Over the past few years every time the Browns get "cute" when calling a play I think to myself, "Why in God's name do the Browns give Paul DePodesta so much influence?"
That stated, I do believe that the strategy of questioning conventional wisdom has paid dividends for the Browns if only in their selection of Andrew Berry as General Manager. Having a "team first" leader at the top of any organization who can put his ego to the side and use a facts based approach that questions conventional wisdom when necessary is a huge plus. Andrew Berry is all of those things and much more.
It does appear that Browns are moving past the era where every single bit of conventional NFL wisdom is second guessed and that's a good thing too. Bringing your second string quarterback into the game to replace your $230,000,000 quarterback on 4th down and less than 1 yard vs a Division Rival and then throwing a 40 yard incomplete pass to the end zone may well challenge conventional wisdom but sometimes conventional wisdom is actually wise.
Here’s a comparison of the use of analytics in professional baseball versus professional football, focusing on the significance and relevance of measurable variables:
Baseball Analytics
1. Nature of the Game:
- Discrete Events: Baseball is composed of a series of discrete events (pitches, at-bats, fielding plays), which are easier to isolate and analyze.
- Large Sample Size: Each player has many opportunities (e.g., hundreds of at-bats or innings pitched) to generate data, leading to robust statistical analysis.
2. Measurable Variables:
- Pitching Metrics: ERA, WHIP, strikeout rates, and pitch velocities.
- Hitting Metrics: Batting average, on-base percentage, slugging percentage, exit velocity, and launch angle.
- Fielding Metrics: Defensive runs saved (DRS), ultimate zone rating (UZR), and fielding percentage.
3. Impact of Analytics:
- Player Evaluation: Sabermetrics have revolutionized how players are valued (e.g., Moneyball).
- In-Game Decisions: Data-driven decisions on shifts, pitch selection, and batting order.
- Long-Term Strategy: Teams use analytics for draft strategy, player development, and contract negotiations.
Football Analytics
1. Nature of the Game:
- Complex Interactions: Football involves complex interactions among 22 players on the field, making it harder to isolate individual contributions.
- Smaller Sample Size: Fewer games and plays per season result in a smaller data set for analysis.
2. Measurable Variables:
- Player Metrics: Yardage (passing, rushing, receiving), touchdowns, and interceptions.
- Team Metrics: Points scored, points allowed, and time of possession.
- Advanced Metrics: Expected points added (EPA), win probability, and player tracking data (e.g., Next Gen Stats).
3. Impact of Analytics:
- Player Evaluation: Combines traditional scouting with data on player efficiency and situational performance.
- In-Game Decisions: Fourth-down decision-making, play-calling tendencies, and clock management.
- Long-Term Strategy: Roster construction, injury prevention, and game planning.
Conclusion
Baseball:
- Statistical Significance and Relevance: Due to the discrete nature of events and the large sample size, baseball analytics yield statistically significant and relevant insights. These metrics can be highly predictive of future performance and directly influence game outcomes and player evaluation.
Football:
- Data Insufficiency: The complex interactions on the field, smaller sample sizes, and greater variability make it more challenging to derive statistically significant and actionable insights. While analytics are useful for strategic planning and situational decisions, they are less deterministic compared to baseball.
In summary, while analytics play a critical role in both sports, the measurable variables in baseball are more statistically significant and relevant due to the nature of the game, leading to clearer and more direct applications. In contrast, football's complexity and variability mean that while data can inform strategy, it often requires a more nuanced and context-dependent approach.