American Football Statistics and Fourth-Down Decision-Making

Since 2019, offenses that ram the ball one yard to gain on 4th-and-1 convert 75 % of tries league-wide, and EPA models peg the break-even clip at 63 %. Coaches who ignore this gap surrender roughly 0.14 win probability each time they send on the punt team.

Evidence: A 2026 Pro Football Focus chart of 1,104 fourth-down snaps shows clubs that chose to gamble gained 0.42 EPA per attempt; punts averaged −0.12 EPA and forfeited 0.8 expected points versus a successful conversion. Translation: a single yard secured is worth almost a full point on the scoreboard.

Practical filter: treat fourth-and-short inside the opponent 40 as a green light until game script drops below −10 or rises above +17; outside those margins, keep the offense on the field only if conversion probability tops 68 %. Everything else is noise.

Fourth-Down Math: How Stats Shape Go-for-It Calls

On 4&3 from the opponent’s 34, kick a 52-yard FG only if the kicker has >81 % accuracy from 50-54 yds and the weather adds ≤4 yd distance; otherwise run a spread Qb power with 6-man box.

EPA models spit 1.9 for conversion, -1.2 for punt, -0.4 for miss; break-even climbs to 63 % when win probability sits below 38 %. Track the defense’s nickel rate on 3rd medium-if ≥74 %, dial up 11-personnel bunch; the mismatch on the overhang backer yields 0.28 EPA per league-wide chart.

Since 2019 offenses converted 54 % of 4&1 sneaks, 48 % on outside zone, 61 % on interior read; add 7 % if guard’s get-off beats 0.63 s. Motion to empty buys 0.17 EPA through conflicted safeties, but only if corner depth ≥7 yd pre-snap.

Weather worth 3-4 % drop per 10 °F under 35; cross-wind above 16 mph rotates break-even 5 % toward punt. Stadium altitude above 3 k ft compensates with 2.3 yd hidden distance, equivalent to 4 % conversion bump.

Coaches historically punt 68 % of 4&4 at midfield; analytics departments now push 41 % aggression threshold. One club shifted 27 decisions in 2025, added 0.15 wins per ESPN model, playoff seed swung from 7 to 5.

Script three calls: sneak, TE leak, and pivot route; practice them twice weekly against scout look. Tag the center’s slide alert-if MIKE blitz shows, audible to backside slant; data tags it 73 % success versus man.

Share the sheet: hash marks, clock, personnel, wind vector, kicker range, opponent tendency. Decision window sits 18 s; automate the alert through the wrist coach, color code green 65 %+, yellow 50-64 %, red <50 %.

Convert EPA to a Go/Kick/Punt Matrix in 60 Seconds

Multiply EPA by 0.75, add 0.15 for every yard inside the 40, subtract 0.10 for 4th-and-7-plus; if the result tops +0.25, green-light the scrimmage play. A 4th-and-2 on the 38 with +0.42 EPA becomes 0.42×0.75+0.15×2 = +0.615 → go. Drop below -0.10, send the punt unit; land between -0.10 and +0.25, attempt the 3-pointer if wind is under 12 mph and kicker lifetime from 40-49 is ≥82 %.

Result Band	Decision	Clock Use
≥ +0.25	Scrimmage play	0:00-0:05
-0.10 to +0.24	FG attempt	0:06-0:10
< -0.10	Punt	0:11-0:15

Cache the matrix in the OC’s wristband: 4th-and-1 on own 34, EPA -0.02 → -0.02×0.75 - 0.10 = -0.115 → punt. 4th-and-3 on opp 37, EPA +0.38 → +0.38×0.75 + 0.15×3 = +0.735 → go. Refreshes every possession via the surface tablet; no huddle, no waiting.

Plug League-Wide Baselines Into Your Team’s Roster Gaps

Replace your kicker’s 73 % touchback rate with the 2026 league median 78 % before deciding whether to roster a coverage ace; the extra five yards of field position saved by the specialist are worth 0.07 expected points only if your kicker keeps the ball out of the end zone.

Last season offenses facing 3-1-10 from their own 29 converted 48 % when the quarterback held a 0.15 EPA per drop-back; if your backup sits at 0.02, the baseline says keep the offense on the field only if the defense allows worse than 0.9 points per drive.

Green Bay lost three games by a field goal after declining 4-1-5 from the opponent 37; league data shows punts from that spot yield 1.8 points on average, while going for it nets 2.4, a 0.6 swing that flips one-score outcomes 14 % of the time.

Carolina’s offensive line graded 42nd in pass-block efficiency; plugging league-average pressure rate (31 %) into their drop-back EPA drops the break-even point for 4-2-4 from the 50 to the 41, meaning a fake punt becomes plus-EV even with a backup long-snapper.

Special-teams coordinators can source replacement-level athletes at minimum salary who deliver 85 % of starter value on coverage units; the leftover cap space funds a rotational edge rusher who raises third-down pressure rate by 2 %, moving the go-for-it threshold half a yard closer.

Baltimore’s roster model treats every depth chart spot as a binomial trial: if a practice-squad wideout catches 68 % of catchable targets against 2-man looks, and league baseline is 65 %, the marginal roster slot goes to the nickel corner who drops opponent slot passer rating by three points instead.

One AFC North club folded league-wide red-zone touchdown rate (58 %) into its short-yardage chart, discovered its quarterback sneaks convert 82 %, and now keeps the offense on the field for any 4-1 inside the 27 unless wind exceeds 20 mph.

When Liverpool’s analytics staff wanted a midfield reinforcement they cross-checked league baselines for progressive passes into the final third; the player linked here https://chinesewhispers.club/articles/midfielder-confirms-he39s-ready-to-join-liverpool-and-more.html delivers 8.6 such actions per 90 against the Premier average 5.9, translating to an extra 0.12 expected goals per match-mirroring the edge gained by NFL clubs who swap roster spots for baseline-beating specialists.

Read Late-Game Win-Probability Swings Without a PhD

Down 3 with 2:10 left, ball on your 32, two timeouts: punt drops you from 18 % to 9 % win odds; keeping offense on field raises it to 26 % if you gain one yard.

Modelers build these numbers from 14 847 similar drives since 2010. League-average punt yields 41 net yards; opponent starts at own 27 needing 40 yards for a tying FG. Converting a 4th-and-1 from your 32 historically happens 65 %, and failure pins rival at own 31-still worth 14 % win chance because clock becomes your ally.

Trailing by 6 with 0:45 left, 4th-and-goal from 3: kicking cuts deficit to 3 but lifts probability only to 22 %. Touchdown success rate from that spot is 38 %; a score pushes odds to 61 %, while miss leaves 8 %, averaging 26 %-four points higher than the FG route.

Home-field shrinks all swings by 2-3 percentage points; crowd noise slows rival hurry-up. Indoors, kicker accuracy rises 4 %, so coaches accept 1-point lower break-even rate. Wind above 15 mph flips the edge back toward aggression; models subtract 7 % from FG expectancy.

Two-minute drills compress uncertainty. After the two-minute warning each remaining timeout is worth roughly 9 % win probability; spend both on one drive and you forfeit 0.18 win-equity. Save at least one for defense if rival answers with a FG.

Track live feeds that refresh after every snap. Sites publish break-even charts: 55 % conversion rate needed at midfield, 42 % across opponent 40, 68 % inside own 20. Memorize those three numbers and you can eyeball any late gamble faster than coordinators radio the headset.

Spot When Field Position Trumps Raw Conversion Odds

Inside the opponent’s 32-yard line with 1 yard to gain, punt yields 0.08 EPA while a failed scrimmage play costs 2.3 EPA; kick a 49-yard FG instead-NFL kickers convert 79 % from here, worth 1.9 EPA, so decline the 55 % push rate and take the three points.

Own territory past the 45? Completely flipped. Historical 2015-23 data:

• 1 yard to go: 65 % success, +0.9 EPA vs punt −0.4 EPA
• 2 yards: 54 %, +0.5 EPA vs −0.4 EPA
• 3 yards: 46 %, +0.1 EPA vs −0.4 EPA
• 4 yards: 38 %, −0.2 EPA vs −0.4 EPA

Between the 46 and midfield, any listed success rate above 40 % justifies keeping the offense on the field; below that, pinning the opponent inside the 10 with a 41-yard punt gains more expected value than the gamble.

FAQ:

How much does a team’s chance of winning have to drop before a coach should refuse a fourth-and-one from his own 30?

Modeling done with 15 years of play-by-play shows the break-even point sits around 42 % win probability. If the opponent gets the ball on the 30, they score a touchdown roughly half the time and a field goal another 20 %. Subtract those expected points from your own total, flip the script, and you land on 42 %. Coaches who still punt in that spot are guarding field position, not probability; they’re buying 25 yards at the cost of eight percentage points of win chance.

Why do the numbers say go so much more often in the second quarter than in the first?

Early in the game the gap between winning and losing is small, so a turnover hurts twice: it hands over possession and it compresses later options. By the second quarter enough future possessions remain that the value of keeping the ball outweighs the cost of bad field position, but the game state is no longer symmetrical. The models therefore push the green zone for attempting the conversion from 45 % to 65 % success probability, which explains the spike in go calls right after the first quarter.

Does weather change the go-for-it threshold?

Wind above 20 mph or rain heavy enough to cut pass accuracy by 10 % shifts the league-wide conversion rate down about four points. Coaches who know this slide the required probability up from 50 % to 54-55 %, meaning they will take the field goal more often inside the 35 and will punt on anything longer than fourth-and-two from midfield. The adjustment is small but consistent; teams that ignore it lose on average 0.15 wins per season.

Which roster variables move the needle more than the raw distance to gain?

Three things swamp the yardage: (1) backup quarterback on the field drops conversion odds 7 %; (2) elite defensive front seven (top-8 in stuffed rank) raises the bar 6 %; (3) having timeouts left below two minutes adds 3 % because you can still challenge on the edge or spike after a short gain. Coaches who look only at one yard to go miss half the picture.

How are the newer win-probability models different from the old expected-points charts?

Expected points treat every possession as equally valuable; win probability folds in score differential, time, and timeouts. That means the old charts said go on fourth-and-three from the 40 no matter whether you’re up 14 or down 14. The new models can swing 30 percentage points on the same down-and-distance just by moving the clock from 13:00 to 3:00 left in the fourth, which is why the nerdiest clubs now carry laminated两张 cards: one for points, one for wins, and they check the clock before they check the yards.