Fourth-Down Decisions: When the Math Says Go For It

For most of football history, fourth down was barely a decision. The chart was unwritten but universal: inside your own territory you punt, in the gray zone you punt, near the fringe you kick, and you only go for it when you're desperate. Coaches who deviated got second-guessed; coaches who punted on fourth-and-1 from the opponent's 40 got a pass, because that was simply what everyone did. Then the math arrived, and it turned out the conventional chart was leaving points — and wins — on the field, drive after drive. Fourth down is now the most-studied decision in the sport, and the gap between what the numbers recommend and what coaches actually do is one of the most revealing things in football.

The one-sentence version: a fourth-down decision is a comparison of three expected outcomes — go, punt, or kick — measured in win probability, not yards — and once you frame it that way, the aggressive choice is correct far more often than tradition ever allowed.

The three choices, and what each is worth

On every fourth down an offense has exactly three options, and each has an expected value you can estimate:

• Go for it. With some probability you convert and keep a live drive at a new spot; with the rest you fail and hand the opponent the ball where you stand. The value is a blend of those two futures, weighted by your conversion chance — which depends almost entirely on distance to go.
• Punt. You give up the ball but push the opponent backward, trading possession for field position. The value is whatever the opponent's expected outcome is from their new, worse starting spot.
• Kick a field goal. With some probability (a function of distance and your kicker) you bank three points; with the rest you miss and hand the opponent the ball near the spot of the kick — often better field position for them than a punt would have allowed.

The right call is simply whichever option has the highest expected value. The revolution in fourth-down thinking wasn't a new option; it was learning to actually compute these three numbers instead of defaulting to the punt.

From expected points to win probability

The first cut uses expected points: every spot on the field, for each team, has an average eventual point value learned from thousands of historical drives. Going for it on fourth-and-1 swaps your current expected points for a probability-weighted mix of "converted, better spot" and "failed, opponent's ball." Punting swaps it for the negative of the opponent's expected points from their new spot. Comparing those two is the classic fourth-down chart in a nutshell.

But expected points isn't the whole story, because not all points are worth the same. Up three with two minutes left, you don't care about average points — you care about winning. That's where win probability comes in. Win-probability models estimate, from the score, time, field position, down, and distance, the chance the team eventually wins. The best fourth-down decisions compare the three options in win-probability terms, which automatically handles game state: late and trailing, you go for it in spots you'd never dare while tied in the first quarter; protecting a lead in the final minutes, you punt in spots the early-game math would call a clear go. Expected points tells you the textbook answer; win probability tells you the right answer for this moment.

Why the math leans aggressive

The consistent finding across this research is that conventional coaching has been too conservative — that teams should go for it considerably more often than they traditionally have, especially in two zones. The intuition behind it:

• Short yardage is easier than it feels. Fourth-and-1 converts at a high rate — offenses sneak, run power, or play-action while the defense has to honor everything. When the conversion probability is high, the expected value of going usually beats both the modest field-position gain from a punt and the three points from a long field goal.
• Punting from the opponent's side of midfield barely helps. Out near the opponent's 35-to-40, a punt nets only a short field-position swing — you can't down it deep without risking a touchback — so you're surrendering a possession to gain very little. Going for it, even at longer distances, often wins the comparison.
• A failed long field goal can be worse than a punt. Miss from 50 and the opponent takes over near the spot of the kick, better field position than a punt would have left them. That hidden cost makes the "safe" field goal less safe than it looks, and the go-for-it option relatively more attractive.

None of this means "always go for it." Deep in your own territory, the downside of a failed conversion — handing the opponent a near-automatic scoring chance — is so large that punting is clearly right. The aggressive shift is concentrated in short-yardage and in the zone just past midfield, exactly where tradition was most timid.

An illustrative example: same fourth-and-2, three game states

Picture an offense facing fourth-and-2 at the opponent's 38. The "right" call flips with the situation. The labels below describe the direction the math points, not measured probabilities from any real model:

Same down, distance, and field position; three different correct answers. That's the entire point of the win-probability framing: the decision isn't a fixed chart entry, it's a function of the score and the clock. (Again, these scenarios are illustrative, not model outputs.)

Why aggressiveness varies so much from coach to coach

If the math is this clear, why doesn't every coach follow it? Because the incentives aren't purely about expected wins. A few forces push coaches back toward the punt:

• Asymmetric blame. A failed fourth-down gamble is vivid and personal — the coach "cost" the team. A punt that quietly lowers win probability is invisible; nobody gets fired for punting. The safe-looking choice protects the coach even when it hurts the team.
• Roster reality. A team with a dominant offensive line and a mobile quarterback converts short yardage far more often than one that can't, so the correct aggressiveness genuinely differs by personnel. Going for it is a better bet when you're more likely to convert.
• Game context the chart can't see. Weather, a gassed defense, an injury, momentum — real information a coach has and a generic model doesn't. Sometimes a "wrong" punt is right for reasons outside the inputs.

The result is a wide spread: some staffs have leaned hard into the analytics and go for it at rates that would have been unthinkable a decade ago, while others still play the old chart. Measuring that spread — how often a team goes for it relative to what a win-probability model recommends — is now a standard way to characterize a coaching staff's identity.

How to measure fourth-down decision-making

To grade a team's fourth-down choices honestly, you don't ask whether the gambles worked — outcomes are noisy, and a correct decision can fail. You ask whether the decisions were right given the information at the snap. The method: for every fourth down a team faced, use a win-probability model to compute the expected win probability of go, punt, and kick, identify which option the model preferred, and record what the coach actually did. Aggregate it and you get a "go-for-it rate over expected" or a total win-probability gained-or-lost from fourth-down calls across the season. That separates process from luck: a coach who consistently picks the higher-win-probability option is making good decisions even if a few of them backfired, and a coach who punts into low-win-probability spots is bleeding small edges no matter how tidy the punts look.

How to use it

• Judge the decision, not the result. A correct go-for-it can fail and a wrong punt can look fine — grade the choice against the win probability at the snap, not the outcome.
• Let the score and clock drive it. The same fourth-and-2 has different right answers tied in the first quarter, trailing late, and protecting a lead. There is no single chart.
• Adjust for the roster. Teams that convert short yardage well should be more aggressive; the right rate isn't the same for everyone.
• Remember the field-goal trapdoor. A missed long kick can hand the opponent better field position than a punt — the "safe" three points carries a hidden cost.
• Read aggressiveness as identity, not heroism. A staff that goes for it more isn't gambling; it's usually just following the math the conservative staffs are leaving on the table.

Fourth down used to be where coaches stopped thinking and reached for the punt team. Now it's where the thinking is sharpest — three options, each with a number, compared in the only currency that matters at the end of a game: the probability of winning. The chart was never wrong because it was a chart; it was wrong because it stopped at field position. Put win probability on the axis, and "go for it" turns out to be the answer far more often than the punt team ever wanted to admit.