Advanced analytics in sports have revolutionized how we understand and report on athletic performance. By crunching numbers and tracking detailed data, teams and journalists can now dive deeper into player skills, game strategies, and overall team dynamics.

This shift towards data-driven analysis has changed how sports are played, coached, and covered in the media. From player evaluation to in-game decision-making, advanced metrics offer fresh insights that go beyond traditional stats, reshaping how we view and talk about sports.

Advanced analytics in sports reporting

Defining advanced analytics

• Advanced analytics, also known as sports analytics, is the use of data and statistical analysis to measure performance and make decisions to gain a competitive sports advantage
• The growth of technology has enabled teams, broadcasters and journalists to capture more granular data about player and team performance than traditional statistics (player tracking, pitch/hit tracking)
• Key aspects of sports that analytics aim to quantify and optimize include:
  • On-field strategy
  • Player health and development
  • Roster construction
  • Play-by-play analysis
• Advanced analytics provide an objective lens to augment subjective observations and conventional wisdom in evaluating sports performance

Applications of advanced analytics

• Sports analytics are used for varied purposes such as:
  • Game preparation
  • In-game decision making
  • Player evaluation
  • Sports betting
  • Predictive analysis
• The rise of analytics has transformed how sports are played, coached, managed and reported on in the media
  • Teams leverage analytics for a competitive advantage (player acquisition, in-game tactics)
  • Media use analytics to provide deeper insight into performance and strategy
  • Betting models rely heavily on advanced metrics to set lines and identify value

Advanced metrics for sports

Sport-specific advanced metrics

• Basketball analytics emphasize metrics like:
  • True shooting percentage
  • Rebounding percentage
  • Player efficiency rating (PER)
  • Plus-minus
  • Defensive rating to analyze player impact
• Advanced baseball metrics include:
  • Wins above replacement (WAR)
  • Fielding independent pitching (FIP)
  • Ultimate zone rating (UZR)
  • Spin rate
  • Exit velocity
  • Sabermetrics, the empirical analysis of baseball statistics, was a forerunner to modern sports analytics
• Football analytics feature metrics like:
  • Defense-adjusted value over average (DVOA)
  • Yards above replacement (YAR)
  • Expected points added (EPA)
  • Completion probability
• Hockey utilizes advanced stats such as:
  • Corsi
  • Fenwick
  • Zone starts
  • PDO (shooting percentage plus save percentage) to measure possession and performance
• Soccer analytics incorporate:
  • Expected goals (xG)
  • Expected assists (xA)
  • Pass completion rate
  • Defensive actions
  • Pressures to assess players and teams
• Golf analytics emphasize strokes gained as a method to compare players' performance relative to the field in:
  • Driving
  • Approach
  • Short game
  • Putting

Cross-sport applications

• While the specific metrics vary, advanced analytics share common goals across sports:
  • Evaluate player value and efficiency
  • Quantify key skills and events
  • Measure and predict team performance
  • Identify optimal tactics and strategies
• Many advanced metrics are based on comparing individual performance to a league average or replacement level baseline
• Analytics can be applied to evaluate both individual players and broader five-man units, pairings and team performance
• Spatial tracking data is being used across sports to quantify movement, spacing and events in more detail

Performance evaluation with analytics

Evaluating players and teams

• Possession metrics like Corsi and Fenwick quantify which hockey teams and players are controlling the puck and outshooting their opponents, which is predictive of success
• Wins Above Replacement (WAR) estimates the total number of wins a baseball player adds to his team compared to a readily available minor league replacement player
  • WAR encapsulates a player's total value across hitting, base running, fielding and pitching
• Adjusted plus-minus in basketball measures how much a player impacts their team's point differential while on the court, adjusted for strength of teammates and opponents
  • Plus-minus quantifies the net point differential when a player is on the court
• Expected goals (xG) measures the likelihood of a soccer shot resulting in a goal based on shot distance, angle and type, and evaluates player finishing and chance creation
  • xG can quantify how many goals a player "should have" scored based on the quality of their shots
• Yards Above Replacement (YAR) in football compares the performance of skill position players and passers to a theoretical replacement-level baseline
  • A similar concept to WAR in baseball
• Advanced golf metrics break down player performance across driving distance and accuracy, approach proximity, short game and putting compared to the PGA Tour average
  • Strokes gained quantifies how many strokes a player gains on the field in each facet of the game

Contextualizing analytics

• Analytics provide a useful evaluation tool but require context:
  • Role: How a player is deployed in their team's system and their primary responsibilities
  • Competition: The strength of a player's league and the quality of their opponents
  • Sample size: Whether the performance sample is large enough to draw reliable conclusions
  • Teammates: How a player's teammates impact their performance and efficiency
• Box score stats, film study and the eye test are still important complements to analytics
• Analytics are descriptive of past performance, with imperfect predictive power for future outcomes

Limitations of sports analytics

Contextual limitations

• Analytics are based on a sample of past data and may not always be predictive of future results, especially for players with smaller sample sizes
  • Young players have limited data which may not represent true talent level
  • Aging curves are not always predictive for older players
• Certain advanced metrics like plus-minus and WAR can be noisy in small samples, requiring appropriate context about reliability
  • Single-game plus-minus is not a reliable indicator of performance
  • WAR stabilizes over larger samples (multiple seasons)
• Analytics are often taken out of context as singular metrics of player value without factoring in the team system, role and complementary pieces
  • Players can be miscast in roles that do not maximize their value
  • Lineup synergies and interactions impact individual performance

Philosophical limitations

• The widespread use of analytics can lead to flawed group think if certain metrics become over-valued in player evaluation and decision-making
  • Teams can become over-reliant on metrics at the expense of other evaluation methods
  • Metrics like pitcher wins and RBIs were historically overweighted in baseball
• Analytics can fail to capture difficult to quantify skills and intangibles such as:
  • Leadership
  • Locker room presence
  • Effort
  • Mental makeup
• Overemphasis on analytics in sports reporting can come at the expense of game storytelling, personalities and other intangible factors core to fan interest
  • Analytics provide informational value but do not capture the full viewing experience
  • The human element and unknowability of sports is part of the appeal