Depth charts and who's out
Who's out
Ravens
Ja'Kobi Lane WRWristIR
T.J. Tampa CBKneeIR
Chris Moore WRAnkleQ
Zay Flowers WRHamstringQ
Malaki Starks SHamstringQ
2 more
Ronnie Stanley OTToeQ
Trey Hendrickson LBFingerQ
Cowboys
Cobie Durant CBHamstringOUT
DeMarvion Overshown LBHamstringOUT
Malik Hooker SForearmOUT
P.J. Locke SFootOUT
Malik Davis RBHipIR
Every read on this page assumes the man is on the field. Designations as of 2026-09-20.
The game
Ravens ahead means Henry carries it late; Cowboys chasing means Prescott throws, and Lamb, Pickens and Ferguson get the volume.
BAL has to
Run at the Cowboys zone eyes
Explosive run rate 14.5%, 1st. Cowboys play man 25.0%, 24th: lighter boxes, backs turned.
DAL has to
Find Pickens against man
1.00 EPA a target vs man, 2nd of 106. Ravens play man 32.1%, 13th.
The model's play
model fades them by 0.5+, favorite, worse QB, elite QB
Won 62.5% of 40 past bets like it. The worst case we can be 80% sure of is 52.4%, which is still short of the 52.4% break-even.
Best price
The matchup
Cowboys: send it at Jackson.
Cowboys: find Pickens against man.
Ranks are 2025 regular season, 1st to 32nd, 1st best at the job either side. top 8 bottom 8
The game plan in words, and every rank (2025 regular season)
What they like to do Minter's Ravens run it more than almost anyone (lowest) and play slow (4th-slowest).
Where they win and lose They break explosive runs (1st) and run it well (2nd) and they get sacked (4th-worst).
What the Cowboys defense brings They load the box (5th-lowest) and they give it up (2nd-worst), never take it away (3rd-worst) and let everything get completed (5th-worst).
What works against them Bring pressure: they take sacks (4th-worst).
Where the two meet Cowboys: send it at Jackson. Blitz rate 28.5%, 9th. Jackson falls to 0.11 EPA when blitzed, 13th of 36. Ravens: run at the Cowboys zone eyes. Explosive run rate 14.5%, 1st. Cowboys play man 25.0%, 24th: lighter boxes, backs turned.
What they like to do Schottenheimer's Cowboys play fast (1st).
Where they win and lose They complete everything (4th), move the ball (5th) and throw it well (5th).
What the Ravens defense brings They rarely play two-high (3rd-lowest) and they cannot get a sack (2nd-worst).
What works against them Nothing on the ledger stands out; it comes down to the matchup below.
Where the two meet Cowboys: find Pickens against man. 1.00 EPA a target vs man, 2nd of 106. Ravens play man 32.1%, 13th. Cowboys: give Prescott a clean pocket. 0.36 EPA a dropback clean, 6th of 36. Ravens get pressure 27.5%, 8th-lowest.
Baltimore
Dallas
The bets
What the model likes
4 more
The information
Player angles
Also posted, with no read either way: Lamar Jackson 240.5 passing yards (over in 12 of his 32, 2024-2026), Justice Hill 13.5 rushing yards (over in 10 of his 26, 2024-2026), Emari Demercado 8.5 rushing yards (over in 14 of his 26, 2024-2026), Rashod Bateman 46.5 receiving yards (over in 10 of his 31, 2024-2026), Ryan Flournoy 30.5 receiving yards (over in 6 of his 22, 2024-2026), Jake Ferguson 29.5 receiving yards (over in 16 of his 33, 2024-2026), LaJohntay Wester 19.5 receiving yards, Javonte Williams 16.5 receiving yards (over in 12 of his 35, 2024-2026), Derrick Henry 9.5 receiving yards (over in 16 of his 36, 2024-2026), CeeDee Lamb 6.5 receptions (over in 14 of his 30, 2024-2026), George Pickens 5.5 receptions (over in 12 of his 33, 2024-2026), Mark Andrews 4.5 receptions (over in 7 of his 36, 2024-2026), Jake Ferguson 3.5 receptions (over in 19 of his 33, 2024-2026), Rashod Bateman 3.5 receptions (over in 7 of his 31, 2024-2026), Ryan Flournoy 3.5 receptions (over in 5 of his 22, 2024-2026), Justice Hill 2.5 receptions (over in 11 of his 26, 2024-2026), Chris Moore 1.5 receptions (over in 4 of his 10, 2025-2026), Derrick Henry 1.5 receptions (over in 10 of his 36, 2024-2026) and LaJohntay Wester 1.5 receptions.
Player trends
5 more player trends
Game angles
First quarter: Ravens -0.5 and a total of 10.
Game trends
A lean is where a record points. Only a record past the luck line is a bet; the rest is reading. The bar's tick is the line to beat. Confidence runs from 50, a coin flip, to 100.
Every read in full: the scores, the story and the receipts
The player lines
Over 82.5 receiving yards for CeeDee Lamb leads 31 lines. On 7, the player's history and the matchup agree.
Why 80
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 11 of 16 (69%) against a 36% baseline, 53% after shrinking for sample size.
- Agreement 2 other signals on the same side.
- The grid No held angle applies to this line.
- The matchup BAL allow 18% more than league to WRs, 4th most.
- The price -115, which needs 53% to break even.
- Where it went Unproven and priced at minus money, with no 1-unit bet on the game to ride alongside.
Why 76
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 4 of 5 (80%) against a 36% baseline, 51% after shrinking for sample size.
- Agreement 2 other signals on the same side.
- The grid No held angle applies to this line.
- The matchup BAL allow 13% more than league to QBs, 5th most.
- The price -115, which needs 53% to break even.
- Where it went Unproven and priced at minus money, with no 1-unit bet on the game to ride alongside.
Why 73
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 4 of 5 (80%) against a 40% baseline, 51% after shrinking for sample size.
- Agreement 1 other signal on the same side.
- The grid No held angle applies to this line.
- The matchup BAL allow 29% more than league to RBs, 2nd most.
- The price -160, which needs 62% to break even.
See all 31 player lines
Why 70
- The story Against it: the Ravens play slow.
- The evidence 11 of 17 (65%) against a 40% baseline, 49% after shrinking for sample size.
- Agreement No other signal on this side.
- The grid No held angle applies to this line.
- The matchup DAL allow 52% more than league to QBs, 1st most.
- The price -120, which needs 55% to break even.
Why 58
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 5 of 7 (71%) against a 54% baseline, 47% after shrinking for sample size.
- Agreement 1 other signal on the same side.
- The grid No held angle applies to this line.
- The matchup BAL allow 18% more than league to WRs, 4th most.
- The price -115, which needs 53% to break even.
- Where it went Unproven and priced at minus money, with no 1-unit bet on the game to ride alongside.
Why 54
- The story With it: a 53.5 total is a throwing script; Cowboys are without Hooker in the secondary.
- The evidence 6 of 10 (60%) against a 48% baseline, 40% after shrinking for sample size.
- Agreement No other signal on this side.
- The grid No held angle applies to this line.
- The matchup DAL allow 19% more than league to WRs, 3rd most.
- The price -110, which needs 52% to break even.
Why 53
- The story No view on this one.
- The evidence 11 of 16 (69%) against a 58% baseline, 53% after shrinking for sample size.
- Agreement No other signal on this side.
- The grid No held angle applies to this line.
- The matchup BAL allow 14% less than league to RBs, 26th most.
- The price -115, which needs 53% to break even.
Why 73
- The story Against it: the Cowboys play fast.
- The evidence 5 of 5 (100%) against a 54% baseline, 75% after shrinking for sample size.
- Agreement No other signal on this side.
- The grid No held angle applies to this line.
- The matchup BAL allow 3% less than league to QBs, 18th most.
- The price -120, which needs 55% to break even.
Why 66
- The story No view on this one.
- The evidence 11 of 16 (69%) against a 45% baseline, 53% after shrinking for sample size.
- Agreement 2 other signals on the same side.
- The grid No held angle applies to this line.
- The matchup DAL allow 2% more than league to RBs, 17th most.
- The price -115, which needs 53% to break even.
Why 58
- The story With it: a 53.5 total is a throwing script; Cowboys are without Hooker in the secondary.
- The evidence 18 of 35 (51%) against a 48% baseline, 41% after shrinking for sample size.
- Agreement 1 other signal on the same side.
- The grid No held angle applies to this line.
- The matchup DAL allow 19% more than league to WRs, 3rd most.
- The price -145, which needs 59% to break even.
Why 57
- The story No view on this one.
- The evidence 13 of 26 (50%) against a 48% baseline, 38% after shrinking for sample size.
- Agreement No other signal on this side.
- The grid No held angle applies to this line.
- The matchup DAL allow 34% more than league to RBs, 3rd most.
- The price -115, which needs 53% to break even.
Why 52
- The story Against it: a 53.5 total is a throwing script; Cowboys are without Hooker in the secondary.
- The evidence 13 of 18 (72%) against a 56% baseline, 57% after shrinking for sample size.
- Agreement No other signal on this side.
- The grid No held angle applies to this line.
- The matchup DAL allow 3% less than league to TEs, 16th most.
- The price -115, which needs 53% to break even.
Lines from Bovada as of 2026-09-25. A record is the player against the consensus closing line, every game since 2023, graded on the box score. Overs hit 46 to 48% league-wide; that is the bar, not 50. "Player and matchup agree" means he clears this number 60%+ in the situation this game puts him in AND the opponent allows 10%+ over league to his position. None of these has a graded record yet.
Player trends
Henry: 10 of his last 10 games as a favorite with 50+ rushing yards.
Why 80
- The story No view on this one.
- The evidence 10 of 10 (100%) against a 63% baseline, 86% after shrinking for sample size.
- Agreement 2 other signals on the same side.
Why 80
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 9 of 10 (90%) against a 45% baseline, 72% after shrinking for sample size.
- Agreement 2 other signals on the same side.
- Where it went Unproven and priced at minus money, with no 1-unit bet on the game to ride alongside.
See all 5 player trends
Why 70
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 8 of 10 (80%) against a 48% baseline, 60% after shrinking for sample size.
- Agreement 2 other signals on the same side.
- Where it went Unproven and priced at minus money, with no 1-unit bet on the game to ride alongside.
Why 66
- The story With it: a 53.5 total is a throwing script; Cowboys are without Hooker in the secondary.
- The evidence 9 of 10 (90%) against a 62% baseline, 72% after shrinking for sample size.
- Agreement 1 other signal on the same side.
Why 57
- The story With it: the Cowboys chase, so Prescott throws; a 53.5 total is a throwing script.
- The evidence 8 of 10 (80%) against a 58% baseline, 60% after shrinking for sample size.
- Agreement 1 other signal on the same side.
- Where it went Unproven and priced at minus money, with no 1-unit bet on the game to ride alongside.
The second number is the point. Everyone posts the first one.
Team trends
The last time the Ravens were here, Sep 20, 2020: beat the Texans by 17 as a 7.5-point favorite.
Inside the luck line: 198 games, and shuffled results produce records like this one. Reading, not a bet; the lean is the Ravens.
The last time the Cowboys were here, Sep 23, 2007: beat the Bears by 24 as a 3-point underdog.
Inside the luck line: 93 games, and shuffled results produce records like this one. Reading, not a bet; the lean is the Ravens.
See all 7 team trends
The last time the Cowboys were here, Sep 14, 2014: beat the Titans by 16 as a 3-point underdog.
Inside the luck line: 92 games, and shuffled results produce records like this one. Reading, not a bet; the lean is the Cowboys.
Why 76
- The evidence 7 of 8 (88%) against a 49% baseline, 66% after shrinking for sample size.
- Agreement 2 other signals on the same side.
Why 70
- The evidence 10 of 13 (77%) against a 48% baseline, 59% after shrinking for sample size.
- Agreement 2 other signals on the same side.
Why 61
- The evidence 8 of 11 (73%) against a 50% baseline, 53% after shrinking for sample size.
- Agreement 2 other signals on the same side.
Against the spread and the total, as far back as the current head coach and starting quarterback go (at least three seasons; each card says how far, and what changed). The second bar is the same team everywhere else, which is the number that tells you whether the first one means anything.
Model plays carry a live record. Prop spots and trends are situational reads, and no prop verdict has a graded record yet. When two sections disagree, both stay.
History
Last meetings
- 2024 wk 3
Ravens at
CowboysRavens won, Ravens covered28-25 - 2020 wk 13
Cowboys at
RavensRavens won, Ravens covered17-34
Against the spread in these meetings: 2-0 to Ravens, on closing numbers.
The card
Betting PassThe number. The market has the Ravens by 3.5 at a total of 53.5: Ravens 28.5, Cowboys 25 on paper.
How it plays. The total of 53.5 is high. Ravens play slow (29th in plays a game). Cowboys play fast (1st in plays a game).
What changed. Ravens: WR Zay Flowers questionable. Cowboys: FS Malik Hooker out; LCB Cobie Durant out; LILB DeMarvion Overshown out.
Model favorites
Three favorites from 39 signals: three player lines. None sized at this price. Data as of: forecast Fri 12:00 PM ET; injuries Fri 11:49 AM ET; model Fri 12:04 PM ET.
A player line
4 reasons forA player line
4 reasons forA player line
4 reasons forHow the score works
Score is how much we trust a bet. 50 is a coin flip; the number is how far past it the evidence pushes: reasons for minus reasons against, and a reason from something with a years-long record counts more than a trend. Units follow it: a bet backed by a tracked record is 1 unit, 1.5 at 80 or more; a bet whose logic is sound but unproven is half a unit, and only at plus money or alongside a 1-unit bet on the same game. Click a bet for its reasons.