Fernandez L. vs Tjen J.

Match & Event

Executive Summary

Totals

Game Spread

Key Risks: CRITICAL DATA LIMITATION - Player 1 (Fernandez L.) statistics are MISSING/INVALID. Data scraper matched to retired player “Mary Joe Fernandez” with 0 matches played. Cannot perform hold/break analysis without valid Player 1 data.

CRITICAL DATA QUALITY ISSUE

DATA COLLECTION ERROR DETECTED:

The briefing data for Player 1 (Fernandez L.) is completely invalid:

• Player matched: Mary Joe Fernandez (RETIRED PLAYER)
• Matches played: 0
• Hold %: 0%
• Break %: 0%
• All statistics: 0 or missing

Root Cause: The data scraper likely matched “Fernandez L.” to the wrong player in the database. Mary Joe Fernandez is a retired American player (career: 1985-2000), not the current active player “Fernandez L.” (likely Leylah Fernandez).

Impact on Analysis:

• Cannot calculate expected hold/break rates for Player 1
• Cannot model game distributions
• Cannot calculate expected total games
• Cannot calculate expected game margin
• Both totals and spread markets are unanalyzable

Recommendation:

1. PASS on all markets for this match
2. Fix data scraper to correctly identify active “Fernandez L.” player
3. Re-collect briefing data before attempting analysis
4. Verify player name matching logic in stats_scraper.py

Fernandez L. - Complete Profile

INVALID DATA WARNING

All statistics below are INVALID (matched to retired player Mary Joe Fernandez):

Rankings & Form

Surface Performance (All Surfaces)

Hold/Break Analysis

Game Distribution Metrics

Serve Statistics

Return Statistics

Physical & Context

Tjen J. - Complete Profile

Rankings & Form

Surface Performance (All Surfaces)

Hold/Break Analysis

Game Distribution Metrics

Serve Statistics

Return Statistics

Physical & Context

Recent Form Analysis

Clutch Statistics

Playing Style

Matchup Quality Assessment

ANALYSIS NOT POSSIBLE

Cannot assess matchup quality without valid Player 1 data.

Available Player 2 Context:

• Janice Tjen has valid L52W statistics (14 matches, 78.5% hold, 30.1% break)
• Recent form is improving (7-2 in last 9)
• Playing style is error-prone (W/UFE ratio 0.91)
• Tiebreak performance is strong (83.3% win rate, but small sample n=6)

Missing Player 1 Context:

• Cannot determine Elo comparison
• Cannot determine form advantage
• Cannot determine hold/break differential
• Cannot model expected game distributions

Game Distribution Analysis

MODELING NOT POSSIBLE

Set Score Probabilities: Cannot calculate without Player 1 hold/break data

Match Structure: Cannot calculate without Player 1 statistics

Total Games Distribution: Cannot model without both players’ data

Totals Analysis

Market Analysis

Market Implied Probabilities (No-Vig):

• P(Over 20.5): 52.1%
• P(Under 20.5): 47.9%

Market Lean: Slight Over lean (market line 20.5 with Over juice)

Model Cannot Validate: Without Player 1 hold/break data, cannot determine if this line is accurate or if edge exists.

Factors Driving Total (If Data Were Available)

Player 2 Context (Tjen J.):

• Hold rate: 78.5% (moderate - suggests competitive service games)
• Break rate: 30.1% (decent return game)
• Avg total: 22.2 games
• Three-set frequency: 55.6% (competitive matches)
• Recent form avg: 25.0 games (trending higher)

Expected Pattern (Speculative): If Player 1 has similar hold/break rates (~75-80% hold), expect:

• Moderate hold rates → 10-11 games per set
• Low tiebreak probability
• Competitive sets → likely 2-1 outcome
• Total games estimate: 21-24 games range

But this is pure speculation without Player 1 data.

Handicap Analysis

Spread Coverage Probabilities

Market Analysis

Market Line: L. Fernandez -4.5 (implied favorite)

• Fernandez -4.5 @ 1.86 (No-vig: 49.7%)
• Tjen +4.5 @ 1.84 (No-vig: 50.3%)

Market Implication: Nearly even spread market (close to pick’em on -4.5 spread)

Model Cannot Validate: Without Player 1 data, cannot determine expected game margin or spread edge.

Head-to-Head (Game Context)

No H2H data available in briefing.

Market Comparison

Totals

Game Spread

Market Vig Analysis:

• Totals vig: 0.9% (fair market)
• Spread vig: 2.6% (typical market)

Cannot calculate edges without model probabilities.

Recommendations

Totals Recommendation

Rationale: Cannot analyze totals market without valid hold/break statistics for Player 1 (Fernandez L.). Data scraper matched to retired player “Mary Joe Fernandez” with 0 matches played. PASS recommendation is mandatory when critical data is missing.

Game Spread Recommendation

Rationale: Cannot model expected game margin without Player 1 hold/break statistics. The -4.5 spread market implies a close match, but we cannot validate this without calculating hold/break differentials. PASS is the only responsible recommendation.

Pass Conditions

• PRIMARY: Missing critical hold/break data for Player 1
• Data collection error (wrong player matched)
• Cannot model game distributions
• Cannot calculate expected totals or game margins
• Edge calculation impossible without model probabilities

Action Required:

1. Fix data scraper player matching logic
2. Re-collect briefing for correct “Fernandez L.” player (likely Leylah Fernandez)
3. Re-run analysis with valid data
4. Only then can totals/spread edges be calculated

Confidence Calculation

Base Confidence (from edge size)

Base Confidence: PASS (no edge calculable due to missing data)

Adjustments Applied

Data Quality Impact:

• Player 1 completeness: INVALID (0 matches, wrong player)
• Player 2 completeness: HIGH (14 matches, valid stats)
• Overall completeness: LOW
• Confidence multiplier: 0.0 (cannot proceed)

Final Confidence

Key Supporting Factors:

1. None (data invalid)

Key Risk Factors:

1. CRITICAL: Player 1 statistics completely invalid (matched to retired player)
2. Cannot model expected game distributions without both players’ hold/break data
3. Missing data = cannot calculate edge = mandatory PASS per methodology (2.5% minimum edge)

Risk & Unknowns

Variance Drivers

• Data Collection Failure: Primary risk is invalid data, not market variance
• Player Identification Error: Scraper matched “Fernandez L.” to wrong player
• Cannot Assess Market Variance: Without model, cannot determine if market is tight or wide

Data Limitations

CRITICAL LIMITATIONS:

• Player 1 (Fernandez L.) hold % = 0% (INVALID)
• Player 1 break % = 0% (INVALID)
• Player 1 matches played = 0 (INVALID - matched to retired player)
• Player 1 tiebreak data = 0 TBs (INVALID)
• Player 1 recent form = No data (INVALID)
• Player 1 all statistics = Missing/Invalid

Player 2 Data Quality:

• Tjen J. statistics appear valid (14 matches, reasonable percentages)
• Small sample size (n=14 matches) but sufficient for baseline analysis
• Tiebreak sample small (n=6 TBs) but above minimum threshold (>5)

Technical Issues to Resolve

Data Scraper Issues:

1. Player name matching logic needs improvement
2. “Fernandez L.” matched to “Mary Joe Fernandez” (retired 2000)
3. Should match to active “Leylah Fernandez” (or other active L. Fernandez)
4. Add validation: Flag if matches_played = 0
5. Add validation: Flag if player last match > 5 years ago

Recommended Fixes:

# In stats_scraper.py - add validation
if player_stats['matches_played'] == 0:
    logger.warning(f"Player {player_name} has 0 matches - likely wrong player match")

# Add recency check
if player_stats['last_match_date'] < (today - 365 days):
    logger.warning(f"Player {player_name} last match was >1 year ago - verify active status")

Correlation Notes

• Cannot assess correlation without model probabilities
• If analysis were possible, totals and spread would be correlated:
  • Higher total → more games → spread less meaningful
  • Lower total → fewer games → spread more decisive

Sources

1. TennisAbstract.com - Player statistics attempted (Last 52 Weeks Tour-Level Splits)
   • Player 1 ERROR: Matched to Mary Joe Fernandez (retired) instead of active player
   • Player 2 VALID: Janice Tjen statistics appear accurate (14 matches)
2. Sportsbet.io - Match odds (totals: O/U 20.5, spread: Fernandez -4.5)
3. Briefing File: fernandez_l_vs_tjen_j_briefing.json (collected 2026-01-19T08:22:28Z)

Verification Checklist

Core Statistics

• Hold % collected for both players (surface-adjusted) - FAILED (P1 invalid)
• Break % collected for both players (opponent-adjusted) - FAILED (P1 invalid)
• Tiebreak statistics collected (with sample size) - FAILED (P1 invalid)
• Game distribution modeled - FAILED (cannot model)
• Expected total games calculated with 95% CI - FAILED (cannot calculate)
• Expected game margin calculated with 95% CI - FAILED (cannot calculate)
• Totals line compared to market - FAILED (no model line)
• Spread line compared to market - FAILED (no model line)
• Edge ≥ 2.5% for any recommendations - N/A (no edges calculated)
• Confidence intervals appropriately wide - N/A (no modeling)
• NO moneyline analysis included - PASSED

Enhanced Analysis (New)

• Elo ratings extracted (overall + surface-specific) - FAILED (P1 missing)
• Recent form data included (last 10 record, trend, dominance ratio) - PARTIAL (P2 only)
• Clutch stats analyzed (BP conversion, BP saved, TB serve/return) - PARTIAL (P2 only)
• Key games metrics reviewed (consolidation, breakback, sv_for_set/match) - FAILED (P1 missing)
• Playing style assessed (winner/UFE ratio, style classification) - PARTIAL (P2 only)
• Matchup Quality Assessment section completed - FAILED (P1 missing)
• Clutch Performance section completed - FAILED (P1 missing)
• Set Closure Patterns section completed - FAILED (P1 missing)
• Playing Style Analysis section completed - FAILED (P1 missing)
• Confidence Calculation section with all adjustment factors - PASSED (PASS recommendation)

Data Quality Issues

• Data limitation clearly documented in report
• PASS recommendation issued for both totals and spread
• Root cause identified (player matching error)
• Recommended fixes outlined
• Report explains why analysis cannot proceed

Overall Verification Status: INCOMPLETE - Data collection error prevents analysis. PASS recommendation is appropriate and well-justified.

Appendix: Technical Notes for Development Team

Data Scraper Improvements Needed

Issue: stats_scraper.py matched “Fernandez L.” to retired player “Mary Joe Fernandez”

Likely Cause:

• Search query “Fernandez L.” returned multiple results
• Scraper selected first/wrong result without validation
• No recency check on player activity

Recommended Solutions:

1. Add Active Player Filter:

   # Filter to players with matches in last 52 weeks
   if player_stats['matches_played'] == 0:
       logger.warning("Player has 0 matches - trying alternative search")
       # Try more specific search or request manual confirmation
   

2. Add Disambiguation Logic:

   # If multiple "Fernandez L." results:
   # - Prioritize players with recent matches (last 12 months)
   # - Check tour (ATP vs WTA) against match metadata
   # - Verify ranking/Elo is reasonable (not 0)
   

3. Add Data Validation:

   def validate_player_stats(stats, player_name):
       if stats['matches_played'] == 0:
           raise ValueError(f"{player_name}: 0 matches played - likely wrong player")
       if stats['hold_pct'] == 0 and stats['break_pct'] == 0:
           raise ValueError(f"{player_name}: All stats are 0 - data collection failed")
       return True
   

4. Use Full Name When Available:
   • If briefing metadata can provide full first name, use it
   • “Leylah Fernandez” vs “L. Fernandez” improves matching accuracy
5. Manual Verification Mode:
   • For ambiguous players, output options and request selection
   • Cache correct player mappings for future matches

Testing:

• Test with other common surnames: Williams, Djokovic, Zverev
• Test with abbreviated first names: A. Zverev (Alexander vs Mischa)
• Test with retired vs active players of same name