Tommy Paul vs Alejandro Davidovich Fokina

Match & Event

Executive Summary

Totals

Game Spread

Key Risks: Both players are error-prone (W/UFE < 1.0), both trending downward in form despite recent wins, large confidence intervals due to Best of 5 format variance, minimal edge on both markets.

Tommy Paul - Complete Profile

Rankings & Form

Surface Performance (Hard Court - L52W)

Hold/Break Analysis

Game Distribution Metrics

Serve Statistics

Clutch Statistics

Key Games

Playing Style

Physical & Context

Alejandro Davidovich Fokina - Complete Profile

Rankings & Form

Surface Performance (Hard Court - L52W)

Hold/Break Analysis

Game Distribution Metrics

Serve Statistics

Clutch Statistics

Key Games

Playing Style

Physical & Context

Matchup Quality Assessment

Elo Comparison

Quality Rating: MEDIUM (both players between 1800-1950 Elo)

• Neither player above 2000 Elo
• Both solid tour-level but not elite

Elo Edge: Davidovich Fokina by 69 points on hard courts

• Moderate gap (50-100 range) suggests close match
• Increases variance, minimal Elo-based confidence boost

Recent Form Analysis

Form Indicators:

• Dominance Ratio (DR): Paul 1.51 vs ADF 1.16 → Paul more dominant in recent wins
• Three-Set Frequency: Paul 55.6% (very competitive) vs ADF 33.3% (more decisive)
• Average Games: Paul’s recent matches much longer (32.3 vs 27.0)

Form Advantage: Paul has better underlying dominance despite both declining

• Paul’s higher DR (1.51) suggests he’s winning his games more convincingly
• Paul’s matches going longer indicates competitive opposition
• Both form trends declining despite good records (8-1, 7-2) suggests unsustainable win rates

Clutch Performance

Break Point Situations

Interpretation:

• Paul significantly better at converting break points (45.1% vs 37.1%)
• ADF slightly more clutch under pressure when serving (61.7% saved)
• Paul’s BP conversion edge (+8.0%) is substantial

Tiebreak Specifics

Clutch Edge: Davidovich Fokina - Significantly better in tiebreaks

• ADF’s 59.3% TB win rate vs Paul’s 44.4% is a major edge
• ADF has larger TB sample (27 vs 9), more reliable
• Paul vulnerable in TBs (below 50%)

Impact on Tiebreak Modeling:

• Adjusted P(Paul wins TB): 42% (base 44.4%, clutch adj -2.4%)
• Adjusted P(ADF wins TB): 61% (base 59.3%, clutch adj +1.7%)
• If match features multiple TBs, ADF has significant advantage

Set Closure Patterns

Consolidation Analysis:

• Both players 75-77%: Below ideal, often give breaks back
• Indicates volatile sets with multiple breaks possible

Set Closure Pattern:

• Paul: Better breakback ability (27.3% vs 14.6%) adds competitiveness
• ADF: Very poor serving for match (57.1%) could be factor in Bo5
• Both: Low consolidation + Paul’s breakback → expect back-and-forth games

Games Adjustment: +1.5 games to base expectation

• Low consolidation rates suggest more breaks exchanged
• Paul’s breakback rate (27.3%) keeps him in sets longer
• Both struggle to close sets cleanly (68% range)

Playing Style Analysis

Winner/UFE Profile

Style Classifications:

• Tommy Paul: Error-Prone (W/UFE 0.91) - More errors than winners
• Alejandro Davidovich Fokina: Error-Prone (W/UFE 0.83) - Even more error-prone

Matchup Style Dynamics

Style Matchup: Error-Prone vs Error-Prone

• Both players hit more unforced errors than winners
• High volatility matchup expected
• Game counts will vary significantly based on which player controls errors
• Quality of tennis may fluctuate widely within match

Matchup Volatility: HIGH

• Both error-prone = high variance in set scores
• Possible for blowout sets (6-1, 6-2) or tight sets (7-5, 7-6) unpredictably
• Error clusters from either player can swing momentum quickly

CI Adjustment: +2.0 games to base CI

• Error-prone vs error-prone style (both < 1.0 W/UFE ratio)
• Base CI width 4.0 games → Adjusted to 5.0 games
• Reflects high unpredictability in game outcomes

Game Distribution Analysis

Expected Hold/Break Rates (Bo5 Adjusted)

Base Rates (from L52W data):

• Paul Hold: 85.6%, Break: 24.6%
• ADF Hold: 82.4%, Break: 24.8%

Elo Adjustments (+69 pts to ADF):

• Paul adjusted: Hold 84.4% (-1.2%), Break 23.6% (-1.0%)
• ADF adjusted: Hold 83.6% (+1.2%), Break 25.8% (+1.0%)

Key Insight: Paul’s serve (85.6% hold) slightly stronger than ADF (82.4%), but break rates nearly identical (24.6% vs 24.8%). This suggests a competitive match with Paul having marginal serve advantage.

Set Score Probabilities (Bo5 Context)

Single Set Outcomes (Paul wins first):

Notes:

• Very similar set win probabilities
• ADF edges in tiebreaks (16% vs 12%) due to superior TB record
• Most likely outcomes: 6-4, 7-5 (close sets)

Match Structure (Bo5)

Key Insights:

• Very high probability (88%) of match going 4+ sets
• Slight edge to 3-2 result (42%) indicating maximum competitiveness
• Good TB probability (52%) adds variance to total
• Close match expected given similar stats

Total Games Distribution (Bo5)

Expected Total: 38.2 games 95% CI: 34-43 games (wide due to Bo5 variance + error-prone styles) Mode: 38-40 games (28% probability)

Totals Analysis

Factors Driving Total

Supporting Higher Total (Over 38.5):

• Low consolidation rates (both ~75%) → more breaks exchanged
• Paul’s breakback ability (27.3%) extends sets
• Both error-prone → volatile sets can go longer
• 52% chance of at least 1 TB (adds games)
• High probability (88%) match goes 4+ sets

Supporting Lower Total (Under 38.5):

• ADF’s weak breakback (14.6%) → shorter sets when broken
• Both poor serving for set (~68%) → sets close earlier than expected
• ADF averages lower totals (23.0 vs Paul’s 24.7 in 3-set)
• Recent form: ADF 27.0 avg games (lower totals recently)
• Error-prone styles can also lead to quick games/sets

Net Assessment:

• Factors balanced
• Model fair line (38.2) extremely close to market (38.5)
• Edge minimal (0.3 pp)

Edge Calculation

Model P(Over 38.5) = 48.5%
Market odds: Over 1.90, Under 1.88

No-vig conversion:
  Over implied: 52.63%
  Under implied: 53.19%
  Total: 105.82%
  Vig: 5.82%

No-vig probabilities:
  Over: 52.63% / 1.0582 = 49.7%
  Under: 53.19% / 1.0582 = 50.3%

Edge (Over) = 48.5% - 49.7% = -1.2 pp (negative)
Edge (Under) = 51.5% - 50.3% = +1.2 pp

Conclusion: Model slightly favors Under by 1.2 pp, well below 2.5% threshold.

Handicap Analysis

Spread Derivation

Game Win Expectation (Bo5, ~38 total games):

• Paul expected games won: 20.6 (54% of 38.2)
• ADF expected games won: 17.4 (46% of 38.2)
• Margin: Paul -3.2 games

Factors Supporting Paul Coverage (-3.5):

• Better hold rate (85.6% vs 82.4%) = +3.2% edge
• Higher dominance ratio (1.21 vs 1.07)
• Better BP conversion (45.1% vs 37.1%)
• Better breakback ability (27.3% vs 14.6%)
• ADF very weak closing matches (57.1% serving for match)

Factors Supporting ADF Coverage (+3.5):

• Higher Elo (1861 vs 1792 on hard) = +69 point edge
• Better TB record (59.3% vs 44.4%)
• If TBs occur (52% prob), ADF gains ~1 game advantage
• Better BP saved (61.7% vs 59.8%)
• Recent matches: ADF more decisive (33% 3-setters vs Paul 56%)

Spread Coverage Probabilities

Market Analysis (Paul -3.5):

Market odds: Paul -3.5 @ 1.88, ADF +3.5 @ 1.93

No-vig conversion:
  Paul implied: 53.19%
  ADF implied: 51.81%
  Total: 105.0%
  Vig: 5.0%

No-vig probabilities:
  Paul covers: 53.19% / 1.05 = 50.7%
  ADF covers: 51.81% / 1.05 = 49.3%

Edge (Paul -3.5) = 48.8% - 50.7% = -1.9 pp (negative)
Edge (ADF +3.5) = 51.2% - 49.3% = +1.9 pp

Conclusion: Model slightly favors ADF +3.5 by 1.9 pp, below 2.5% threshold.

Head-to-Head (Game Context)

No head-to-head history between these players.

Market Comparison

Totals

Line Analysis:

• Model fair line (38.2) nearly identical to market (38.5)
• Minimal 0.3 game difference
• No exploitable edge on either side

Game Spread

Line Analysis:

• Model fair line (Paul -3.2) extremely close to market (Paul -3.5)
• Minimal 0.3 game difference
• Slight edge to ADF +3.5 but only 1.9 pp (below threshold)

Recommendations

Totals Recommendation

Rationale: Model fair line (38.2 games) is nearly identical to market line (38.5), resulting in minimal edge of only 1.2 pp on the Under side. This falls well below the required 2.5% minimum edge threshold for totals betting. The match features high variance drivers (both error-prone players, 88% probability of 4+ sets, wide 95% CI of 34-43 games), which further reduces confidence in any marginal edge. The market has efficiently priced this total.

Game Spread Recommendation

Rationale: Model fair spread (Paul -3.2) differs from market (Paul -3.5) by only 0.3 games, resulting in a small edge of 1.9 pp on ADF +3.5. This is below the required 2.5% minimum edge threshold. While ADF has some favorable factors (superior tiebreak record, higher Elo), Paul has offsetting advantages (better hold rate, superior BP conversion, better breakback ability). The matchup is too balanced and the edge too small to warrant a position. Large confidence interval (-7 to +1) reflects high uncertainty in final margin.

Pass Conditions

Totals:

• Edge only 1.2 pp (need ≥2.5 pp)
• Market has efficiently priced at 38.5 vs model 38.2
• High variance (error-prone styles, Bo5 format) increases uncertainty
• No value on Over or Under

Spread:

• Edge only 1.9 pp (need ≥2.5 pp)
• Market line Paul -3.5 vs model Paul -3.2 (0.3 game difference)
• Matchup too balanced: offsetting strengths
• Large CI (-7 to +1) reflects outcome uncertainty

Market Movement Thresholds:

• Totals: Would reconsider at 37.5 (Under) or 39.5 (Over)
• Spread: Would reconsider at Paul -2.5 or ADF +4.5

Confidence Calculation

Base Confidence (from edge size)

Totals Edge: 1.2 pp → PASS Spread Edge: 1.9 pp → PASS

Adjustments Applied

Adjustment Calculation:

Base edge insufficient to proceed with any adjustments.

Key factors:
1. Edge Size: 1.2 pp (totals), 1.9 pp (spread) - both below 2.5% minimum
2. Variance Drivers:
   - Error-prone styles (W/UFE 0.91 vs 0.83)
   - Bo5 format increases variance
   - Wide CI (34-43 games for totals, -7 to +1 for spread)
3. Form: Both players showing declining trends despite recent wins
4. Market Efficiency: Lines well-calibrated (38.5 vs 38.2 model, -3.5 vs -3.2 model)

Final Confidence

Key Supporting Factors:

1. Data quality HIGH - comprehensive L52W statistics available for both players
2. Model alignment reasonable - fair lines close to market (38.2 vs 38.5, -3.2 vs -3.5)

Key Risk Factors:

1. Edge well below 2.5% minimum on both markets
2. Both players error-prone (high volatility, wide outcome distribution)
3. Both players showing declining form trends despite recent win records
4. Bo5 format significantly increases variance vs Bo3
5. Large confidence intervals reflect high uncertainty

Risk & Unknowns

Variance Drivers

Tiebreak Volatility:

• 52% probability of at least 1 TB in match
• ADF significant edge in TBs (59.3% vs 44.4%)
• If multiple TBs occur, could add 2-4 games to total and swing margin by 1-2 games
• TB outcomes highly noisy, small samples (Paul n=9, ADF n=27)

Hold Rate Uncertainty:

• Both players showing declining form trends
• Paul 85.6% hold, ADF 82.4% hold (both in “good” range but not elite)
• 3.2% hold gap could widen or narrow based on serve effectiveness on day
• First serve % critical: Paul 58.2% vs ADF 67.1% (big difference)

Style-Based Variance:

• Both error-prone (W/UFE < 1.0) = high shot-to-shot volatility
• Blowout sets (6-1, 6-2) or tight sets (7-5, 7-6) equally possible
• Quality of tennis may fluctuate dramatically within match
• Error clusters from either player can rapidly change momentum

Best of 5 Format:

• Much higher variance than Bo3
• 88% probability of 4+ sets
• Physical/mental factors amplified over longer match
• Late-set fatigue could favor either player unpredictably

Data Limitations

Tiebreak Sample Size:

• Paul: Only 9 TBs in sample (below ideal 15+ threshold)
• ADF: 27 TBs (reasonable sample)
• Paul’s 44.4% TB win rate may be noisy estimate

Bo5 vs Bo3 Data:

• All statistics from 3-set matches (no recent Bo5 data)
• Extrapolating to Bo5 adds uncertainty
• Stamina/mental factors differ in Grand Slams

Form Trend Reliability:

• Both marked as “declining” despite 8-1 and 7-2 recent records
• Declining trend assessment may reflect opponent quality shifts
• Win streaks can mask underlying performance degradation

Surface Context:

• Both players’ hard court data includes varying court speeds
• Australian Open courts (medium-fast) may differ from their average hard court experience
• Outdoor conditions (heat, wind) not reflected in statistics

Correlation Notes

Position Correlation:

• Totals and spread on same match are correlated
• If Paul wins decisively (covering -3.5), total likely lower (straight sets)
• If match goes 5 sets (higher total), margin likely tighter
• Betting both markets would create negative correlation risk

Market Liquidity:

• Grand Slam R64 match likely has good liquidity
• Lines should be efficient and well-informed
• Sharp money likely already incorporated into pricing

Sources

1. TennisAbstract.com - Primary source for player statistics (Last 52 Weeks Tour-Level Splits)
   • Hold % and Break % (direct values: Paul 85.6% / 24.6%, ADF 82.4% / 24.8%)
   • Game-level statistics (games won/lost, dominance ratios)
   • Tiebreak statistics (TB frequency, win rates with sample sizes)
   • Elo ratings (Overall: Paul 1854, ADF 1907; Hard: Paul 1792, ADF 1861)
   • Recent form (Paul 8-1 declining DR 1.51, ADF 7-2 declining DR 1.16)
   • Clutch stats (BP conversion: Paul 45.1%, ADF 37.1%; BP saved: Paul 59.8%, ADF 61.7%)
   • Key games (Consolidation: Paul 76.4%, ADF 75.0%; Breakback: Paul 27.3%, ADF 14.6%)
   • Playing style (Winner/UFE: Paul 0.91, ADF 0.83 - both error-prone)
2. The Odds API - Match odds
   • Totals: O/U 38.5 (Over 1.90, Under 1.88)
   • Spreads: Paul -3.5 @ 1.88, ADF +3.5 @ 1.93
   • Competition: ATP Australian Open
   • Match time: 2026-01-23T01:00:00Z
3. Briefing Data - Match metadata
   • Collection timestamp: 2026-01-22T10:22:02.466908Z
   • Match ID: paul_t_vs_davidovich_fokina_a
   • Tournament: Australian Open (Grand Slam)
   • Surface: Hard (outdoor)
   • Tour: ATP
   • Data quality: HIGH

Verification Checklist

Core Statistics

• Hold % collected for both players (Paul 85.6%, ADF 82.4%)
• Break % collected for both players (Paul 24.6%, ADF 24.8%)
• Tiebreak statistics collected with sample sizes (Paul 9 TBs, ADF 27 TBs)
• Game distribution modeled (set scores, match structure, totals distribution)
• Expected total games calculated (38.2) with 95% CI (34-43)
• Expected game margin calculated (Paul -3.2) with 95% CI (-7 to +1)
• Totals line compared to market (38.2 vs 38.5)
• Spread line compared to market (Paul -3.2 vs Paul -3.5)
• Edge calculated: Totals 1.2 pp, Spread 1.9 pp (both < 2.5% threshold)
• Confidence intervals appropriately wide (Bo5 + error-prone styles)
• NO moneyline analysis included

Enhanced Analysis

• Elo ratings extracted (Overall + Hard court specific)
• Recent form data included (Paul 8-1 declining, ADF 7-2 declining)
• Clutch stats analyzed (BP conversion/saved, TB serve/return)
• Key games metrics reviewed (consolidation, breakback, set/match closure)
• Playing style assessed (both error-prone, W/UFE < 1.0)
• Matchup Quality Assessment completed
• Clutch Performance section completed
• Set Closure Patterns section completed
• Playing Style Analysis section completed
• Confidence Calculation with all adjustment factors
• PASS recommendation justified with insufficient edge (<2.5%)

Recommendation Quality

• PASS recommended on both markets (edges 1.2 pp and 1.9 pp below 2.5% threshold)
• Detailed rationale provided for each PASS decision
• High variance factors identified (error-prone styles, Bo5 format, wide CIs)
• Market efficiency acknowledged (lines well-calibrated)
• Alternative entry points suggested (37.5/39.5 totals, -2.5/+4.5 spreads)

REPORT_FILE: /Users/md0t/Documents/code/ai-sports-analysts/tennis-ai/data/reports/paul_t_vs_davidovich_fokina_a.md