Behind the Sticks

Jon Sumrall addresses reporters after Florida's game against Oklahoma on Nov. 7, 2026. Frame from Florida Gators/UAA video.

Florida scored 4.44 points a drive while it was 4-0 and 1.34 in the five games since. Nine games say the offense stopped winning early downs, and three remain to find out if it can again.

By Brenden Martin

Twenty plays

The Oklahoma drive began at the Oklahoma 11-yard line with 2:41 left in the third quarter. It ended with 4:56 left in the fourth. In between, the Sooners ran 20 plays, gained 89 yards and held the ball for 12 minutes, 45 seconds.

Six of those snaps came on third down. Each one either moved the chains or scored, and the last was a 6-yard touchdown on third-and-6. Before the drive, ESPN's win probability model gave Florida a 32.3% chance to beat Oklahoma. After it, 1.7%. That is a 30.6-point swing, the largest of any drive in the game.

Here is the number that did not make most of the postgame talk. Florida held the ball for 2 minutes, 3 seconds of the fourth quarter. Oklahoma held it for 12:57. The final was 23-7, in front of 90,442 at Ben Hill Griffin Stadium, and the game was decided in a quarter in which Florida's offense barely touched the ball.

Asked afterward what he liked about his team, head coach Jon Sumrall answered in part by talking about the offense's identity. "Right now, we feel a little bit lost," he said.

Nine games into 2026, the box scores can say where Florida got lost and roughly when. This is that attempt.

The season in one paragraph

Florida is 5-4 and 3-4 in the SEC. It opened with four wins, over Florida Atlantic, Campbell, Auburn and Ole Miss, by an average of 30.75 points. It has since lost four of five, to Missouri, South Carolina, Georgia and Oklahoma, by an average of 20.5, and won at Texas by three. The 4-0 start averaged 53.5 points a game. The five games since have averaged 15.6. Florida was ranked No. 21 before it played Ole Miss and No. 8 before it played Missouri. It has not scored more than 27 points in a game since Oct. 3.

Three regular-season games remain: at Kentucky, Vanderbilt at home and at Florida State. One more win would make Florida bowl eligible, provided the Campbell win counts. The NCAA lets a team count one FCS win toward eligibility, and only if that school meets a scholarship threshold. GatorBait has not confirmed that Campbell does.

The thread

A useful season storyline has to hold up in most games and name its exceptions. After going through all nine drive by drive, here is the one that does:

Florida's offense stopped winning first and second down as the run game quit working. Rushing PPA, an expected-points measure of every carry, averaged +0.35 a game in the 4-0 start and -0.23 over the five games since.

Three numbers carry it.

First, points per drive. Florida scored 200 points on 45 drives in its first four games, 4.44 a drive. In the last five it scored 78 points on 58 drives, 1.34 a drive. (These count offensive points only. Return touchdowns are excluded, including two Vernell Brown III punt returns against Campbell.)

Second, the line between winning and losing. Florida won all five games in which it scored 2.25 points a drive or more. It lost all four in which it scored below 1.6. Opponent scoring does not sort the games that way: Florida won while allowing 3.55 points a drive to Auburn and lost while allowing 2.30 to Oklahoma.

Third, the rushing numbers do not overlap. In the first four games, Florida's rushing PPA was +0.45, +0.35, +0.21 and +0.38. In the last five it was -0.52, -0.09, -0.04, -0.50 and -0.01. The worst game of the start was better than the best game since. Weighted by carries rather than averaged by game, the figures are about +0.34 and -0.19.

Two terms carry the rest of the story. PPA is predicted points added: how many expected points a play gains or loses, per College Football Data's model. Success rate counts a play as a success if it gains 50% of the yards to go on first down, 70% on second down, or all of them on third and fourth.

By down, the shift is plain. First-down PPA went from +0.13 to -0.12, and second-down PPA from +0.64 to -0.05. Third-down PPA held up much better, +0.65 to +0.46. Florida's success rate, weighted by plays, fell from 51.6% to 37.3%. Stuff rate, the share of carries stopped at or behind the line, averaged 12.4% a game in the first four and 29.8% in the last five (13.3% and 28.7% weighted by carry).

How much of the slide was offense? About 82%. Florida's per-drive scoring margin fell 3.80 points between the two stretches. The offense accounts for 3.10 of that, from 4.44 to 1.34. Opponents' scoring per drive rose 0.70, from 1.86 to 2.55. The defense gave up more; most of the change came on offense.

What this does not say is why. The numbers show that carries stopped working. They cannot show whether the cause was blocking, play design, a change in what opposing defenses did or something else, and GatorBait has not watched the film to say. That limit comes up again below.

What we tested and threw out

A season thread earns trust by surviving the alternatives. Five did not.

The quarterback change. Aaron Philo threw most of the passes at Missouri and South Carolina, where the slide began. Tramell Jones Jr. threw every pass at Texas and Oklahoma and replaced Philo in the second quarter against Georgia. But the offense was already struggling before Jones played a full game. Florida's offensive PPA was +0.02 at Missouri and 0.00 against South Carolina. Its passing PPA at Texas, where Jones threw every pass, was +0.27; at South Carolina, where Philo threw most of them, it was +0.08. The data does not explain the rotation. The reasons for it are a question for the program.

Turnovers. Florida's turnover margin was +4 in the first four games and -7 in the last five, and it decided Georgia. It did not decide Missouri, where neither team turned the ball over. It barely moved South Carolina, where Florida's three giveaways produced only 7 points. At Oklahoma, neither team scored off a turnover. Turnovers are a chapter, not the season.

Third down and the red zone. Florida converted 21 of 43 third downs (48.8%) in the first four games and 26 of 69 (37.7%) in the last five. But third-down efficiency is downstream of early-down failure, and third-down PPA held. In the red zone, Florida scored on 23 of 24 trips early and 10 of 13 since. The finishing is worse, but the larger change is arrival: 6.0 red-zone trips a game fell to 2.6.

Time of possession and a worn-down defense. Florida's defense faced fewer snaps, not more, 70.0 a game early and 65.2 since. Possession is a result of efficiency, not a cause. The defense did slip, with opponents' success rate rising from 40.5% to 45.5%, but that is the smaller share of the story.

Opponent quality and penalties. Florida beat No. 4 Ole Miss and No. 7 Texas, and it lost to No. 25 Missouri. Ranking does not sort the games. Florida was flagged eight or more times in five games, split across wins and losses. We cannot fully rule out that the later opponents defended the run better than the early ones, and we say so below.

Chapter one: what 66-21 and 52-3 told us, and hid

Florida Atlantic arrived on Sept. 5 as a 26.5-point underdog, according to ESPN's odds, and the game played like it. Nine of Florida's 11 drives ended in touchdowns, a tenth in a field goal, and it never punted. It went 8-for-8 in the red zone and gained 624 yards. Jadan Baugh carried 14 times for 160 yards and three touchdowns, and Philo completed 16 of 21 passes for 275 yards and three more.

The hidden cost was on the other side. Florida Atlantic, trailing 45-14 early in the third quarter, ran 90 plays and converted 11 of 22 third downs. Its three touchdowns came on drives of 13, 12 and nine plays. "Our defense was atrocious at times quite honestly," Sumrall said, and then he said something larger: "I don't think we're as good as we think we are."

Campbell followed on Sept. 12 and the final was 52-3. The line is flattering. Fourteen of the 52 points came on two Brown punt returns, of 64 and 70 yards. The offense scored 38 points on 11 drives, 3.45 a drive, against an FCS defense, and went 3-for-9 on third down. On passing downs its success rate was 22.2%.

Sumrall saw it that night. "If we play like we did in the second half, we're about to have a real rude awakening," he said. He added: "What we're about to walk into in conference play, we will get absolutely embarrassed if we don't improve a ton."

The head coach said it in September, before the numbers did.

Chapter two: two proofs, Auburn and Ole Miss

The conference games that followed were wins, and they came in two shapes.

Auburn, on Sept. 19, was a 44-39 win at Jordan-Hare Stadium, and its shape was explosion. Florida's explosiveness score of 1.57 was the highest of its nine games, while its 39.0% success rate was the second-lowest of its five wins. The run game still graded positive (+0.21 PPA), anchored by Baugh's 26 carries for 162 yards and three touchdowns, including a 53-yard run on fourth-and-1 with 12:38 left in the fourth quarter that moved ESPN's win probability from 89.4% to 98.4%. Florida went 3-for-3 on fourth down.

It was also an unusual game. Florida drew 16 penalties for 138 yards and still won. Its defense sacked Auburn seven times, then allowed five touchdown drives of 75 yards or more. Florida held the ball 39:16 to 20:44 and needed all of it.

Ole Miss, on Sept. 26, was the cleanest evidence of what the run game could be. No. 4 Ole Miss came to Gainesville and left down 52-28. Florida ran 53 times for 302 yards, 5.7 a carry, with a 59.7% success rate and no turnovers. It held the ball 36:59 to 23:01.

The win was not a rout in the way the score implies. Ole Miss gained more per play, 7.0 yards to Florida's 6.6, and threw for 298 yards. Florida won on volume, 76 plays to 61, and on special teams. Its four kick returns covered 151 yards, including a 74-yard return by London Montgomery that set up a touchdown, and Jayden Woods's sack-fumble in the third quarter was recovered by Myles Graham and became a 38-21 lead. "I think we won the special teams battle," Sumrall said. He also said: "I felt it was going to take 42 points to win tonight, potentially."

That is a coach planning around his offense. For four games, it carried him.

Chapter three: the break, at Missouri

The slide begins on Oct. 3 in Columbia. No. 8 Florida, a 5.5-point favorite, lost 45-17 at No. 25 Missouri. Florida is 2-5 all time at Missouri.

The run game stopped. Florida carried 25 times for 42 yards, a figure that includes yards lost on sacks, and its rushing PPA was -0.52. A stuff rate of 36% meant more than one carry in three was stopped at or behind the line. Baugh carried 12 times for 13 yards. Of Florida's 12 drives, nine gained 25 yards or fewer. The offense scored 1.42 points a drive.

Turnovers cannot be the reason. There were none for either team. Missouri's own success rate was 45.6%, under half its plays, yet it scored 45 because its good plays were big. It had eight scrimmage plays of 20 yards or more. Two runs by Jamal Roberts, of 77 and 80 yards, produced 157 of Missouri's 220 rushing yards. Those came when Florida's win probability was already under 3%, so they widened the loss more than they caused it. The game turned before that.

The biggest swing came late in the second quarter. Florida tied the score at 10, and on the next series Austin Simmons hit Caleb Goodie Sr. for 31 yards and a 17-10 Missouri lead with 2:05 left in the half. ESPN's win probability for Florida went from 57.5% to 42.8%. Florida then ran three plays for minus-1 yard, and Missouri kicked a 50-yard field goal as the half expired. Missouri outscored Florida 10-0 in the final 4:26 of the half.

In the third quarter, Missouri hit a 36-yard pass on third-and-7 from its own 28. A 15-yard Florida personal foul, a 14-yard pass and a 5-yard Roberts run made it 27-10. Florida's win probability dropped to 12.1%.

Fairness requires one paragraph. Missouri, South Carolina, Texas, Georgia and Oklahoma may simply defend the run better than Florida Atlantic, Campbell, Auburn and Ole Miss did. The data this story uses does not include opponents' season rushing-defense numbers, so we cannot separate "the Florida run game broke" from "the Florida run game met better defenses." Ole Miss, ranked fourth, gave up 302 rushing yards, which argues against a purely opponent-driven reading. It does not close the question.

No full UF media relations transcript was available for this game, so there are no postgame quotes.

Chapter four: third-and-long, South Carolina

South Carolina visited Gainesville on Oct. 10 and won 38-19. The yardage was nearly even, South Carolina 393 and Florida 387. By the series table, it was Florida's first home loss to South Carolina since 2014, and the box score does not say where it came from.

It came from third-and-long. South Carolina converted five of nine third downs that needed seven yards or more, and five of seven that needed 10 or more. Florida was 1 of 4 on third-and-7-plus. Two of LaNorris Sellers's four touchdown passes came on third down.

Turnovers did not explain it. Florida gave the ball away three times, and South Carolina scored only 7 points off them.

The swing came in the third quarter, with South Carolina leading 17-13. On third-and-14, Sellers hit Nyck Harbor for 35 yards. On the next snap, Donovan Murph caught a 26-yard touchdown pass. The score was 24-13, and Florida's win probability dropped from 50.7% to 25.9%.

Florida's offense gained its 387 yards but went 1-for-4 on third-and-7-plus, took three sacks and stalled at the goal line, where runs of 4, 2 and 0 yards from the South Carolina 9 led to a field goal. Its rushing PPA that day was -0.09.

Sumrall's explanation was direct, and it is his, not ours. "Disappointing football today by us. No excuse for it," he said, and he said South Carolina's "defensive front won the battle at the line of scrimmage over us in the run and the pass game." We attribute that statement and do not adopt it. Without film, the numbers can show that Florida's runs were not working, but not why.

Chapter five: the win that does not fit, at Texas

On Oct. 17 in Austin, Florida beat No. 7 Texas 27-24, and by the series table it was the first Florida win in Austin. It was also the least efficient win of the season.

Florida's success rate was 33.7%, the lowest of its nine games. Its rushing PPA was -0.04. It had no play of 20 yards or more from scrimmage; the longest was a 17-yard catch by Dallas Wilson. Texas was a 13.5-point favorite, and ESPN gave Florida a 12.4% chance before kickoff.

The thread explains the offense, not that result. Florida won because of everything else. It did not turn the ball over, and Texas had one turnover. Texas went 2-for-10 on third down. Florida made five sacks and 11 tackles for loss, ran 81 plays to Texas's 50, and held the ball 38:11 to 21:49. Jones completed 22 of 35 passes for 155 yards and two touchdowns with no interceptions.

Florida was down 17-3 in the second quarter before it got going. Ty Jackson stopped Texas on fourth-and-1, Jones hit Wilson for 17 yards on fourth-and-12 and Lacota Dippre caught a 7-yard touchdown with 18 seconds left in the half.

The game turned in the fourth quarter. Up 20-17, Florida got a third-and-1 sack, and with 8:34 left Brendan Bett stopped Texas's Arch Manning on fourth-and-2. Six plays and 45 yards later, Baugh scored on an 11-yard run. ESPN's win probability for Florida moved from 38.7% to 92.5% between the 9:10 and 6:22 marks of the quarter. Later, with 2:03 to go and Florida up 27-24, Graham's sack-fumble, recovered by Kanye Clark, pushed it from 63.0% to 86.5%. Texas's Ryan Niblett also returned a punt 63 yards for a touchdown, and Florida still won.

This is the thread's exception and also its proof. An offense with a rushing PPA of -0.04 won at Texas only by being close to flawless everywhere else. That is not a plan for the remaining three games.

Chapter six: where turnovers did decide, Georgia

Georgia beat Florida 27-8 on Oct. 31 in Atlanta, and this is the one game where the turnover story is the story. Georgia leads the series 58-44-2 and has won six in a row. Florida's last win was Nov. 7, 2020, 44-28.

Florida committed four turnovers, three interceptions and a fumble, and Georgia committed one. Seventeen of Georgia's 27 points followed Florida's giveaways: a touchdown after the first interception, a field goal after the second, and a 66-yard fumble return by Chris Cole with 29 seconds left in the first half. The other 10 points came on drives of 81 and 61 yards.

The sequence that broke the game was short. Philo threw an interception in the first quarter and another in the second, which moved ESPN's win probability for Georgia from 87.6% to 93.2%, the largest swing of the game. Jones replaced him. With 29 seconds left in the half, Jones was sacked at the Georgia 34 and fumbled, and Cole took it back.

The run game did not work. Florida's stuff rate was 40%, its line yards were 0.68 a carry and its rushing PPA was -0.50. Baugh had 17 carries for 38 yards. The box score's rushing total of 54 yards includes sack losses.

The defense deserves a fair line. Georgia gained only 4.9 yards a play, and its explosiveness score of 0.55 was the lowest any Florida opponent posted in the nine games. A forced fumble by DJ Coleman, returned 42 yards by Graham, set up Florida's only touchdown. But Georgia ran for 217 yards.

There is no UF media relations transcript for this game, so there are no postgame quotes.

Chapter seven: the quarter, Oklahoma

Back to the drive.

Through three quarters of the Nov. 7 game, the possession split was 24:53 for Oklahoma and 20:07 for Florida, hardly a mismatch. The game turned in a quarter in which Florida's offense did not have the ball.

Oklahoma went 10-for-18 on third down. Florida went 2-for-9. Florida's defense was better than the score: it posted eight quarterback hurries, four pass breakups, a sack and five tackles for loss, and its success rate allowed was 46.7%. But it got no takeaways, and it could not get off the field on third down.

The decision on fourth-and-3 at the Oklahoma 17 came with 6:22 left in the third quarter. Florida trailed 13-7 and had just moved there on a 33-yard catch by Duke Clark. Jones's pass was intercepted in the end zone, and Florida's win probability fell from 41.4% to 33.0%. Sumrall explained the call. "I felt like it was going to take 14 points at least," he said. Whether to go was a judgment call that numbers cannot settle. It did not work, and then Oklahoma took the 12:45 drive described above.

Graham spoke for the locker room. "Just don't get used to this because this is not us," he said. And: "It's not on the offense. It's not on anybody but us."

Jones was specific about the fix. "We just got to sustain drives, finish on third down and it'll move on from there," he said.

Florida is now 1-2 all time against Oklahoma: the Gators won the Jan. 8, 2009 BCS title game, 24-14, and lost the Dec. 30, 2020 Cotton Bowl, 20-55.

By the numbers

Game: FAU

• Result: W 66-21

• Pts/drive, UF (opp): 6.00 (1.75)

• Rush PPA: +0.45

• Success rate, UF (opp): 54.4% (40.2%)

• 3rd down: 5-9

• TO margin: +2

Game: Campbell

• Result: W 52-3

• Pts/drive, UF (opp): 3.45 (0.21)

• Rush PPA: +0.35

• Success rate, UF (opp): 54.0% (23.9%)

• 3rd down: 3-9

• TO margin: 0

Game: at Auburn

• Result: W 44-39

• Pts/drive, UF (opp): 4.00 (3.55)

• Rush PPA: +0.21

• Success rate, UF (opp): 39.0% (43.3%)

• 3rd down: 7-15

• TO margin: 0

Game: No. 4 Ole Miss

• Result: W 52-28

• Pts/drive, UF (opp): 4.33 (2.33)

• Rush PPA: +0.38

• Success rate, UF (opp): 59.7% (55.6%)

• 3rd down: 6-10

• TO margin: +2

Game: at No. 25 Missouri

• Result: L 17-45

• Pts/drive, UF (opp): 1.42 (3.46)

• Rush PPA: -0.52

• Success rate, UF (opp): 35.4% (45.6%)

• 3rd down: 6-16

• TO margin: 0

Game: South Carolina

• Result: L 19-38

• Pts/drive, UF (opp): 1.58 (3.17)

• Rush PPA: -0.09

• Success rate, UF (opp): 45.5% (43.1%)

• 3rd down: 6-13

• TO margin: -3

Game: at No. 7 Texas

• Result: W 27-24

• Pts/drive, UF (opp): 2.25 (1.55)

• Rush PPA: -0.04

• Success rate, UF (opp): 33.7% (41.2%)

• 3rd down: 8-19

• TO margin: +1

Game: vs. No. 3 Georgia

• Result: L 8-27

• Pts/drive, UF (opp): 0.67 (2.00)

• Rush PPA: -0.50

• Success rate, UF (opp): 33.9% (50.0%)

• 3rd down: 4-12

• TO margin: -3

Game: Oklahoma

• Result: L 7-23

• Pts/drive, UF (opp): 0.70 (2.30)

• Rush PPA: -0.01

• Success rate, UF (opp): 37.0% (46.7%)

• 3rd down: 2-9

• TO margin: -2

4.44 to 1.34. Florida's offensive points per drive, first four games to last five.

82%. Share of the 3.80-point per-drive margin swing that came from the offense.

+0.35 to -0.23. Rushing PPA a play, averaged by game, first four games to last five (about +0.34 to -0.19 weighted by carry).

23 of 24, then 10 of 13. Red-zone scoring, first four games to last five.

21 of 43, then 26 of 69. Third-down conversions, first four to last five.

12.4% to 29.8%. Stuff rate, averaged by game, first four to last five.

The line. Florida is 5-0 when it scores 2.25 points a drive or more and 0-4 when it scores below 1.6.

Definitions. Success rate: a play gains 50% of yards to go on first down, 70% on second, 100% on third or fourth. PPA: expected points added per play, College Football Data. Points per drive: offensive points only, so return touchdowns are excluded. The 82% figure is a GatorBait calculation. Box-score totals govern; play-derived figures are used for rates only.

The historian's note

Florida has been 4-0 and lost four of the next five before. The near-twin is 2011.

That team opened 4-0, with wins over Florida Atlantic (41-3), UAB (39-0), Tennessee (33-23) and at Kentucky (48-10). It then lost four straight: Alabama 10-38, at LSU 11-41, at Auburn 6-17 and Georgia 20-24. A 26-21 win over Vanderbilt put it at 5-4 after nine games, the same record as this team. The 2011 Gators finished 6-6 in the regular season and won the Jan. 2, 2012 Gator Bowl over Ohio State, 24-17.

2010 is a partial match. That team started 4-0, including a 48-14 win over Kentucky, lost three in a row to Alabama, LSU and Mississippi State, beat Georgia in overtime, 34-31, and Vanderbilt, 55-14, and finished 7-5. It then beat Penn State, 37-24, in the Jan. 1, 2011 Outback Bowl.

The contrast is 2024, which was not a 4-0 team at all. That team was 4-5 after nine games, then beat LSU, Ole Miss and Florida State, and beat Tulane 33-8 in the Gasparilla Bowl to finish 8-5.

The honest limit: three seasons is a small sample, and a parallel is a pattern to note, not a forecast. We have not sourced earlier eras and leave them out.

The look ahead: Kentucky, Vanderbilt, Florida State

The arithmetic is short. Florida is 5-4, and one more win would make it bowl eligible if the Campbell win counts. The possible finishes: 0 wins, 5-7, not bowl eligible on record; 1 win, 6-6; 2 wins, 7-5; 3 wins, 8-4. If Campbell does not count, Florida would have four countable wins and need two more. In the SEC, Florida is 3-4 with Kentucky and Vanderbilt left, so 5-4 is the ceiling.

At Kentucky, Saturday, Nov. 14, 6:15 p.m. ET on ESPN. Florida leads the series 54-22 but is 1-4 in the last five meetings, with its lone win a 48-20 victory in 2024. Kentucky won last season's game 38-7.

Vanderbilt, Saturday, Nov. 21, Gainesville. Florida leads 44-11-2, including 24-2-1 at home. The last meeting was Oct. 7, 2023, a 38-14 Florida win. Florida's last loss to Vanderbilt was Nov. 19, 2022, 24-31 in Nashville.

At Florida State, Friday, Nov. 27, 3:30 p.m. on ABC. Florida won in 2024, 31-11, and in 2025, 40-21. Florida leads the all-time series 39-28-2 through 2025, and the last 10 meetings are split 5-5.

None of this is a prediction. It is a measuring stick. The question for the three games is whether the run game and early-down efficiency return to the range Florida had in its first four. That range is specific. Florida's lowest rushing PPA in the 4-0 start was +0.21. Its lowest scoring rate in a win this year is 2.25 points a drive. In 2026, Florida has won every time it cleared that mark and lost every time it fell short.

Nine games gave Florida a record, 5-4, and one measurable problem. After Oklahoma, Sumrall said his team feels a little bit lost. The next three games will show whether the early downs have a way back, and the first test is Saturday, Nov. 14, at Kentucky.

Sources and method

Box-score, play-by-play and win-probability figures come from ESPN's game summary and play data (site.api.espn.com and sports.core.api.espn.com; game ids 401856637, 401856672, 401856687, 401856699, 401856708, 401856714, 401856724, 401856734 and 401856739) and UF box scores on FloridaGators.com. PPA, success rate, explosiveness, stuff rate and line yards are College Football Data advanced stats. Series records come from FloridaGators.com opponent-history pages; season parallels from ESPN schedule data. Postgame quotes are from UF media relations transcripts. Per-drive, share-of-swing, red-zone, third-down-split and turnover-margin figures are GatorBait calculations from those sources. Play counts are box-score totals (rushing attempts plus pass attempts, including sacks), not College Football Data counts. Unrounded components: Florida's offensive points per drive were 200 on 45 drives (4.4444) and 78 on 58 (1.3448); opponents' were 91 on 49 (1.8571) and 143 on 56 (2.5536); the offense's share of the 3.80-point swing is 81.65%, shown as 82%.