Case Study: The Fourth Down Revolution

"The numbers are on our side. Why wouldn't we go for it?" — Lincoln Riley

Executive Summary

This case study examines how analytics transformed fourth-down decision-making in college football, creating one of the most visible examples of data changing the game. You will trace the evolution from conservative punting to aggressive conversion attempts, understand the analytical framework behind these decisions, and examine why adoption has been uneven despite clear evidence.

Skills Applied: - Understanding expected value calculations - Analyzing decision-making under uncertainty - Recognizing behavioral factors in analytics adoption

Background

The Traditional Approach

For most of football history, fourth-down decisions followed a simple rule: punt almost always.

The logic seemed obvious. Failing to convert gives the opponent good field position. Punting moves them back 40 yards. Why risk giving away points?

Coaches learned this from their coaches, who learned from their coaches. Fourth-down conservatism was baked into football culture. A coach who went for it and failed faced criticism from fans, media, and administrators. A coach who punted and lost faced less scrutiny—"he played it safe."

This created what economists call an agency problem. The optimal choice for the team (go for it more often) differed from the safe choice for the coach's career (punt). Coaches who deviated and failed faced consequences; those who played conservatively and lost could blame the players or circumstances.

The Analytical Challenge

Starting in the 2000s, researchers began quantifying fourth-down decisions rigorously.

The key insight: both options—going for it and punting—have uncertain outcomes. Analytics compares the expected values of each option, accounting for:

If you go for it: - Probability of conversion × Value of first down at that spot - Probability of failure × Cost of opponent having ball there

If you punt: - Expected net punt distance × Value of opponent having ball at that field position

When the expected value of going for it exceeds the expected value of punting, analytics says go for it.

What the Numbers Showed

Multiple independent analyses reached the same conclusion: teams punt too often.

The math is particularly compelling near midfield. Converting extends the drive toward scoring territory. Failing gives opponents decent field position—but so does a punt from there. The incremental field position gain from punting doesn't justify giving up a conversion attempt with 50%+ success odds.

The Pioneers

Kevin Kelley and Pulaski Academy

The most aggressive adopter wasn't a college or NFL coach—it was Kevin Kelley, head coach of Pulaski Academy, an Arkansas high school.

Kelley's approach was radical: - Never punt - Always attempt onside kicks after scoring - Go for two-point conversions frequently

His reasoning: keeping possession maximizes offensive opportunities. Field position matters less than having the ball.

The results? Pulaski Academy won seven state championships under Kelley, routinely outscoring opponents by wide margins.

Critics noted that high school football differs from college and NFL ball. Skill disparities are larger. Onside kick success rates are higher against less-prepared teams. But Kelley demonstrated that aggressive application of expected value principles could work at his level.

The NFL Vanguard

At the professional level, the Philadelphia Eagles under Doug Pederson became known for aggressive fourth-down decisions, culminating in the famous "Philly Special" during the 2018 Super Bowl—a fourth-down conversion that became one of the most celebrated plays in NFL history.

Analytics-friendly coaches like Sean McDermott (Buffalo) and Brandon Staley (Los Angeles Chargers) pushed boundaries further, sometimes controversially. Staley's Chargers drew criticism for aggressive fourth-down attempts that failed in high-profile games—even as the underlying analytics supported the decisions.

College Football Adoption

The Trend Lines

College football has seen dramatic increases in fourth-down attempt rates over the past decade.

FBS Fourth-Down Attempt Rate (4th & 3 or shorter)

This represents a massive behavioral shift. Nearly half of 4th-and-short situations now result in conversion attempts, compared to less than a third eight years earlier.

Leading Adopters

Some programs have embraced analytics-informed decision-making more than others.

Lincoln Riley (USC, formerly Oklahoma): One of the most aggressive fourth-down coaches in recent history. Riley's Oklahoma teams consistently ranked among leaders in fourth-down attempts. He openly discussed using analytics to inform decisions.

Chip Kelly (UCLA): Kelly's Oregon teams pioneered tempo and analytical approaches. His philosophy: "If the math says we should go for it, why wouldn't we?"

Kyle Whittingham (Utah): Less discussed but statistically aggressive. Utah has ranked among the top 20 in fourth-down attempt rate for consecutive seasons.

Laggards and Skeptics

Not everyone has adopted. Some coaches remain conservative despite evidence:

Kirk Ferentz (Iowa): Iowa's conservative approach extends to fourth-down decisions. Ferentz has explicitly questioned analytics, stating that the models don't account for weather, momentum, and psychological factors.

Mack Brown (North Carolina): The veteran coach has been quoted saying he trusts his experience over numbers in critical moments.

The generational pattern is notable: younger coaches and those who entered coaching in the analytics era are generally more aggressive than long-tenured coaches from earlier eras.

Analysis: Why the Adoption Gap?

Factors Supporting Adoption

Clear evidence: Multiple independent analyses using different methods reached the same conclusion. The case for going for it more often is robust.

Competitive pressure: As some teams gained edges from better decisions, others faced pressure to catch up.

Changing media environment: Analytics-literate media became more likely to criticize punting decisions, reducing the career risk of aggression.

Generational turnover: Younger coaches entered the profession already familiar with expected value frameworks.

Factors Resisting Adoption

Loss aversion: The psychological pain of a failed fourth-down attempt exceeds the pleasure of success. Coaches feel the failures more acutely.

Visibility bias: Failed fourth-down attempts are obvious and criticized. Extra wins from better decisions accumulate subtly and attract less attention.

Institutional inertia: Athletic directors, boosters, and fans may still expect traditional play-calling. Coaches who deviate carry career risk.

Legitimate exceptions: Sometimes team-specific factors (poor short-yardage offense, elite punter, opponent factors) justify deviation from general analytics recommendations.

Model uncertainty: Coaches can (sometimes correctly) argue that their situation differs from the historical data underlying the models.

Calculating Expected Value: A Worked Example

Consider a specific decision: 4th and 2 from your own 40-yard line, neutral game situation (tied in second quarter).

Option 1: Go For It

Historical conversion rate for 4th and 2: approximately 60%

If successful (60% probability): - You have 1st and 10 from opponent's 42-yard line - Expected Points from opponent's 42: approximately +2.3 EP

If unsuccessful (40% probability): - Opponent has ball at your 40-yard line - Expected Points for opponent there: approximately +2.1 EP - Your perspective: -2.1 EP

Expected Value of Going For It: $$EV_{go} = (0.60 \times 2.3) + (0.40 \times -2.1) = 1.38 - 0.84 = +0.54 \text{ EP}$$

Option 2: Punt

Average net punt from your 40: approximately 40 yards Opponent receives ball at their 20-yard line

Expected Points from opponent's 20 (their perspective): - Approximately +0.8 EP - Your perspective: -0.8 EP

Expected Value of Punting: $$EV_{punt} = -0.8 \text{ EP}$$

Decision

$$EV_{go} - EV_{punt} = 0.54 - (-0.8) = +1.34 \text{ EP}$$

Going for it is worth 1.34 expected points more than punting in this situation. Over a season, a team that makes better fourth-down decisions in situations like this gains multiple expected points—translating to additional wins.

Discussion Questions

1. Why do you think college football adoption has been faster than NFL adoption in recent years? Consider factors like coaching tenure, media environment, and organizational structure.

2. Is there a threshold of conservatism where deviation from analytics becomes clearly wrong? If a coach punts on 4th and 1 from the opponent's 30-yard line, is that a fireable offense?

3. How might analytics-informed fourth-down decisions change opponent defensive strategy? If teams go for it more often, how should defenses adjust?

4. What would you need to see to convince a skeptical coach that analytics-informed decisions are better? What evidence would be most persuasive?

5. Are there ethical considerations in recommending aggressive fourth-down decisions? Consider the pressure on players executing in high-stakes moments.

Your Turn: Mini-Project

Option A: Historical Analysis

Pick a team from the 2023 season. Research their fourth-down decisions throughout the season: - How often did they go for it in various situations? - Use an expected points model to evaluate whether they were too aggressive, too conservative, or about right - Estimate how many points they gained or lost through fourth-down decisions

Option B: Decision Model

Build a simple fourth-down decision model: - List the key variables that should influence the decision - Create a decision tree or flowchart for 4th and short situations - Identify where your model might need team-specific adjustments

Option C: Interview Study

Interview a football coach (any level) about fourth-down decision-making: - What factors influence their decisions? - How do they think about analytics versus experience? - Have their approaches changed over the years?

Write a 500-word summary comparing their perspective to the analytics framework.

Key Takeaways

1. Fourth-down conservatism persisted for decades despite being suboptimal, demonstrating how culture and incentives can sustain inefficient behavior.

2. Analytics provided clear evidence that teams punt too often, especially in neutral game situations near midfield.

3. Adoption has been significant but uneven. College football has seen dramatic increases in fourth-down attempts, but variation across coaches remains substantial.

4. Expected value frameworks quantify the tradeoffs between going for it and punting, accounting for success probabilities and field position outcomes.

5. Behavioral factors including loss aversion, visibility bias, and career incentives explain why optimal strategies take time to spread.

References

• Burke, Brian. "Expected Points and Expected Points Added Explained." Advanced Football Analytics.
• Romer, David. "Do Firms Maximize? Evidence from Professional Football." Journal of Political Economy, 2006.
• The New York Times Fourth Down Bot (interactive tool)
• Lopez, Michael. "Football Analytics: Fourth Down Decision-Making in the NFL."