Chapter 14: Exercises - Player and Team Comparison Charts

Exercise Overview

These exercises build skills in creating effective comparison visualizations for football analytics. Progress from basic comparisons to complex multi-entity analysis.

Level 1: Conceptual Understanding

Exercise 1.1: Comparison Type Classification

For each scenario below, identify the appropriate comparison type and justify your choice:

a) Comparing a quarterback's 2023 vs 2024 statistics b) Ranking all SEC running backs by yards per carry c) Showing how four teams perform across offensive, defensive, and special teams metrics d) Tracking weekly AP Poll rankings for top 10 teams e) Comparing a recruit to current roster players

Answer Framework: | Scenario | Comparison Type | Justification | |----------|-----------------|---------------| | a | | | | b | | | | c | | | | d | | | | e | | |

Exercise 1.2: Chart Type Selection

Match each visualization need to the most appropriate chart type:

  1. Show 12 teams' performance across 6 metrics simultaneously
  2. Compare two QBs' stats with their percentile ranks
  3. Show how conference standings changed week by week
  4. Display improvement in rushing yards from Week 1 to Week 12
  5. Rank 25 teams by defensive EPA

Chart Options: A. Horizontal bar chart B. Radar chart comparison C. Bump chart D. Slope chart E. Dumbbell chart

Exercise 1.3: Fairness Analysis

Identify the fairness issues in these comparison scenarios:

a) Comparing a starter's total yards to a backup's yards per game b) Ranking players by touchdowns without adjusting for games played c) Comparing a 2024 quarterback to a 1990 quarterback using raw passing yards d) Creating a radar chart using only metrics where one player excels

Level 2: Implementation Basics

Exercise 2.1: Simple Ranking Bar Chart

Create a horizontal bar chart ranking SEC teams by points per game:

teams = ['Georgia', 'Alabama', 'LSU', 'Tennessee', 'Ole Miss',
         'Texas A&M', 'Kentucky', 'Missouri', 'Florida', 'Auburn',
         'South Carolina', 'Arkansas', 'Mississippi State', 'Vanderbilt']

ppg = [38.2, 35.8, 34.1, 32.5, 31.8, 28.4, 26.1, 25.8, 25.2,
       24.1, 23.5, 22.8, 21.4, 18.9]

conference_avg = 27.3

Requirements: - Sort from highest to lowest - Add conference average reference line - Highlight top 3 teams - Include value labels

Exercise 2.2: Basic Radar Chart

Create a radar chart for this quarterback:

metrics = ['Completion %', 'Yards/Att', 'TD Rate', 'INT Rate (inv)',
           'Rush Yards', 'Sack Rate (inv)']

# Values normalized to 0-100 scale (higher is better)
qb_stats = {
    'Completion %': 72,
    'Yards/Att': 85,
    'TD Rate': 68,
    'INT Rate (inv)': 88,  # Already inverted: 100 = no interceptions
    'Rush Yards': 45,
    'Sack Rate (inv)': 62
}

Requirements: - Proper polar projection - Filled area with transparency - Clear metric labels - Appropriate title

Exercise 2.3: Dumbbell Chart

Create a dumbbell chart showing offensive improvement:

teams = ['Georgia', 'Ohio State', 'Alabama', 'Michigan', 'Texas']

# EPA per play
epa_first_half = [0.15, 0.22, 0.18, 0.12, 0.08]
epa_second_half = [0.28, 0.19, 0.24, 0.21, 0.22]

Requirements: - Color connectors based on improvement direction (green for up, red for down) - Show change magnitude labels - Sort by improvement amount

Level 3: Intermediate Applications

Exercise 3.1: Multi-Player Radar Comparison

Create a radar chart comparing three running backs:

metrics = ['Yards/Carry', 'TD Rate', 'YAC/Carry', 'Broken Tackles/Att',
           'Receiving Yards', 'Pass Block Grade', 'Fumble Rate (inv)']

players = {
    'Bijan Robinson': {
        'Yards/Carry': 92,
        'TD Rate': 78,
        'YAC/Carry': 88,
        'Broken Tackles/Att': 95,
        'Receiving Yards': 82,
        'Pass Block Grade': 71,
        'Fumble Rate (inv)': 94
    },
    'Jahmyr Gibbs': {
        'Yards/Carry': 85,
        'TD Rate': 72,
        'YAC/Carry': 91,
        'Broken Tackles/Att': 76,
        'Receiving Yards': 96,
        'Pass Block Grade': 68,
        'Fumble Rate (inv)': 89
    },
    'Devon Achane': {
        'Yards/Carry': 88,
        'TD Rate': 68,
        'YAC/Carry': 94,
        'Broken Tackles/Att': 72,
        'Receiving Yards': 84,
        'Pass Block Grade': 58,
        'Fumble Rate (inv)': 92
    }
}

Requirements: - Distinct colors for each player - Semi-transparent fills - Legend with player names - Professional styling

Exercise 3.2: Conference Bump Chart

Create a bump chart showing Big Ten standings:

weeks = ['Week 1', 'Week 4', 'Week 8', 'Week 12', 'Final']

standings = {
    'Ohio State': [2, 1, 1, 2, 2],
    'Michigan': [3, 2, 2, 1, 1],
    'Penn State': [1, 3, 3, 3, 3],
    'Wisconsin': [5, 4, 5, 4, 4],
    'Iowa': [4, 5, 4, 5, 5],
    'Minnesota': [7, 6, 6, 7, 6],
    'Maryland': [6, 7, 8, 6, 7],
    'Illinois': [8, 8, 7, 8, 8],
    'Purdue': [10, 9, 9, 9, 9],
    'Nebraska': [9, 10, 10, 10, 10]
}

Requirements: - Highlight top 3 teams with bold lines - Rank 1 at top of y-axis - Labels for highlighted teams - Gridlines for each rank position

Exercise 3.3: Grouped Bar Comparison

Create a grouped bar chart comparing teams across efficiency metrics:

teams = ['Georgia', 'Alabama', 'Ohio State', 'Michigan']

metrics = {
    'Offensive EPA/Play': [0.22, 0.19, 0.24, 0.15],
    'Defensive EPA/Play': [-0.18, -0.15, -0.12, -0.21],  # Negative is good
    'Special Teams EPA/Play': [0.04, 0.02, 0.01, 0.05]
}

Requirements: - Grouped bars for each team - Appropriate color scheme - Legend identifying metrics - Note that negative defensive EPA is positive for defense

Level 4: Advanced Techniques

Exercise 4.1: Percentile Profile Chart

Create a percentile profile visualization:

# Target player stats
player = "Caleb Williams"
player_stats = {
    'Completion %': 66.8,
    'Yards/Attempt': 9.2,
    'TD Rate': 7.8,
    'INT Rate': 2.1,
    'QBR': 88.4,
    'Adjusted Comp %': 78.2,
    'Big Time Throw %': 6.4,
    'Turnover Worthy Play %': 2.3
}

# Population distributions (all FBS QBs with 200+ attempts)
population = {
    'Completion %': [58.2, 59.1, 60.4, ..., 72.3],  # 130 values
    'Yards/Attempt': [5.8, 6.1, ..., 10.2],
    # ... etc
}

Requirements: - Horizontal bars showing percentile position (0-100 scale) - Color zones: red (<25), yellow (25-50), light green (50-75), dark green (>75) - Show raw value and percentile for each metric - 50th percentile reference line

Exercise 4.2: Similarity Heatmap and Dendrogram

Create clustering visualization for player similarity:

import numpy as np

players = ['Player A', 'Player B', 'Player C', 'Player D',
           'Player E', 'Player F', 'Player G', 'Player H']

# Feature matrix (8 players x 6 metrics)
features = np.array([
    [72, 8.5, 6.2, 2.1, 85, 120],
    [65, 7.8, 5.8, 3.2, 72, 280],
    [70, 8.2, 6.0, 2.4, 82, 150],
    [64, 7.5, 5.5, 3.5, 68, 310],
    [71, 8.4, 6.1, 2.2, 84, 130],
    [63, 7.2, 5.2, 3.8, 65, 340],
    [68, 7.9, 5.7, 2.8, 76, 220],
    [66, 8.0, 5.9, 2.6, 78, 190]
])

Requirements: - Create similarity heatmap (8x8 grid) - Create dendrogram showing hierarchical clustering - Use ward linkage method - Identify natural clusters

Exercise 4.3: Small Multiples by Conference

Create small multiples showing rushing yards distribution by conference:

conferences = {
    'SEC': [rushing_yards for 40 players],
    'Big Ten': [rushing_yards for 35 players],
    'Big 12': [rushing_yards for 32 players],
    'ACC': [rushing_yards for 38 players],
    'Pac-12': [rushing_yards for 30 players]
}

# Reference: FBS average = 485 yards
fbs_average = 485

Requirements: - 5 histogram subplots (1 row, 5 columns) - Consistent x-axis range across all plots - FBS average reference line - Conference mean marked in each plot - Clear conference labels

Level 5: Professional Applications

Exercise 5.1: Complete Player Comparison Dashboard

Build a multi-panel dashboard comparing two quarterbacks:

Data Provided: - Player statistics (15 metrics each) - Population distributions for percentile calculations - Game-by-game trends (12 games each) - Situational splits (by down, by quarter)

Required Panels: 1. Radar chart overlay 2. Metric-by-metric bar comparison 3. Percentile comparison chart 4. Season trend comparison (line chart) 5. Situational efficiency heatmap comparison

Design Requirements: - Consistent color scheme (Player A: blue tones, Player B: red tones) - Professional typography - Clear titles for each panel - Cohesive layout with proper spacing

Exercise 5.2: Recruiting Comparison Tool

Create visualization for comparing a recruit to current roster players:

Data:

recruit = {
    'name': 'Arch Manning',
    'position': 'QB',
    'height': 76,  # inches
    'weight': 215,
    'arm_strength': 92,  # 0-100 rating
    'accuracy': 88,
    'mobility': 75,
    'football_iq': 90,
    'leadership': 95
}

roster_qbs = [
    {'name': 'Quinn Ewers', 'height': 74, 'weight': 210, ...},
    {'name': 'Maalik Murphy', 'height': 76, 'weight': 230, ...},
    {'name': 'Cole Lourd', 'height': 73, 'weight': 195, ...}
]

Required Visualizations: 1. Radar chart: recruit vs roster average 2. Bar chart: where recruit ranks on each attribute 3. Similarity scores: which roster player is most similar 4. Gap analysis: where recruit exceeds/trails roster

Exercise 5.3: Conference Power Rankings Dashboard

Create an interactive-style static dashboard showing conference comparisons:

Components: 1. Bump chart: Top 10 teams across weekly polls 2. Diverging bar chart: Conference aggregate EPA vs FBS average 3. Small multiples: Each conference's offensive vs defensive efficiency 4. Stacked bars: Conference breakdown by team tier (Elite/Good/Average/Below)

Requirements: - Publication-quality design - Cohesive visual theme - Comprehensive annotations - Data source citations

Challenge Exercises

Challenge A: Fair Comparison System

Design a comparison system that automatically adjusts for: - Games played - Opponent strength - Home/away splits - Weather conditions

Create visualizations that clearly communicate both raw and adjusted values.

Challenge B: Historical Player Comparison

Build a system to compare players across different eras: - Era adjustment factors for passing stats (1980 vs 2020) - Visualization showing both raw and era-adjusted comparisons - Uncertainty ranges reflecting adjustment imprecision

Challenge C: Transfer Portal Analytics

Create comparison visualizations for transfer portal analysis: - Player production at previous school - How similar players have performed after transferring - System fit analysis (player style vs new team's offensive scheme) - Uncertainty in projections

Submission Checklist

For each exercise, ensure your submission includes:

  • [ ] Complete Python code with comments
  • [ ] Generated visualization file (.png or .pdf)
  • [ ] Brief description of design choices
  • [ ] Discussion of any limitations or caveats
  • [ ] For Level 4+: Alternative approaches considered

Grading Criteria

Criterion Weight
Technical correctness 25%
Visual clarity 25%
Comparison fairness 20%
Code quality 15%
Design aesthetics 15%