TennisEdge's AI model finds no betting value in S. Shimabukuro vs N. Budkov Kjaer (Montreal on 2026-08-01) — the market is efficient, with S. Shimabukuro favoured at 1.66.
Independent estimates of who wins, each from a different method. The blend is what the bot prices against the market.
| Method | S. Shimabukuro | N. Budkov Kjaer |
|---|---|---|
| Elo ratings-based, surface-weighted |
40.8% | 59.2% |
| Machine learning 29-feature gradient-boosted model |
52.5% | 47.5% |
| Serve / return point-by-point simulation |
39.9% | 60.1% |
| Blend what the bot actually prices off |
47.9% | 52.1% |
| Market implied de-vigged from the opening odds |
55.9% | 44.1% |
We publish how well these score against the closing market, including where they lose: how accurate are tennis predictions?
| Player | Surface | Elo | Record | Matches |
|---|---|---|---|---|
| S. Shimabukuro | hard | 1657 | 165–116 | 281 |
| N. Budkov Kjaer | clay | 1628 | 52–30 | 82 |
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. 4, 2026.
S. Shimabukuro leads 1–0 over 1 meeting. Full record.
| Date | Event | Surface | Winner | Score |
|---|---|---|---|---|
| Aug. 1, 2026 | Montreal | hard | S. Shimabukuro | 2 - 0 |
| L | M. Cilic | 0 - 2 | Aug. 4, 2026 |
| W | N. Budkov Kjaer | 2 - 0 | Aug. 1, 2026 |
| L | R. Pacheco Mendez | 1 - 2 | July 29, 2026 |
| L | J. Faria | 1 - 3 | June 30, 2026 |
| L | F. Tiafoe | 2 - 0 | June 17, 2026 |
| L | S. Shimabukuro | 2 - 0 | Aug. 1, 2026 |
| L | S. Gueymard Wayenburg | 0 - 2 | July 21, 2026 |
| L | A. Pellegrino | 0 - 2 | July 14, 2026 |
| L | O. Virtanen | 1 - 3 | June 25, 2026 |
| W | M. Dodig | 2 - 0 | June 24, 2026 |
All Aug. 1, 2026 results & picks →
═══ S. Shimabukuro vs N. Budkov Kjaer — Montreal (hard), round labeled "Final" ═══
DATA CAVEAT: labeled a Montreal Masters Final between #147 (Shimabukuro) and #124
(Budkov Kjaer), but NEITHER player has any Montreal matches in 90d history. Venue/path
context is therefore unavailable; analysis rests on form + models only.
MODELS vs MARKET
ELO prob A = 0.408 (ELO favors B ~59%)
ML prob A = 0.525 (ML slightly favors A)
Blend prob A = 0.479 (near coinflip, marginal B)
Market: odds_a 1.66 (impl 0.558), odds_b 2.10 (impl 0.476); value_pick B, edge 0.079.
ELO and ML DISAGREE with each other; ML sits on top of the market on A; the blend is
a coinflip. No model-market consensus, but no clean single-model conviction either.
ELO/RANK SANITY
B ranked 124 yet ELO-favourite — rating inflated by a deep grass/clay run (Wimbledon
final, Valencia title, French semi, Rome final) against mostly 1500-1750 fields. His
LAST TWO results are losses (clay). ELO overstates current level → discount the B edge.
A ranked 147, ELO underdog, consistent with weak current form. No credible edge to lean.
RECENT FORM (last 5 each)
A Shimabukuro (29): L Los Cabos v Pacheco Mendez 1327 (Weak) 7-6 5-7 2-6 — lost to a
weak player after a ~month layoff; L Wimbledon v Faria; L Halle v Tiafoe; W Halle v
Griekspoor; L Stuttgart v Shelton. Peak was Stuttgart grass title (early June).
Shape: FADING (loss to 1327 in 3 sets, off long layoff).
B Budkov Kjaer (19): L Tampere v Gueymard Wayenburg 1516 6-7 3-6; L Bastad v Pellegrino
3-6 2-6 (blowout); L Wimbledon FINAL v Virtanen; W Wimbledon SF v Dodig; W Wimbledon
QF v Reis Da Silva. Big body of work but micro-form = 2 straight clay losses.
Shape: FADING recently (2 losses, one a blowout).
SURFACE TRANSITION
B's last two matches are CLAY (Tampere, Bastad); this match is HARD → TRANSITION for B
(depressed level + wider margin variance). A is already on hard (Los Cabos). Transition
is a RISK on the side (B) the value points to — cuts against backing B, not for it.
FATIGUE (Step 2 calibration)
A: last match 07-29 (3d ago), 158min 3-setter but LOST it → no 3-set-hangover haircut
(applies only to yesterday's WINNERS, and it was 3d ago regardless). Age 29 baseline.
B: last match 07-21 (11d ago) → mild rust risk (no on-court reps for 11 days), age 19
recovers fast. Neither side hits an extreme fatigue spot; treat as roughly equal.
NARRATIVE
Neither reached a Final in the printed window at THIS event (no venue data). B lost the
Wimbledon final 06-25 (>3wks ago, no live let-down). No motivation signal in the data.
GATES
G1 edge-origin: FAIL — book-priced only (no betfair_ml); best B edge is blend 0.521 −
1/2.10 = 4.4% (value_edge 7.9% is optimistic), both BELOW the 8% book-priced bar, and
the qualitative read gives B a TRANSITION RISK, not confirmation. No pre-narrative edge
clears the bar.
G2 consensus: PASS (stake-tier only; never blocks) — would be lone/contaminated (ELO)
→ capped tier, but moot given G1.
G3 fatal-risk: n/a (ML market; no margin bet).
PRICING
No exchange price (alt_markets null). Book-priced ML requires edge >= 8% vs book AND a
flag — the segment the backtest loses on. B edge does not reach 8%; A has no edge (ML =
market). min_odds for B at ~52% = 1.05/0.52 = 2.02; book 2.10 clears min-odds but the
edge is too thin and book-only.
VERDICT: PASS. Models disagree without a clean, articulable qualitative edge; the value
side (B) carries a clay->hard transition risk and 2 straight losses; ELO on B looks
inflated and is discounted; no exchange liquidity and sub-8% book edge. Nothing to stake.