly 1200-1500). Her 42.7% is classic
  weak-field stat inflation. Discount the ML side, do not stake it.

RECENT FORM (max 5 shown; all fetched matches used for classification)
  A — W. Falkowska (ITF W35/W50 level, mostly clay)
   08-03 Warsaw R32   W vs A. Paszun (1293, Weak)      6-0 6-2   DOM
   07-29 W50 Dublin   L vs C. Noel (1500, Avg)         4-6 6-3 7-6  GRIND loss
   07-15 Istanbul2 Q  L vs H. Vandewinkel (1500)       3 sets
   07-14 Istanbul2 Q  W vs D. Dilek (1500)             7-5 6-7 6-2  GRIND (TB)
   07-13 Istanbul2 Q  W vs V. Panshina (1500)          6-4 6-4   DOM (QUALI final, not a title)
  B — N. Noha Akugue (WTA main-tour / 125 level)
   08-03 Warsaw R32   W vs J. Avdeeva (1644, Avg-top)  6-2 6-3   DOM
   07-22 Hamburg R16  L vs A. Bondar (1791, Strong)    1-6 6-3 5-7  competitive loss
   07-21 Hamburg R32  W vs V. Steiner (1530, Avg)      6-3 2-6 6-2  GRIND
   07-17 Rome QF      L vs F. Jones (1714, Strong)     3-6 1-6   DOM loss
   07-15 Rome R16     W vs D. Astakhova (1651, Strong) 5-7 6-2 6-3  GRIND over Strong

STEP 1b — SHAPE / PATH
  A shape: STABLE at her own level. The 6-0 6-2 is a DOM over a 1293 — exactly
    what her level predicts, not over-performance. Against 1500-grade opponents
    she is going three sets and losing (Dublin 7-6 in the 3rd, Istanbul quali).
    Judged against the 1650+ bracket she meets today, that is UNDER-performing.
  B shape: SHARPENING — 6-2 6-3 over a 1644 yesterday is a clear beat of bracket,
    on the back of a competitive 3-set loss to Bondar (1791) and a 3-set win over
    Astakhova (1651). Level of opposition is 150-400 ELO above A's entire window.
  Path: BOTH players have exactly 1 match at Warsaw this week (both won R32
    yesterday). No sharpness asymmetry, no rust on either side.
  Level gap is the real story: A's 90d field ELO ~1200-1600, B's ~1470-1790.
    That is a genuine class edge to B — but it is already what the 1.40 prices.

STEP 2 — FATIGUE
  A: 2 matches/14d, 174 min, back_to_back 0, last 08-03 (straight sets).
  B: 3 matches/14d, 305 min, back_to_back 1, last 08-03 (straight sets, 2 sets).
  Neither played a 3-set match yesterday -> the calibrated -4.5pp 3-set-hangover
  haircut does NOT apply to either side. No extreme spot (no same-day double, no
  3rd match in 3 days, no 150min+ match yesterday). B's larger cumulative minutes
  is CONTEXT ONLY per the 07-27 calibration and is further softened by age 22
  (recovers fast). Verdict: roughly equal. Fatigue plays no part in this decision.

STEP 3 — NARRATIVE
  Let-down: A's most-recent "Final" is 07-13 Istanbul 2 QUALIFYING final (22d) —
    not a main-draw title and outside any let-down window. B: no Final. -> None.
  Venue momentum: neither player has prior Warsaw history in the window.
  Title recency: A won W35 Amstelveen 07-05 (clay, 30d) — a W35 title does not
    transfer as evidence against a rank-296 opponent on hard.
  Surface transition: B came off clay (Hamburg 07-22) but has ALREADY played and
    dominated one hard match here (08-03), so the transition is spent. A is on
    hard for a 3rd straight event. No usable transition signal either way.

MARKET CHOICE
  alt_markets: none at analysis time
  handicap: AH: not quoted
  totals:   not quoted
  betfair_ml: absent -> no executable exchange price. Book-only path applies:
    an ML bet requires edge >= 8% vs the BOOK quote.
      B @ 1.40: best supporting model (ELO 76.5%) gives +5.1% vs the 1/1.40
        quote. FAILS the 8% book-only bar. My own honest read of B is ~70-73%
        (class edge real, but A won her opener and the surface is neutral) —
        at 70-73% the edge vs 71.4% is ~0 to negative, and min_odds at 0.72
        would be 1.458 > 1.40 available. Fails on price too.
      A @ 2.75: ML 42.7% gives +6.3% vs the 1/2.75 quote. Also fails the 8%
        bar, AND the supporting model is the contaminated weak-field one, AND
        min_odds at 42.7% is 2.46 — the only test it passes. Not stakeable on a
        signal I have just argued is inflated.
    Pinnacle: null — no sharp corroboration available for either side.
  No probe tier: the probe requires an EXECUTABLE exchange price (betfair_ml),
  which is absent. Book-only quotes get no probe.

GATES
  G1 edge-origin: FAIL — no exchange price; best pre-narrative book edge is +5.1% (B, ELO-only) vs the required 8% book-only bar, and ML points the other way.
  G2 consensus:   PASS (stake-tier only) — would be lone-model/capped anyway; ELO and ML disagree by 19pp.
  G3 fatal-risk:  PASS (n/a — ML market, G3 does not block; no margin bet considered).

VERDICT — PASS
  The class gap to Akugue is real (rank 296 vs 504, 1650-1790 opposition vs
  1200-1550), but 1.40 already prices it and then some: the ONLY model backing
  her clears just 5.1% over the raw book quote, below the 8% no-exchange bar, and
  my honest 70-73% read sits at or under the 71.4% implied. The ML's 42.7% for
  Falkowska is weak-field inflation, not a live dog edge. Two models pointing
  opposite ways, no exchange liquidity, no fatigue or narrative asymmetry — this
  is the small-event, book-priced segment the backtest loses on. No bet.