AI verdict
Model consensus and the Glicko-2 rating lean to Miyoshi K. Full breakdown below.
Picks locked before kickoff results verified after public accuracy log →
| Herbert P-H | Miyoshi K | |
|---|---|---|
| Win % | 52.4% | 62.9% |
| Hard % | 56.8% | 63% |
| Clay % | 42.6% | 60% |
| Grass % | 54.8% | 0% |
| Deciding set % | 54.9% | 52.8% |
| Tie-breaks % | 47.9% | 42.9% |
| Titles (total) | 2 | 2 |
| Last 10 | LLLLWWLWWW | LWWWLWWWWL |
Source: scorego-app.com
| Herbert P-H | Miyoshi K | |
|---|---|---|
| Aces per match | 8.00 | 5.53 |
| Double faults per match | 4.69 | 1.84 |
| Breaks per match | 1.92 | 10 |
Source:tennisstats.com (per-match averages)
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| Win Percentage 2026 Calendar Year | 60.0% 9 / 15 Matches | 75.0% 27 / 36 Matches |
| Win Percentage Last 12 Months | 50.0% 13 / 26 Matches | 70.5% 31 / 44 Matches |
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| Win Percentage 2026 Calendar Year | 37.5% 3 / 8 Matches | N/A 0 / 0 Matches |
| Win Percentage Last 12 Months | 30.0% 3 / 10 Matches | N/A 0 / 0 Matches |
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| Win Percentage 2026 Calendar Year | 66.7% 2 / 3 Matches | N/A 0 / 0 Matches |
| Win Percentage Last 12 Months | 66.7% 2 / 3 Matches | N/A 0 / 0 Matches |
Source:tennisstats.com
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| H2H wins | 0 (0%) | 0 (0%) |
| Sets | 0 (0%) | 0 (0%) |
| Wins this year | 50.0% 14 / 28 Matches | 75.0% 27 / 36 Matches |
| Wins in 12 mo. | 43.9% 18 / 41 Matches | 70.5% 31 / 44 Matches |
| Current rank | 287 | 495 |
Source: tennisstats.com
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| Prize money (career) | $9,511,808 USD | $11,954 USD |
| Rating | 287 | 495 |
| Age | 35 | 22 |
| Height | 1.88m | 1.85m |
| Weight | 75kg | 70kg |
| Playing hand | Right-handed | Right-handed |
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| Match Wins | 50 % | 75 % |
| Match Wins (3 Sets) | 54 % | 75 % |
| Match Wins (5 Sets) | 0 % | N/A |
| Wins in Straight Sets | 29 % | 58 % |
| Wins From Behind | 13 % | 31 % |
| Won At Least 1 Set | 68 % | 86 % |
| Set 1 Win | 46 % | 75 % |
| Set 2 Win | 50 % | 69 % |
Source:tennisstats.com
| Pierre-Hugues Herbert | Kenta Miyoshi | |
|---|---|---|
| Aces | 2 | 1 |
| Double Faults | 9 | 7 |
| 1st serve percentage | 52% | 43% |
| 1st serve points won | 71% (32/45) | 60% (18/30) |
| 2nd serve points won | 39% (16/41) | 55% (22/40) |
| Break Points Saved | 5/8 | 4/8 |
| 1st return points won | 50% (15/30) | 38% (17/45) |
| 2nd return points won | 45% (18/40) | 61% (25/41) |
| Break Points Converted | 4/8 | 3/8 |
| Net points won | 0% (0/0) | 0% (0/0) |
| Service Points Won | 57% (48/84) | 57% (40/70) |
| Return Points Won | 43% (30/70) | 43% (36/84) |
| Total Points Won | 51% (78/154) | 49% (76/154) |
| Last 10 balls | 3 | 7 |
| Match points saved | 0 | 0 |
| Service games won | 75% (9/12) | 64% (7/11) |
| Return games won | 36% (4/11) | 25% (3/12) |
| Total games won | 57% (13/23) | 43% (10/23) |
Source:tennisstats.com
Outcome percentages for totals and props: values for each player and a match-total column.
| Pierre-Hugues Herbert | Kenta Miyoshi | Match total | |
|---|---|---|---|
| Average Games in a Set | 10.57 | 9.33 | 9.95 |
| Over 6.5 | 96 % | 94 % | 95 % |
| Over 7.5 | 96 % | 83 % | 90 % |
| Over 8.5 | 92 % | 56 % | 74 % |
| Over 9.5 | 73 % | 33 % | 53 % |
| Over 10.5 | 35 % | 22 % | 29 % |
| Over 11.5 | 35 % | 22 % | 29 % |
| Pierre-Hugues Herbert | Kenta Miyoshi | Match total | |
|---|---|---|---|
| Aces Per Match | 8.00 | 5.53 | 13.53 |
| Most Aces in a match | 77 % | 89 % | N/A |
| 1+ | 100 % | 100 % | 100 % |
| 2+ | 100 % | 89 % | 98 % |
| 3+ | 100 % | 68 % | 92 % |
| 4+ | 96 % | 63 % | 90 % |
| 5+ | 88 % | 53 % | 84 % |
| Pierre-Hugues Herbert | Kenta Miyoshi | Match total | |
|---|---|---|---|
| Breaks Per Match | 1.92 | 10 | 11.92 |
| Over 0.5 | 96 % | 100 % | 100 % |
| Over 1.5 | 38 % | 100 % | 100 % |
| Over 2.5 | 27 % | 100 % | 96 % |
| Over 3.5 | 8 % | 100 % | 77 % |
| Over 4.5 | 4 % | 100 % | 60 % |
| Over 5.5 | 0 % | 100 % | 56 % |
| Pierre-Hugues Herbert | Kenta Miyoshi | Match total | |
|---|---|---|---|
| Tie Breaks Per Match | 0.42 | 0.31 | 0.36 |
| Over 0.5 | 35 % | 31 % | 33 % |
| Over 1.5 | 4 % | 0 % | 2 % |
| Under 0.5 | 65 % | 69 % | 33 % |
| Under 1.5 | 96 % | 100 % | 2 % |
| Pierre-Hugues Herbert | Kenta Miyoshi | Match total | |
|---|---|---|---|
| Double Faults Per Match | 4.69 | 1.84 | 6.53 |
| Over 0.5 (1+) | 100 % | 53 % | 100 % |
| Over 1.5 (2+) | 92 % | 42 % | 95 % |
| Over 2.5 (3+) | 73 % | 32 % | 85 % |
| Over 3.5 (4+) | 46 % | 16 % | 69 % |
| Over 4.5 (5+) | 35 % | 11 % | 58 % |
| Over 5.5 (6+) | 31 % | 5 % | 42 % |
| Pierre-Hugues Herbert | Kenta Miyoshi | Match total | |
|---|---|---|---|
| Net Points Won Per Match | 9.43 | N/A | N/A |
| Over 4.5 (5+) | 86 % | N/A | N/A |
| Over 5.5 (6+) | 86 % | N/A | N/A |
| Over 6.5 (7+) | 82 % | N/A | N/A |
| Over 7.5 (8+) | 77 % | N/A | N/A |
| Over 8.5 (9+) | 68 % | N/A | N/A |
| Over 9.5 (10+) | 55 % | N/A | N/A |
Source:tennisstats.com
Herbert P-H take on Miyoshi K on August 22, 2026 in the ATP. Uinston-Seylem, International.
Recent form (W-D-L): Herbert P-H 1-0-4, Miyoshi K 3-0-2. PropickAI's Glicko-2 model gives the home side 45% to win. Read on for detailed stats, previous meetings, the bookmaker line and the PropickAI neural-network verdict.
Our model prediction (Glicko-2)
2 Miyoshi K — 55%
Model estimate: win Herbert P-H 45.1%, win Miyoshi K 54.9%. Model favourite — Miyoshi K.
Informational estimate, not a betting recommendation.
Match winner — Herbert P-H
How often this pick wins on our settled history — not a profit forecast.
Math-model prediction
Based on the collected prematch data, the favourite looks like Herbert P-H.
For reference
Odds source: consensus of 4 books (Fonbet, Kambi BetMGM, Kambi Unibet, Leon). This is the fair probability after removing the bookmaker margin from the odds — reference information, not a prediction or a betting recommendation.
For reference
↓ odds shortened (money on this outcome), ↑ drifted (line weakened). For reference, not a betting recommendation.
Model consensus
Modeli dayut edinyy signal. Takie matchi my pomechaem kak bolee silnye.
Market consensus
Consensus probability with the bookmaker margin removed; the number on the right is the best available odds. For reference only, not a bet.
AI overview
AI analysis:1/7 ready · Still preparing:DeepSeek, ChatGPT, Claude, Qwen, Kimi, GLM 5.2
Pierre-Hugues Herbert is scheduled to play Kenta Miyoshi in the first qualifying round of the ATP 250 Winston-Salem Open on August 22, 2026, at 18:00 MSK.
Current Form and Head-to-Head
Pierre-Hugues Herbert enters this match in search of consistency on hard courts, holding a recent record of a single loss (L) based on the baseline indicators. The 35-year-old French veteran relies heavily on his experienced serve-and-volley style, though his singles ranking has slipped to No. 287. Conversely, 22-year-old Kenta Miyoshi from Japan enters with much higher competitive match rhythm, boasting a 3-2 record over his last five professional matches (LWWLW) primarily on the ATP Challenger circuit. Miyoshi has reached a career-high ranking of No. 495 this season and is highly motivated to make his mark at the ATP main tour level. This will be the first professional head-to-head (H2H) meeting between the two players.
Injury and Motivation Analysis
There are no fresh injuries or physical limitations reported for either player ahead of this qualifying match. Motivation will be exceptionally high for both. For Herbert, the hard-court season in North America represents a final opportunity to leverage his extensive experience and secure a spot in a main draw. For Miyoshi, transitioning onto the faster outdoor hard courts of Winston-Salem is an invaluable test against top-tier opposition, and his recent deep runs in Challenger tournaments indicate excellent baseline physical conditioning and match sharpness.
Probability Assessment and Line Discrepancies
The baseline market-calibrated Glicko model values this matchup closely, assigning a 51% win probability to Pierre-Hugues Herbert and 49% to Kenta Miyoshi. Our analytical estimate aligns identically with these calculations: P1 (Herbert win) at 51% and P2 (Miyoshi win) at 49%. While Miyoshi has a clear advantage in recent singles match fitness and youthful coverage, Herbert’s elite world-class service placement and tactical court positioning on quick outdoor hard courts perfectly balance out the Japanese player's baseline momentum. Consequently, a razor-thin margin is expected with no significant divergence from the underlying model parameters.
Additional Markets and Set Betting
Given how evenly matched both competitors project on a fast hard court, a highly competitive two-set battle or a full three-set layout is heavily anticipated. A total games line of Over 22.5 serves as the most viable approach, as both players possess strong holds and tiebreak tendencies. The most probable exact score by sets is a narrow 2:1 victory in favor of Pierre-Hugues Herbert, leaning on his veteran composure during the critical points of a deciding set.
Conclusion
The primary recommendation for this match is to target the total games market, specifically backing the game total to exceed the line due to the highly balanced skillsets and severe contrast in playing styles.
| AI | 1 | 2 | Score | Total games | Bet |
|---|---|---|---|---|---|
| Google AI | 51% | 49% | 2:1 | Over 22.5 | Total Games Over 22.5 |
Get the 7-AI consensus, math-model probabilities, value picks and full reports for every match with Full Access.
Play free: you start with $10,000 in virtual money. Bet on this match at our odds, then track your profit and climb the global leaderboard. No real money, no sign-up.
V baze uchteny nedavnie matchi oboikh uchastnikov.
We mark in green only strong signals where the model passed our publication filter. The other rows are for audit and training.
Tennis math model. Tennis without draws: the base signal comes from player rating, surface, form, serve/return and tournament fatigue. For live logic the set state and who is serving matter more than the overall points score.
Glicko-2 is a team-strength rating: it weighs the line, history and consistency, but is not a guarantee.
Model = Glicko + Elo strength prior (heuristic, historical calibration in progress). Glicko45.1% · +Elo prior (weight0.52).
Tennis without draws: the base signal comes from player rating, surface, form, serve/return and tournament fatigue. For live logic the set state and who is serving matter more than the overall points score.
Betting notes are built from the line, market and our math model: take a signal only when the model and market agree.
| Outcome | Fonbet |
|---|---|
| 1.55 64.5% | |
| 1.83 54.6% | |
| 1.73 57.8% | |
| 1.25 80.0% | |
| 1.04 96.2% | |
| 1.72 58.1% | |
| 1.90 52.6% | |
| 1.78 56.2% | |
| 1.90 52.6% | |
| 1.75 57.1% | |
| 2.00 50.0% | |
| 2.30 43.5% | |
| 1.87 53.5% | |
| 2.00 50.0% | |
| 3.50 28.6% | |
| 7.70 13.0% | |
| 2.00 50.0% | |
| 1.80 55.6% |
| Outcome | Fonbet |
|---|---|
| 1.01 99.0% | |
| 10.50 9.5% | |
| 1.01 99.0% | |
| 12.00 8.3% |
| Outcome | Fonbet |
|---|---|
| Over 17.5 | 1.73 57.8% |
| Over 18.5 | 2.02 49.5% |
| Over 20.5 | 2.10 47.6% |
| Over 21.5 | 2.02 49.5% |
| Over 22.5 | 2.02 49.5% |
| Over 23.5 | 1.92 52.1% |
| Over 24.5 | 1.70 58.8% |
| Over 25.5 | 1.55 64.5% |
| Over 26.5 | 2.00 50.0% |
| Over 27.5 | 7.70 13.0% |
| Over 28.5 | 3.60 27.8% |
| Over 30.5 | 2.95 33.9% |
| Under 17.5 | 2.00 50.0% |
| Under 18.5 | 1.70 58.8% |
| Under 20.5 | 1.65 60.6% |
| Under 21.5 | 1.70 58.8% |
| Under 22.5 | 1.70 58.8% |
| Under 23.5 | 1.78 56.2% |
| Outcome | Fonbet |
|---|---|
| Over 2.5 | 1.01 99.0% |
| Under 2.5 | 10.50 9.5% |
| Bookmaker | 1 | 2 |
|---|---|---|
| Pari | 2.02 | 1.75 |
| Winline | 2.03 | 1.77 |
| Fonbet | 2.02 | 1.75 |
| Betcity | 2 | 1.75 |
Herbert P-H
Miyoshi K
The Herbert P-H - Miyoshi K breakdown — outcomes, probabilities, risks and history — is available after a free sign-in.
By the PropickAI model, Herbert P-H is favoured — about 55% win probability.
The consensus of 7 AI models and the Glicko-2 rating leans to Herbert P-H. Full probabilities and bookmaker odds are shown on this page.
August 22, 2026.