East Carolina never gained seven yards on a first down

For about ninety seconds in the second quarter, East Carolina had a game. On second-and-4, Mitch Griffis threw deep to Tyler Johnson and the ball came down at the Alabama 1: three plays, seventy-five yards, the kind of drive a 28-point underdog builds a whole night's belief on. Jeremiah Melvin scored from a yard out. Alabama 17, East Carolina 7, 13:41 left in the half.

Alabama scored three more touchdowns before halftime, one of them without putting its offense on the field, and it was 38-7 at the break. "Hat's off to Coach DeBoer and the Alabama football team," Blake Harrell said afterward. "Thought they did what good football teams do. When you make a mistake, they take advantage of it. They go put the ball in the end zone. They three-and-out you, and take control of the game. And they certainly did that."

What we had, and what we missed

Our margin model made Alabama the pick before kickoff by 18.2 points, with an 88 percent chance of winning. Almost all of that came from one thing: raw roster talent, worth 11.1 points by itself. Last season's body of work added 3.4 more. Returning production added 1.3. Then the model handed 2.5 points back to East Carolina at quarterback, because Griffis had started college football games and Keelon Russell had never started one.

Quarterback is precisely where Alabama won the game. Russell's 335 total yards were the largest single input to the blowout, and the one position the model had scored as an Alabama weakness was his.

Right side, wrong size. We had Alabama by 18; Alabama won by 38. Our 80 percent range ran from a one-point East Carolina win to a 38-point Alabama win, so the result landed on the outer edge of what we thought plausible rather than outside it. That is cold comfort. The market, at Alabama minus 28.5, was closer to the truth than we were, our against-the-spread lean was East Carolina, and it missed by nine and a half points of margin. That one goes in the loss column and stays there.

What the model did get right was the shape of the mismatch. It read Alabama's offense against East Carolina's defense as a strength and East Carolina's offense against Alabama's defense as a weakness, and both showed up almost exactly as described: Alabama at plus 0.29 expected points per play on a 52 percent success rate, East Carolina at minus 0.20 on 22 percent. The units behaved. The scale is what we underpriced.

The quarterback ran, and a ceiling moved

Russell's official line was 13 carries for 86 yards and a touchdown. Strip out the sacks, which the play-by-play files as passes, and he had 11 carries for 97 yards, 23 percent of Alabama's rushing attempts and 43 percent of its rushing yardage. No Alabama running back topped nine carries or 47 yards. In a game Alabama won on the ground, the quarterback was the leading rusher, and it was not close.

Alabama has not played this way. Across all 14 games last season, its quarterbacks (Ty Simpson, with cameos from Austin Mack and Russell) took 15.6 percent of the team's non-sack carries for 383 yards, about 27 per game. The single-game high for the entire room was six carries for 59 yards. Russell's 97 in one night is more than one and a half times that season best and three and a half times the per-game average. One caveat, because a bad version of this comparison is circulating: do not measure it against Simpson's official 90-carry season line. The NCAA box charges sacks to the quarterback's rushing column, and Simpson's actual rushing was 56 carries for 319 yards. The apples-to-apples number is 97 in one game against a 59-yard season high.

So what does a night like Russell's usually mean? I built the comparison structurally rather than by reputation: games in which a team's primary passer rushed for 85-plus official yards, when that program's quarterback room had averaged 40 or fewer rushing yards per game the prior season, and when the single game beat the program's entire prior-season single-game high. That screen catches the night a ceiling breaks rather than the fifth repeat by an established running quarterback, and it returns 32 games. Those teams went 25-7 with an average margin of nearly ten points, and 22-10 against the spread, beating the closing number by 6.6 points a game. The market was a touchdown behind each of those 32 offenses.

What followed is where the story narrows, and it narrows honestly. The running stuck; the improvement did not. Those quarterbacks averaged 12.1 carries and 46.9 rushing yards in their remaining starts, still more than three times their program's old norm, but only 16 of the 32 ever posted another 85-yard rushing game. The offense as a whole graded 0.245 expected points per play in the breakout and 0.157 afterward, against 0.154 before it. Volume persisted. A better offense did not. Read forward, that puts Russell's legs permanently on Ryan Grubb's menu at roughly a dozen carries a week, and it makes 335 total yards a ceiling rather than a new floor.

Kalen DeBoer's postgame worry was a different one. Russell did not slide at the end of his runs, and the head coach said flatly, "It's something we'll talk about a lot." The 2025 data cannot rule that fear out, and it does not support it either. Across 3,033 quarterback-games last season, a starter held the job into the following week 84.3 percent of the time, and the rate barely moved with workload: 85.2 percent for quarterbacks who carried four times or fewer, 83.8 percent for those who carried fifteen or more. The 32 breakout quarterbacks started their next game 90 percent of the time. The risk these numbers flag is not a broken quarterback. It is that the group won the night and then went 58.8 percent the rest of the way with an offense grading below the FBS average, because the quarterback run was the whole gain. With 32 games from a single season, and position inferred from who threw the most passes, that is a tendency and not a law.

Alabama inverted its own staircase

Here is the strangest thing in Alabama's box score, and it should bother DeBoer more than the timeouts did. Third down was 18.8 percent of Alabama's snaps and produced 78 percent of its offensive value: fifteen third downs, 10.5 yards a play, two-thirds of them successful, 18.24 of the offense's 23.37 total expected points added. First and second down, 64 snaps and the actual body of the offense, produced 4.7 yards a play and plus 0.04 expected points per play. Last season those same early downs were worth plus 0.104. Against an opponent Alabama was favored to bury, the early-down offense got worse.

And the down Alabama threw on more than 70 percent of the time last season became a run down. It ran eight of its fifteen third downs, against a 28.9 percent third-down run rate in 2025. Russell himself carried on five of them, a third of all Alabama third downs, where quarterback runs accounted for 12 percent of them last year. On third-and-6 or longer, Alabama ran four times and converted all four, for 56 yards. This was not fourth-quarter housekeeping: through three quarters Alabama still ran half its third downs, and all five Russell third-down carries came before the fourth. Four of the eight largest swings in the entire game were Alabama third downs: a 58-yard throw to Ryan Coleman-Williams on third-and-11, Russell for 23 on third-and-6, Russell for 17 on third-and-10, and Russell's three-yard touchdown on third-and-3.

Rolled up to the definitions used across eleven seasons of team-game data, this was a behind-schedule offense: 23 snaps in trouble producing 74 percent of the night's value at a 47.8 percent success rate, which is 91st percentile, while the on-schedule offense moved the chains at a fine rate and generated explosiveness in the 23rd percentile. Teams that converted in trouble at a top-decile rate, stayed on schedule at a median-or-better clip, and still could not threaten on early downs: 646 team-games since 2014, 2.6 percent of all of them, spread across 198 programs. They won 81.6 percent by an average of 16.4 points and covered 62.5 percent. That reads like a formula until you meet the control group: same money-down brilliance plus a functioning early-down offense, 1,294 games, won 89.5 percent by 26.7. Roughly ten points of margin and eight points of win probability is what Alabama left on first and second down.

The profile does not stick either. In the following game, behind-schedule conversion in that set fell from 52.4 percent to 34.6 percent, which is simply those teams' own season average, and only 18.2 percent managed it twice. Overall offensive efficiency dropped straight back to the norm. The win rate also splits hard by opponent: against teams within 150 Elo points, those analogs went 65.9 percent at a plus-6.1 margin; against clearly weaker opponents, 91.3 percent at plus-19.9. The forward question for Alabama is not whether Russell can make a third-and-11 throw. He made it. It is whether first and second down can produce more than plus 0.04 a play on a night when the third downs stop converting. A bookkeeping note, since the seam is better shown than hidden: the official tally has Alabama 7-of-15 on third down and my play-by-play derivation has 8-of-15, one play scored differently. The run-pass split is unaffected.

East Carolina's 0-for-9 is the symptom. First down is the disease.

Every recap of this game will lead with 0-for-9 on third down, and every one of them will be describing a consequence. East Carolina took 17 first-down snaps and gained 28 total yards on them, 1.65 a play. The longest first-down gain of the entire night was six yards. Not one first-down snap gained seven. Last season ECU gained 7-plus on 30 percent of its first downs and averaged 6.16 yards a play on the down, against an FBS norm of 6.05. That collapse is what manufactured an average third down of 9.4 yards, six of the nine at seven or more. Their only third-and-short all night, third-and-1 from their own 34, was an Ashton Gray run stuffed for no gain.

The shape of what little East Carolina did produce is stranger still. All three of its explosive gains came on second down: the long ball to Johnson on second-and-4, a 30-yard completion to Melvin on second-and-25, and a 23-yarder from backup Emory Williams to Melvin on second-and-9. None on first, none on third. Those three snaps produced 121 of the offense's 162 yards, three-quarters of the yardage on 7.5 percent of the snaps. The other 37 snaps produced 41 yards, 1.11 apiece, against 3.54 on non-explosive plays last season.

A team taking twelve or more first-down snaps without a single first-down gain of seven yards has happened nine times in FBS play since 2015, nine out of 17,768 team-games, roughly one every fifteen months. Those nine went 1-8 straight up and 1-8 against the number, averaging eight points and 168 yards. The lone win is instructive: Navy beat Temple 27-20 in 2022 with the same first-down shutout, but its average third down was 4.65 yards, because an option offense can grind to manageable distances without ever popping one. Tighten the screen to teams that also faced East Carolina's brand of third-and-long, and it is five games, 0-5, average margin minus 36.2. Separately, 120 team-games in eleven seasons had three or more explosives with every one of them on second down; that group went 43-77 and still averaged 3.77 first-down yards a play, more than double East Carolina's. No team has done both things in the same game in the eleven seasons of play-by-play I can check.

Harrell named third down himself: "On the opposite side of the ball, we get in a third-down situation; we got to keep the stick moving. It can't be a 40-second drive or only 41 plays in a game." He also named the bigger problem, and named it first: "We talked earlier in the week about things that we gotta do to win a football game, and that was run the football and stop the run, and we certainly didn't do those two things. I think that's what stands out to me the most." Alabama ran for 227 yards on 49 carries with five rushing touchdowns.

One line of Harrell's is worth checking against the drive chart. He said Alabama got "some easy short fields or some easy scores," and the second half of that is exactly right. Pierre's pick-six was a score that required no field at all. "I sat and read the quarterback's eyes, and there it was," Pierre said of it. But starting field position was almost dead even: Alabama began its drives an average of 71.6 yards from the end zone, East Carolina 72.3. East Carolina did not lose this on where it started. It lost it on first down, and then again on the down that first down created.

Eight punts and not one fourth down

East Carolina's fourth-down log reads: eight punts, one made field goal, zero offensive plays. Twice it faced fourth-and-1 and punted both times, from its own 34 in the first quarter trailing 3-0, and from its own 26 in the third quarter trailing 38-10. That is out of character. Counting only the fourth-down decisions that were actually decisions last season (an offensive play, a punt or a field-goal attempt, stripping out the timeout and penalty rows that pollute a raw count), ECU went for it on 35 percent of its fourth downs, and at fourth-and-1 or shorter it kept the offense on the field 11 of 13 times. Eighty-five percent. Nor was this a bad offense making a virtue of necessity: ECU's .453 success rate in 2025 was 83rd percentile in FBS, comfortably above the .414 mean.

Two ECU punts are not a philosophy, and the confidence on this one is low. An offense gaining 1.65 yards a snap on first down rarely produces a fourth-and-short worth taking, and the two it did produce came from its own 34 and its own 26, the spots where even aggressive staffs punt. Tomas O'Halloran punted beautifully, with 52, 55 and 57-yarders among them, which quietly made field position East Carolina's least pressing problem. Alabama faced exactly one fourth down all night, went for it on fourth-and-2, and gained 16.

For the broader situation, a good offense erased early in the season by a much stronger opponent, there are 33 comparable team-games since 2014, and they say this mostly does not stick. Those teams' success rates cratered from a .467 prior-season mean to .344 in the blowout, then recovered to .439 for the rest of the year. They won their next game 73 percent of the time. But the recovery was partial: only 8 of 33 got back to their own prior-year offensive level, and barely half finished above .500. The tell for East Carolina is not the score, and it is not the punts. It is whether that 85 percent fourth-and-1 rate reappears next week against an opponent it can actually move.

Five backs, and none of them the back

Alabama came out of the opener without a lead running back at all. Thirty-two running back carries split five ways: Daniel Hill nine for 47 and two touchdowns, EJ Crowell seven for 37 and a touchdown, Trae'shawn Brown seven for five, Kevin Riley six for 17, Khalifa Keith three for 12. The top back held 28.1 percent of the carries. In Alabama's previous 15 games, that has happened twice.

Three injuries shaped that room before kickoff. A.K. Dear was hurt in Alabama's first fall-camp scrimmage and was medically unavailable, with a reported return in late September or early October. Receiver Noah Rogers was out with a fall-camp injury on a similar timeline. Jeremiah Beaman tore an ACL in practice before the opener, had surgery, and will miss the entire season, his second torn ACL in as many years after a 2025 in which he played the Florida State opener and then missed the final fourteen games. Crowell's was the status that resolved: a game-time decision with an ankle issue, about whom offensive coordinator Ryan Grubb had said on the Monday of game week, "he's like Wolverine. Him and Noah Rogers are on a different planet... I didn't see any setbacks yesterday with EJ. He was cutting off with both of his feet." Crowell dressed, played and scored.

Availability does not explain the distribution. Riley, a healthy returning back with 59 carries last season, took zero handoffs in the first half and all six of his after halftime, while staying on the field into the fourth quarter. Riley looks like a back being evaluated rather than trusted. Brown was the only back to carry in all four quarters, seven carries for five yards, 0.71 a pop, and he still got the two-yard touchdown, which is what a backfield sorted by situation rather than by pecking order actually looks like.

Across 3,295 FBS team-games in the two seasons the official player box covers, 127 met the committee bar: five or more non-quarterback rushers with three-plus carries and a top back under 35 percent. Those teams went 108-19, which sounds like a formula until you notice that 88 of the 127 were 21-point wins. The record is the scoreboard doing the splitting, not a plan: the fourteen committee games actually decided inside two touchdowns went 10-4 at an average margin of 2.9. The useful subset is the 34 that happened in week 1, and it says a committee is a one-week condition. Median opening-week top-back share in that group was .280, essentially identical to Alabama's .281. The very next week it jumped to .467, and only five of the 34 were still fragmented. Thirty-three of 34 found a 50-percent lead back at some point, median arrival week 3.

Alabama is its own best comparison here. Its 2024 opener sits in that set at a .257 top-back share, Daniel Hill led it with nine carries, and Jam Miller, five carries that day, went on to 130 carries as the team's rest-of-season leader while Alabama went 8-2 the rest of the way. Which points at the most useful piece of the pattern: only 14 of the 34 opener leaders stayed the leader. In the 20 cases where it changed, the eventual lead back had a median of six carries in the opener, the exact band Riley, Crowell and Brown occupy right now. Two caveats, both worth stating: this table covers only two seasons and has no position column, and this game's carry counts came from play-by-play rather than a final box. Hill's last on-field trace is the second quarter, which could be a rotation call or an early exit. The box will not say.

The parts that weren't clean

For a 38-point win, the night had rough edges, and Alabama named them itself. It burned timeouts on pre-snap operation early. "I think there were some operational things pre-snap that caused us to use timeouts," DeBoer said. "But it was to be expected." Russell took the whole thing on himself: "I'm going to blame everything that happened operations wise on me." Alabama also lost a fumble in the third quarter, on a snap that began with Russell losing seven yards.

And Ryan Coleman-Williams, who led all receivers with five catches for 130 yards including the 58-yard third-and-11 conversion that was the single biggest play of the game, dropped two passes in the second quarter. DeBoer did not flinch. "We're not going to shy away from going to Ryan," he said, and later: "Yeah, he went out there, made some contested catches too. You know we're gonna keep feeding him." On a night when Alabama's early-down offense produced almost nothing, one receiver produced 130 of the 249 passing yards. That is a dependency, not a rotation.

What to actually watch

None of this changes the result, and none of it should be oversold. Two of the comparison sets above are drawn from a single season, one of them has nine games in it, and week 1 is the noisiest week on the calendar. What the history supports is a short list of four things to check in two weeks, not a verdict.

• Alabama, first and second down. Plus 0.04 expected points per play against a team it beat by 38 is the flag. In the historical set, third-down brilliance reverted to the team's own season norm more than 80 percent of the time.
• Alabama, the backfield. History says a 50-percent lead back appears by roughly week 3, and that it is more often not the opener's leading carrier. Kevin Riley's second-half-only usage is the thread to pull.
• Alabama, Russell's carries. Expect the volume to stay, about a dozen a game in the comparable set, and the efficiency not to. The 335 total yards are the ceiling.
• East Carolina, first-down yards per play, not third-down conversion rate. It was 6.16 last season and 1.65 here. The analogs got that number back to league average the very next week and still went 16-39 the rest of the way.

Harrell said the honest thing while it was still fresh: "If we play the game like we did today, we're going to have a long season." The history says he almost certainly will not play it like that again next week. It also says East Carolina's season starts with getting first down back, not with third.