Case Study: The Ripple Effect of a Franchise Quarterback Injury

How One Injury Transforms a Season

Introduction

In Week 4 of the 2023 season, the New York Jets faced a situation familiar to NFL analysts but devastating to experience: their franchise quarterback, Aaron Rodgers, tore his Achilles tendon on just the fourth snap of his Jets debut. This case study analyzes how such an injury ripples through a team's season, affects predictions, and tests our injury models.

Background: Pre-Injury Expectations

The Jets' Preseason Profile

Before Rodgers acquisition: - 2022 Record: 7-10 - Defensive ranking: Top 5 - Offensive ranking: Bottom 10 - Primary weakness: Quarterback play

After acquiring Rodgers: - Vegas Super Bowl odds: Moved from 50:1 to 12:1 - Win total over/under: Set at 10.5 - Point differential expectation: +60 (from -10)

The Analytical Assessment

Our model's pre-injury analysis:

Rodgers Value Add:
- Historical EPA/play: +0.15 (elite tier)
- Previous Jets QB EPA: -0.08
- Differential: +0.23 EPA/play
- Expected plays per game: 35
- Per-game impact: +8.0 EPA ≈ +6 points

Season Projection With Rodgers:
- Expected wins: 10.8
- Playoff probability: 78%
- Division win probability: 22%

The Injury Event

Week 4 vs Buffalo Bills

• Play: Fourth offensive snap of the season
• Injury: Non-contact Achilles tear
• Immediate Result: Rodgers ruled out for season
• Backup: Zach Wilson (2021 #2 overall pick, struggling history)

Immediate Market Response

Model Adjustment Analysis

Step 1: Quarterback Value Reassessment

Rodgers (unavailable):

Tier: Elite
Value over replacement: +6.5 points
Scheme fit: High (offense built for him)

Wilson (replacement):

Historical EPA/play: -0.04
Experience: 24 starts, 12-21 record
Tier: Below average
Value: -1.5 to -2.0 points vs replacement

Net Impact:

Starter loss: +6.5 points value lost
Backup penalty: -1.5 points additional
Total adjustment: +8.0 points

Step 2: Compound Effects

The Rodgers injury created secondary impacts:

Offensive Line Impact: - Protection schemes designed for Rodgers' preferences - Wilson has different mobility profile - Adjustment period required - Estimated additional impact: +0.5 points

Receiver Usage: - Routes designed for Rodgers' ball placement - Timing affected by different release - Estimated impact: +0.3 points

Play Calling: - Offensive coordinator must adjust - Fewer deep shots, more conservative - Estimated impact: +0.2 points

Total Compound Adjustment: +9.0 points per game

Step 3: Season Recalculation

def recalculate_season_projection(team, injury_adjustment):
    """Recalculate season after major injury."""
    original_projected_margin = team['projected_margin']  # +3.5 per game
    games_remaining = 17

    adjusted_margin = original_projected_margin - injury_adjustment
    # +3.5 - 9.0 = -5.5 per game

    expected_wins = calculate_wins_from_margin(adjusted_margin, games_remaining)
    # -5.5 margin ≈ 5.5 expected wins

    playoff_probability = calculate_playoff_odds(expected_wins, division)
    # 5.5 wins ≈ 8% playoff probability

    return {
        'original_wins': 10.8,
        'adjusted_wins': 5.5,
        'change': -5.3,
        'playoff_probability': 0.08
    }

Weekly Tracking

Season Results With Wilson

First 5 Games Post-Injury: - Record: 0-5 (from 1-3 with Rodgers/pre-injury) - ATS Record: 1-4 - Average margin: -8.2 points - Model projected: -5.5 points - Model underestimated impact

Mid-Season Adjustment

After Wilson struggled, the Jets made changes:

Week 11: Tim Boyle starts Week 12-15: Trevor Siemian/Wilson rotation Week 16-18: Various combinations

Each change required model re-assessment:

Wilson value: -1.5 points vs replacement
Boyle value: -0.5 points vs replacement
Siemian value: +0.5 points vs replacement

Backup upgrade = +1.0 to +2.0 points improvement

Model Validation

Comparing Predictions to Results

Analysis: - Point differential prediction was accurate - Win total was underestimated - Reason: Jets won several close games (variance)

ATS Performance Analysis

The market efficiently priced the injury after initial overreaction.

Key Lessons

Lesson 1: Initial Market Reaction

The immediate 8.5-point line swing accurately captured the QB differential: - Model estimate: +9.0 points - Market adjustment: +8.5 points - Markets are efficient for high-visibility injuries

Lesson 2: Compound Effects Are Real

Secondary impacts beyond direct QB loss: - Scheme disruption - Morale/confidence effects - Play calling limitations

Our model's +1.0 compound adjustment was reasonable but possibly underestimated.

Lesson 3: Backup Quality Matters

Wilson's poor performance (-0.04 EPA) contributed significantly: - If Jets had quality backup (+0.02 EPA): ~2 fewer points lost - Total impact would have been ~7 points instead of 9

Lesson 4: Variance Still Dominates

Despite accurate margin predictions: - Jets won 7 games on 5.5 projection - Close games went their way - Defense overperformed expectations

Lesson 5: In-Season Adaptation

Teams adjust to injuries: - Play calling evolved - Wilson eventually benched - Different backup profiles tested

Models should account for adaptation over time.

Alternative Scenarios

Scenario A: Quality Backup

What if Jets had a backup like Jameis Winston (career starter)?

Winston value: +1.0 vs replacement
Impact would be: 6.5 - 1.0 = +5.5 points
Expected wins: 7.5 (vs 5.5 with Wilson)
Playoff probability: 25% (vs 8%)

Scenario B: Rodgers Returns Week 10

If Rodgers had returned mid-season (hypothetically):

Games 1-9 adjustment: -9.0 per game
Games 10-17 adjustment: -2.0 per game (returning from injury)
Weighted average: -5.5 per game
Expected wins: 8.0
Playoff probability: 40%

Scenario C: Injury Happens Week 14

If injury occurred late season:

Games 1-13: Full Rodgers value
Games 14-17: Wilson adjustment
Expected wins: 9.5
Playoff probability: 65%

Timing dramatically affects season outcomes.

Building Better Injury Models

Improvements Identified

1. Better compound effect estimation - Track scheme-specific dependencies - Model play calling adjustments

2. Dynamic backup assessment - Update backup projections weekly - Account for learning/adaptation

3. Confidence interval expansion - Larger uncertainty with backup QBs - Wilson's high variance should widen predictions

4. In-season adjustment - Reduce injury impact over time - Teams adapt to absence

Updated Model Framework

def enhanced_qb_injury_model(team, starter, backup, week_of_injury):
    """Enhanced QB injury adjustment model."""

    # Base differential
    base_impact = starter['value'] - backup['value']

    # Compound effects (scheme dependence)
    scheme_factor = calculate_scheme_dependence(team, starter)
    compound = base_impact * (1 + scheme_factor * 0.1)

    # Time decay (adaptation)
    weeks_since = current_week - week_of_injury
    adaptation_factor = 1 - min(0.2, weeks_since * 0.02)
    adjusted_impact = compound * adaptation_factor

    # Uncertainty expansion
    backup_variance = calculate_backup_variance(backup)
    confidence_expansion = 1 + backup_variance

    return {
        'point_impact': adjusted_impact,
        'confidence_multiplier': confidence_expansion
    }

Conclusion

The 2023 Jets season provides a clear illustration of quarterback injury impact. The immediate adjustment (8-9 points) accurately reflected the starter-backup differential. However, compound effects and backup performance variance created additional challenges. Key takeaways:

1. QB injuries are massive - No other position creates 8+ point swings
2. Markets price efficiently - Major injuries are quickly incorporated
3. Backup quality varies widely - Assessment is crucial
4. Compound effects exist - Add 10-15% for scheme disruption
5. Variance increases - Widen confidence intervals significantly

The framework developed from this case study can be applied to any high-impact injury situation.

Discussion Questions

1. How would your analysis differ if Rodgers was injured in Week 14 instead of Week 4?

2. The Jets defense remained elite despite the QB injury. How should defensive performance be modeled independently?

3. Markets moved 8.5 points immediately. Was this an overreaction, underreaction, or appropriate?

4. How would you model the "emotional" impact of losing a high-profile player?

5. If you could access practice reports and injury data in real-time, how would you modify your approach?

Data Sources

• Game results: Pro Football Reference
• Injury reports: Official NFL injury reports
• Line movements: Historical betting data
• EPA calculations: nflfastR

Technical Appendix: Injury Impact Calculation

# Complete calculation for Rodgers injury

starter_value = {
    'name': 'Aaron Rodgers',
    'epa_per_play': 0.15,
    'tier': 'elite',
    'point_value': 6.5
}

backup_value = {
    'name': 'Zach Wilson',
    'epa_per_play': -0.04,
    'tier': 'below_average',
    'point_value': -1.5
}

# Direct differential
differential = starter_value['point_value'] - backup_value['point_value']
# 6.5 - (-1.5) = 8.0 points

# Compound factor (scheme dependence = 0.12)
compound_multiplier = 1 + 0.12
total_impact = differential * compound_multiplier
# 8.0 * 1.12 = 8.96 ≈ 9.0 points

# Confidence interval
base_std = 13.5  # Normal NFL margin std
backup_variance_factor = 1.15  # Wilson's high variance
adjusted_std = base_std * backup_variance_factor
# 13.5 * 1.15 = 15.5 points std

# 95% confidence interval for margin
ci_width = 1.96 * adjusted_std
# ±30.4 points (very wide due to uncertainty)