WEEK 10 · Nov 5–7, 2026

College Touchdown Cheat Sheet

Free preview

TD chance estimates scoring at least once in the next game. Usage Score ranks scoring-area involvement within this shortlist, from 0 to 100.

Model estimate · backtested on 5,738 player-games ·
Updated Nov 1, 9:07 AM EST

Search player, team or opponent

Team Conference Game

Top 5 of 75Season usage · tap a playerMore metrics →

	Scoring outlook		Red-zone usage					Team inside the 20			Opponent defense		Touchdowns

College Touchdown Cheat Sheet, Week 10, sortable player research
TD chance	TD Usage Score	RZ rush att	GL rush att	RZ targets	EZ targets	RZ opp share	PROE	Pass %	Run %	RZ PROE	RZ TD rate	Expected TDs	Actual TDs	Next game
52%	95	34	12	0	0	40%	-11.2%	25%	75%	+4.1%	63%17 of 27 trips	10.3	11	1.78
47%	100	35	14	5	0	41%	+6.1%	45%	55%	+5.4%	83%19 of 23 trips	13.0	10	1.54
47%	54	23	13	2	0	23%	+4.5%	42%	58%	-2.8%	58%11 of 19 trips	9.1	5	1.53
46%	58	32	15	0	0	27%	+2.6%	34%	66%	+5.6%	56%18 of 32 trips	11.1	4	1.40
46%	84	43	11	1	1	46%	-27.5%	9%	91%	+11.5%	59%13 of 22 trips	11.7	9	1.38

Explore all 75 players.

Every metric, filter and recent-game detail. Free with an account.

Already have an account? Sign in

What do the columns mean?

RZ rush att: Carries starting at or inside the 20; sacks, kneels and team rushes excluded.
GL rush att: Carries starting at or inside the 5.
RZ targets: Attributed targets from at or inside the 20.
EZ targets: Estimated targets with targetYardsToGoal ≤ 0; missing locations remain unknown.
RZ opp share: Player RZ carries plus targets / team qualifying RZ opportunities. A proxy, not snap share.
PROE: Team RZ pass rate over expected, in percentage points versus FBS down/distance.
Pass %: RZ pass choices, including sacks; kneels and spikes excluded.
Run %: RZ run choices among classified plays.
RZ PROE: Opponent RZ pass rate over expected allowed, in percentage points.
RZ TD rate: Offensive touchdowns allowed per observed RZ trip; supporting trips shown below.
Expected TDs: Season xTD: empirical 2026 FBS TD rates by starting-field-position bucket × player volume.
Actual TDs: Season rushing + receiving TDs from box scores. Passing TDs do not count.
Next game: Projected rushing + receiving TDs in the next game, using recent usage and the opponent’s red-zone results.
TD chance: Model estimate of at least one rush or receiving TD. The raw Poisson estimate is calibrated using strictly earlier-week predictions and results. Colors use displayed whole percentages: dark green starts at 45%; light green and amber use fixed shortlist thresholds. Colors are relative, not certainty. See About this data.
TD Usage Score: Ranks scoring-area involvement within this shortlist, from 0 to 100. It is a fixed percentile, not a TD probability; filters do not change scores.

About this data Methodology, coverage & backtest

Coverage through Week 9

56/56 Week 9 FBS games were final at capture. Latest AP poll: Week 9. The shortlist covers Power 4 or AP-ranked matchups in the Nov 5–7 window. Oklahoma State and Sacramento State games are excluded from this displayed slate. Data captured 2026-11-01T14:07:11.829380+00:00.

Usage Score ranks scoring-area involvement within this shortlist, from 0 to 100. It is a percentile within the fixed 75-player shortlist, not the entire player universe and not a scoring probability. The same weighted formula is retained; filters never rescale it. GP counts games with box-score stat entries. RZ opportunities are a proxy; CFBD has no snap counts. EZ targets are estimated.

TD chance is a calibrated model estimate, not a guarantee. The current method is isotonic midpoints, fixed after retrospective comparison. Each backtest fold fits calibration using only earlier-week predictions and outcomes. Equal chance estimates are ordered by the underlying model estimate within a tie, then stable player and game IDs. Green marks the weekly top seven rows; amber marks above-median chance. Colors are relative ranks, not extra certainty. Projected TD counts retain the original Poisson-input model. Availability is not modeled; a player who does not participate cannot score.

Walk-forward results · Weeks 5–9

Brier 0.0981 vs 0.1072 for the RZ-volume baseline; 5,738 player-games. Mean predicted scoring 13.4%; observed 12.8%. The fixed-method backtest has 10-band calibration error of 2.20%, versus 2.87% before calibration. High bands remain sparse: the isotonic midpoints 50%+ backtest band has 10 player-games. Method development followed review of earlier results; this is not untouched external validation. Error remains uneven across bands. Current roster mapping and provider revisions limit point-in-time fidelity.

RZ≤20 yards; GL≤5. Sacks, kneels, team rushes, invalid/no-play records excluded from player rush volume.
Player targets require targetId; partial parses retained only when required fields exist. End-zone targets use targetYardsToGoal≤0 and are estimates, not tracking data. Missing locations are unknown.
RZ opportunity share is a proxy: player RZ carries+targets / team qualifying RZ carries+pass attempts, including unattributed opportunities. CFBD does not provide snap counts.
FBS 2026 empirical TD/volume rates by rush/target and starting-yard buckets1–5,6–10,11–20,21–40,41–60,61–100. Buckets under30 opportunities back off to empirical same-kind rate.
PROE is actual pass choice minus FBS RZ down/distance baseline, with empirical down/global fallback. Sacks count as pass choices; kneels/spikes/non-scrimmage excluded.
RZ trip is a unique game/drive/offense with a counted RZ snap; an eventual offensive rush/pass TD on that drive is a converted trip.
Projected TDs = recent per-game xTD (3-game half-life) × [1+0.5×(opponent RZ TD/trip ÷ FBS rate−1)]. The opponent effect is shrunk halfway toward neutral.
Raw TD chance =1−exp(−projected TDs); displayed chance uses isotonic midpoint calibration, fitted only on earlier-week out-of-fold predictions/outcomes. Values are linearly interpolated between weighted raw-score centroids of pooled monotone blocks, with clipped endpoints. At least1000 earlier rows are required; otherwise the raw chance is retained. The fixed method was chosen after retrospective comparison, not untouched external validation. Expected TD counts are unchanged. Equal chance estimates reflect pooled calibration blocks and are ordered by raw forecast, then stable player and game IDs, without inventing probability differences. Chance color is a fixed weekly-shortlist tier (dark green top seven rows, amber above median), not another probability or guarantee. TD Usage Score is percentile within the fixed displayed 75-player shortlist of0.40 GL carry share+0.25 RZ proxy share+0.20 estimated EZ target share+0.15 opponent factor. Zero EZ denominator gives zero component, not proof of no missing EZ targets.
Actual TDs use CFBD rushing+receiving box scores; unavailable box coverage is null. Passing TDs are excluded from a QB’s scoring outcome.
All historical folds refit rates, baselines and opponents using prior weeks only. Unplayed outcomes are excluded. No claim of validated advantage.

Source: CollegeFootballData v2, cached capture. No new data request is made by this dashboard. The private receipt includes full calibration and dropped-play accounting.

Saturday Index

Getting this view ready

Bringing the latest verified football information into focus.

Scoring outlook

Red-zone usage

Team inside the 20

Opponent defense

Touchdowns

College Touchdown Cheat Sheet, Week 10, sortable player research

Player

TD chance

TD Usage Score

RZ rush att

GL rush att

RZ targets

EZ targets

RZ opp share

PROE

Pass %

Run %

RZ PROE

RZ TD rate

Expected TDs

Actual TDs

Next game

52%

40%

-11.2%

25%

75%

+4.1%

63%17 of 27 trips

10.3

1.78

47%

100

41%

+6.1%

45%

55%

+5.4%

83%19 of 23 trips

13.0

1.54

47%

23%

+4.5%

42%

58%

-2.8%

58%11 of 19 trips

9.1

1.53

46%

27%

+2.6%

34%

66%

+5.6%

56%18 of 32 trips

11.1

1.40

46%

46%

-27.5%

91%

+11.5%

59%13 of 22 trips

11.7

1.38

Walk-forward results · Weeks 5–9

RZ≤20 yards; GL≤5. Sacks, kneels, team rushes, invalid/no-play records excluded from player rush volume.

Player targets require targetId; partial parses retained only when required fields exist. End-zone targets use targetYardsToGoal≤0 and are estimates, not tracking data. Missing locations are unknown.

RZ opportunity share is a proxy: player RZ carries+targets / team qualifying RZ carries+pass attempts, including unattributed opportunities. CFBD does not provide snap counts.

FBS 2026 empirical TD/volume rates by rush/target and starting-yard buckets1–5,6–10,11–20,21–40,41–60,61–100. Buckets under30 opportunities back off to empirical same-kind rate.

PROE is actual pass choice minus FBS RZ down/distance baseline, with empirical down/global fallback. Sacks count as pass choices; kneels/spikes/non-scrimmage excluded.

RZ trip is a unique game/drive/offense with a counted RZ snap; an eventual offensive rush/pass TD on that drive is a converted trip.

Projected TDs = recent per-game xTD (3-game half-life) × [1+0.5×(opponent RZ TD/trip ÷ FBS rate−1)]. The opponent effect is shrunk halfway toward neutral.

Raw TD chance =1−exp(−projected TDs); displayed chance uses isotonic midpoint calibration, fitted only on earlier-week out-of-fold predictions/outcomes. Values are linearly interpolated between weighted raw-score centroids of pooled monotone blocks, with clipped endpoints. At least1000 earlier rows are required; otherwise the raw chance is retained. The fixed method was chosen after retrospective comparison, not untouched external validation. Expected TD counts are unchanged. Equal chance estimates reflect pooled calibration blocks and are ordered by raw forecast, then stable player and game IDs, without inventing probability differences. Chance color is a fixed weekly-shortlist tier (dark green top seven rows, amber above median), not another probability or guarantee. TD Usage Score is percentile within the fixed displayed 75-player shortlist of0.40 GL carry share+0.25 RZ proxy share+0.20 estimated EZ target share+0.15 opponent factor. Zero EZ denominator gives zero component, not proof of no missing EZ targets.

Actual TDs use CFBD rushing+receiving box scores; unavailable box coverage is null. Passing TDs are excluded from a QB’s scoring outcome.

All historical folds refit rates, baselines and opponents using prior weeks only. Unplayed outcomes are excluded. No claim of validated advantage.

Source: CollegeFootballData v2, cached capture. No new data request is made by this dashboard. The private receipt includes full calibration and dropped-play accounting.