"Ban the tush push." Five seasons, 2021 to 2025.

Nine situations, each answered by the one tool from every play of 2021 through 2025, measuring with the definition below. Yellow is the defined play, blue is every other rush with the same yards to go, and the hatched stretch between them is the gap in how often each converted.

Holds, as a steady edge

Over 2021 to 2025, the defined play converted 81.6% of 1,412 plays, every other rush with the same yards to go 71.5% of 6,147, a gap of 10.1 points. Take away the Eagles and the gap is 9.7 points, so it is not one team's play. The play converts about ten points more often than everything else at that distance, season after season. Whether that is grounds for a ban is a judgment about safety and fairness, and no number on this sheet settles it.

0of 247,284 play descriptions name the play

10.1points: the gap, all downs

9.7points: the gap without the Eagles

The verdict follows a rule written in season-sheet.mjs, so you can disagree with it. It reads the all-downs gap: under 100 plays matching the definition is too few to rule; a gap of 8 points or more that stays at 5 or more without the Eagles holds, as a steady edge; a gap under 4 does not hold as stated; between the two is mixed. Whether the play should be legal is not on this sheet. This is the page's rule, not the agent's. Every number is the tool's.

Measured with

"The quarterback takes the snap under center with one yard or less to go and runs straight ahead, with teammates pushing him from behind. In these plays it is any run by the team's primary passer, not from shotgun, with one yard or less to go, on any down."

dictionary/tush-push.json, written 2026-10-26 by the analyst, episode 2. Rules: the runner is the team's primary passer that game (the player with its most pass attempts); not from shotgun; any down; 1 or fewer yards to go. Compared with every other rush with the same yards to go. Cannot see: who pushed the runner or whether anyone pushed at all, so it counts every under-center quarterback sneak, pushed or not, and it cannot tell a sneak from a designed quarterback run off tackle.

The name search first: node ./query-plays.mjs --search "tush push" --seasons 2021-2025 found the words in 0 of 247,284 descriptions. The play has no name in the data; the definition above is how the tool finds it.

All downs

Every rush with one yard or less to go, the defined play against the rest.

node ./query-plays.mjs --play tush-push --seasons 2021-2025

71.5%

81.6%

81.6% the defined play, 1,412 plays; 71.5% every other rush at that distance, 6,147 plays; a gap of 10.1 points.

Third and fourth down

When the drive is on the line.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --downs 3,4

70.1%

82.9%

82.9% the defined play, 1,149 plays; 70.1% every other rush at that distance, 3,515 plays; a gap of 12.8 points.

Fourth down only

The decision the room argues about most.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --downs 4

72.0%

79.9%

79.9% the defined play, 194 plays; 72.0% every other rush at that distance, 1,046 plays; a gap of 7.9 points.

One yard to go

The shortest distance, where the play is most used.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --yards-to-go 1

71.5%

81.6%

81.6% the defined play, 1,412 plays; 71.5% every other rush at that distance, 6,147 plays; a gap of 10.1 points.

Goal to go

At the goal line, where a conversion is a touchdown.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --goal-to-go

56.6%

66.2%

66.2% the defined play, 260 plays; 56.6% every other rush at that distance, 1,104 plays; a gap of 9.5 points.

Every team but the Eagles

Take away the team the play is named for.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --not-team PHI

71.6%

81.2%

81.2% the defined play, 1,246 plays; 71.6% every other rush at that distance, 6,007 plays; a gap of 9.7 points.

The Eagles only

The team that made it an argument.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --team PHI

69.3%

84.3%

84.3% the defined play, 166 plays; 69.3% every other rush at that distance, 140 plays; a gap of 15.1 points.

Regular season only

Leave the playoffs out.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --type REG

71.3%

81.3%

81.3% the defined play, 1,323 plays; 71.3% every other rush at that distance, 5,858 plays; a gap of 9.9 points.

Playoffs only

Thirteen games a season.

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --type POST

75.1%

86.5%

86.5% the defined play, 89 plays; 75.1% every other rush at that distance, 289 plays; a gap of 11.4 points.

By season

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --by season

Season	Defined plays	Rate	Other rushes	Rate	Gap
2021	243	78.6%	1,296	73.0%	5.6
2022	304	85.2%	1,223	71.1%	14.1
2023	304	78.9%	1,126	72.0%	6.9
2024	281	83.3%	1,211	70.4%	12.8
2025	280	81.4%	1,291	71.0%	10.4

By team

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --by team

Team	Defined plays	Rate	Other rushes	Rate	Gap
PHI	166	84.3%	140	69.3%	15.1
BUF	125	84.8%	172	73.3%	11.5
SF	74	90.5%	207	70.5%	20.0
DEN	72	86.1%	162	67.3%	18.8
LAC	57	82.5%	187	63.1%	19.4
NE	56	78.6%	177	70.6%	7.9
NYG	56	85.7%	151	73.5%	12.2
JAX	55	80.0%	192	67.2%	12.8
LV	52	76.9%	163	65.0%	11.9
CIN	52	78.8%	184	70.7%	8.2
PIT	51	82.4%	189	73.5%	8.8
CHI	45	73.3%	162	74.1%	-0.7
MIN	42	83.3%	160	65.6%	17.7
CLE	41	85.4%	185	71.9%	13.5
TEN	40	92.5%	167	75.4%	17.1
NYJ	38	73.7%	138	73.2%	0.5
DAL	36	80.6%	220	76.8%	3.7
IND	35	80.0%	213	72.8%	7.2
NO	34	73.5%	184	72.3%	1.2
ARI	32	71.9%	217	73.7%	-1.9
LA	29	72.4%	234	72.6%	-0.2
HOU	29	93.1%	179	67.0%	26.1
SEA	29	72.4%	173	74.6%	-2.2
CAR	25	80.0%	179	70.9%	9.1
ATL	23	65.2%	211	66.8%	-1.6
BAL	22	86.4%	241	76.8%	9.6
GB	21	85.7%	208	73.6%	12.2
TB	20	85.0%	224	72.8%	12.2
WAS	20	80.0%	258	70.9%	9.1
DET	16	75.0%	223	73.5%	1.5
MIA	13	46.2%	189	66.7%	-20.5

Teams with 10 or more plays matching the definition, most first.

Who ran it, season by season

node ./query-plays.mjs --play tush-push --seasons 2021 --by team (and each season to 2025)

One row per team, one column per season: the plays that matched the definition. Each season's most is in yellow; the first row is every team together. The tool lists a team once it has 10 matching plays in a season, so a smaller count reads "under 10". Teams in order of the five-season column.

Team	2021	2022	2023	2024	2025	2021 to 2025
Every team	243	304	304	281	280	1,412
PHI	15	40	36	48	27	166
BUF	13	15	29	37	31	125
SF	11	18	13	13	19	74
DEN	11	under 10	25	14	13	72
LAC	under 10	13	10	10	15	57
NE	17	under 10	12	under 10	16	56
NYG	under 10	16	10	13	10	56
JAX	under 10	14	under 10	15	16	55
LV	under 10	under 10	15	10	12	52
CIN	under 10	18	10	under 10	under 10	52
PIT	under 10	16	17	13	under 10	51
CHI	under 10	12	16	under 10	10	45
MIN	under 10	under 10	15	13	under 10	42
CLE	under 10	18	under 10	under 10	under 10	41
TEN	10	under 10	under 10	16	under 10	40
NYJ	under 10	under 10	under 10	under 10	15	38
DAL	under 10	13	under 10	under 10	under 10	36
IND	10	under 10	under 10	under 10	12	35
NO	12	under 10	under 10	under 10	under 10	34
ARI	under 10	under 10	under 10	under 10	under 10	32
LA	11	under 10	under 10	under 10	under 10	29
HOU	under 10	under 10	10	under 10	under 10	29
SEA	under 10	under 10	under 10	under 10	under 10	29
CAR	15	under 10	under 10	under 10	under 10	25
ATL	under 10	under 10	under 10	under 10	under 10	23
BAL	under 10	11	under 10	under 10	under 10	22
GB	under 10	under 10	under 10	under 10	under 10	21
TB	under 10	under 10	under 10	under 10	under 10	20
WAS	under 10	under 10	under 10	under 10	under 10	20
DET	under 10	under 10	under 10	under 10	under 10	16
MIA	under 10	under 10	under 10	under 10	under 10	13

Not on this sheet: why, and whether. Every gap here is how often each kind of play converted in the plays kept. It does not say why the play converts, who pushed, or whether anyone was hurt, and it does not say what the rule should be.

What the agent ruled on air

The claim: Whether the tush push should be legal is a question about safety, fairness, and the game people want to watch, and the plays cannot settle it. I tested whether the play converts so much more often than other short-yardage rushes that it is a different kind of play. For that to hold, the defined play needs a large conversion gap over other rushes at the same distance, and the gap must survive without the team that runs it most.

Definition used: I wrote tush-push in this run, at dictionary/tush-push.json, "written 2026-10-26 by the analyst, episode 2", and no entry was on file before it. A search for "tush push" across 2021-2026 (265,845 descriptions, regular season and playoffs) found 0 matches, so the play has no name in the data and the entry defines it by situation.

Five seasons: For 2021-2025, regular season and playoffs, all teams and downs, 1 or fewer yards to go: the defined play converted 78.3% of 1,472 plays vs other short-yardage rushes 70.0% of 6,282 plays, gap 8.3 points.

The read that could prove the claim wrong drops PHI, which ran the defined play most (173 plays): 77.9% of 1,299 plays vs 70.1% of 6,134 plays, gap 7.8 points, almost unchanged.

This season: For 2026, weeks 1-7, with the same filters: the defined play converted 82.0% of 128 plays vs other short-yardage rushes 67.7% of 477 plays, gap 14.3 points. The tool printed no SMALL SAMPLE warning, but 128 plays is about a tenth of the five-season group, so a few conversions either way can move that rate by several points.

The gap between the two answers: The 2026 gap (14.3 points on 128 plays) is larger than the five-season gap (8.3 points on 1,472 plays), but seven weeks of data cannot show that the difference is more than chance.

Ruling: The claim does not hold in its strong form: the defined play converted 8.3 points more often than other short-yardage rushes over five seasons and 7.8 points more without PHI, which is a real edge but not a gap that makes it a different kind of play.

What these plays cannot tell you: They cannot show who pushed the runner or whether anyone pushed at all, so this entry counts every under-center sneak by the primary passer at 1 yard or less, and it cannot explain why the play converts.

Commands run:

node ./query-plays.mjs --search "tush push" --seasons 2021-2026

node ./query-plays.mjs --play tush-push --seasons 2021-2025

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --by team

node ./query-plays.mjs --play tush-push --seasons 2026

node ./query-plays.mjs --play tush-push --seasons 2021-2025 --not-team PHI

From rulings/ruling-2026-10-26-082753.md. The agent ruled with the tool as it was on October 26, 2026. That tool counted 195 two-point tries as failed conversions, so the ruling's numbers differ from the corrected ones above (fixed November 3, 2026).