Exercises: Team Efficiency Metrics

Exercise 1: Basic EPA Calculation

Calculate offensive and defensive EPA/play for a team with the following data:

What is their net EPA/play?

Exercise 2: Success Rate Computation

Given the following play outcomes, calculate the team's overall success rate:

• 245 plays with positive EPA
• 305 plays with negative or zero EPA

Then classify their success rate tier based on league averages.

Exercise 3: Explosiveness Analysis

A team has the following explosive play data:

Calculate their explosive play rates for each category and overall.

Exercise 4: Success-Explosiveness Quadrant

Classify the following teams into their success-explosiveness quadrants:

Exercise 5: Net EPA Interpretation

Match each net EPA/play value with the correct team quality tier:

• Team A: +0.18
• Team B: +0.05
• Team C: -0.02
• Team D: -0.12

Tiers: Elite, Very Good, Above Average, Below Average, Poor

Exercise 6: Pass vs Rush Efficiency

A team has the following efficiency metrics:

Calculate their pass premium and determine if it's above or below league average (~0.08).

Exercise 7: Down-by-Down Success

Calculate success rate by down for a team with:

Compare each to league averages (1st: 48%, 2nd: 42%, 3rd: 38%).

Exercise 8: Defensive Efficiency

A defense allowed the following:

Evaluate this defense's performance relative to league averages.

Exercise 9: Composite Score Calculation

Using the following normalized scores (0-100), calculate a composite team rating:

Exercise 10: Garbage Time Filtering

A team's data before and after garbage time filtering:

How did filtering change the team's apparent quality?

Exercise 11: Year-to-Year Correlation

Two metrics have the following year-to-year correlations:

• Metric A: r = 0.58
• Metric B: r = 0.31

If you're projecting next season, which metric is more reliable? What percentage of variance does each explain?

Exercise 12: Red Zone Efficiency

A team's red zone offense:

Calculate TD rate, points per trip, and overall efficiency.

Exercise 13: Passing Situation Splits

Analyze passing efficiency by game situation:

Which situations show the best and worst performance?

Exercise 14: Rush Direction Analysis

A team's rushing efficiency by direction:

Identify the team's strongest and weakest rushing lanes.

Exercise 15: Score-Adjusted Analysis

Calculate win probability-weighted EPA for:

What is the weighted EPA/play?

Exercise 16: Team Comparison

Compare two teams:

Which team is better overall? Why?

Exercise 17: Correlation Analysis

Given these correlations with wins:

Rank these metrics by predictive value. Why might rush yards have low correlation?

Exercise 18: Situational Defense

A defense's EPA allowed by situation:

Identify the defense's strengths and weaknesses.

Exercise 19: Stability Projection

Project next year's EPA using regression to the mean:

Current EPA: +0.15 Stability coefficient: r = 0.55 League mean: 0.00

Use the formula: Projected = League Mean + (r × (Current - League Mean))

Exercise 20: Multi-Season Analysis

A team's metrics over three seasons:

Calculate 3-year averages and identify trends.

Programming Exercises

Exercise 21: Team EPA Calculator

Write a Python function that calculates team efficiency metrics:

def calculate_team_efficiency(pbp: pd.DataFrame, team: str) -> dict:
    """
    Calculate comprehensive efficiency metrics for a team.

    Returns dict with:
    - off_epa_play: Offensive EPA per play
    - def_epa_play: Defensive EPA per play
    - net_epa: Net EPA per play
    - success_rate: Offensive success rate
    - explosive_rate: Explosive play rate
    """
    pass

Exercise 22: League Rankings

Write a function that ranks all teams by multiple efficiency metrics:

def rank_teams(pbp: pd.DataFrame) -> pd.DataFrame:
    """
    Return DataFrame with all teams ranked by:
    - Offensive EPA (desc)
    - Defensive EPA (asc - lower is better)
    - Net EPA (desc)
    - Success Rate (desc)

    Include composite rank as average of all ranks.
    """
    pass

Exercise 23: Garbage Time Filter

Implement a garbage time filter:

def filter_garbage_time(pbp: pd.DataFrame,
                        min_wp: float = 0.05,
                        max_wp: float = 0.95) -> pd.DataFrame:
    """
    Remove plays where win probability is extreme.

    Args:
        pbp: Play-by-play data
        min_wp: Minimum win probability to include
        max_wp: Maximum win probability to include

    Returns:
        Filtered DataFrame
    """
    pass

Exercise 24: Success-Explosiveness Visualization

Create a visualization function:

def plot_success_explosiveness(team_metrics: pd.DataFrame) -> None:
    """
    Create scatter plot of success rate vs explosive rate.

    - Color by net EPA
    - Add quadrant lines at median values
    - Label each team
    - Add title and axis labels
    """
    pass

Exercise 25: Team Report Generator

Build a complete team report generator:

def generate_team_report(pbp: pd.DataFrame, team: str, season: int) -> str:
    """
    Generate comprehensive text report for a team.

    Include:
    - Overall efficiency summary
    - Offensive analysis (pass/rush splits)
    - Defensive analysis
    - Situational breakdowns
    - League rankings
    - Strengths and weaknesses
    """
    pass

Challenge Exercises

Challenge 1: Opponent-Adjusted Metrics

Calculate opponent-adjusted EPA:

1. For each team's opponents, find their average EPA allowed
2. Adjust team's EPA by subtracting opponent average
3. Compare raw vs adjusted rankings

How much do rankings change?

Challenge 2: Predictive Model

Build a simple wins predictor:

1. Use net EPA to predict wins
2. Calculate correlation and RMSE
3. Compare to point differential model
4. Identify teams that over/underperformed predictions

Challenge 3: Stability Analysis

Analyze metric stability across seasons:

1. Load 3 years of data
2. Calculate year-to-year correlations for: - Offensive EPA - Defensive EPA - Pass EPA - Rush EPA - Success Rate
3. Which metrics are most stable?

Challenge 4: Composite Rating System

Design your own composite rating:

1. Choose which metrics to include
2. Determine appropriate weights
3. Normalize all metrics to common scale
4. Test correlation with wins
5. Compare to simple net EPA

Challenge 5: Real-Time Dashboard

Create an efficiency dashboard:

1. Load current season data
2. Calculate all major efficiency metrics
3. Create visualizations for: - League-wide rankings - Team-specific breakdown - Trend over time (by week)
4. Export to HTML or create interactive display

Solutions

Solutions are available in code/exercise-solutions.py