NFL Week 3 slate →NFL dashboard →

Baltimore Ravens at Dallas Cowboys

Sunday Late · Sun, Sep 27, 4:25 PM ETBAL @ DAL
Baltimore Ravens

Away

Baltimore Ravens

1-1

Dallas Cowboys

Home

Dallas Cowboys

1-1

Model Win Probability

Baltimore Ravens · Away

57%

Model
separation14%

Dallas Cowboys · Home

43%

Signal: Unrated· components disagree · 100% feature coverage
Elo component
65% BAL
Efficiency component
44% BAL
Feature coverage
100%
Data cutoff
2026-09-21T03:30:00Z
Model
pt-nfl-ensemble-v0.3

Signal strength describes model agreement and data quality, not certainty or wagering value.

Team Efficiency Comparison

69
97
Offense
59
91
Passing
97
62
Rushing
53
3
Defense
6
9
Pass Rush
36
80
Situational

0–100 league percentile across all 32 teams. Defensive ratings are inverted so a higher number always means a better unit.

Blended Team Metrics

BALMetricDAL
0.064EPA / Play0.139
46.3%Success Rate48.2%
0.097Dropback EPA0.242
0.045Rush EPA-0.053
14.5%Explosive Play Rate11.2%
40.9%3rd Down Conversion48.1%
16.3%Red Zone TD Rate19.6%
0.007Def. EPA / Play Allowed0.170
4.6%Pressure Rate4.6%

Prior-season form is blended with current-season play and its weight decays as the current sample grows. Metrics without enough data are shown as “—”, never as zero.

Game Context

Kickoff
Sun, Sep 27, 4:25 PM ET
Surface
matrixturf
Away rest
7 days
Home rest
7 days
Site
Neutral site

Player Impact

Lamar Jackson

QB

2026 Weeks 1–2

CMP/ATT
38/56
Pass yds
559
Pass TD
2
INT
1
Y/A
10
Sacks
5

Derrick Henry

RB

2026 Weeks 1–2

Carries
40
Rush yds
212
YPC
5.3
Rush TD
4
Targets
4
Rec
4
Rush share
67.8%

Justice Hill

RB

2026 Weeks 1–2

Carries
8
Rush yds
24
YPC
3
Rush TD
0
Targets
3
Rec
3
Rush share
13.6%

Mark Andrews

TE

2026 Weeks 1–2

Targets
13
Rec
10
Rec yds
98
YPR
9.8
Rec TD
0
Target share
24.5%

Rashod Bateman

WR

2026 Weeks 1–2

Targets
10
Rec
7
Rec yds
88
YPR
12.6
Rec TD
1
Target share
18.9%

Injuries & Availability

Away · Baltimore Ravens

Nnamdi MadubuikeDL

OUT

Neck · Limited Participation in Practice

T.J. TampaCB

OUT

Knee · Did Not Participate In Practice

Teddye BuchananLB

OUT

Knee · Limited Participation in Practice

Zay FlowersWR

DOUBTFUL

Hamstring · Limited Participation in Practice

Home · Dallas Cowboys

Malik HookerS

OUT

Forearm · Did Not Participate In Practice

DeMarvion OvershownLB

OUT

Hamstring · Did Not Participate In Practice

Scoring by Quarter

Sample · 2 games

Away · Baltimore Ravens

Q1
21
Q2
24
Q3
3
Q4
10

Home · Dallas Cowboys

Q1
10
Q2
14
Q3
17
Q4
16

Small sample — treat quarter splits as descriptive context, not a trend.

Latest News

Analytical Angles

Passing profile · BAL

passing

Baltimore Ravens is at the 59th league percentile in blended passing efficiency entering Week 3.

Defense profile · BAL

efficiency

Baltimore Ravens is at the 53th league percentile in blended defensive efficiency; higher values reflect fewer EPA allowed.

Passing profile · DAL

passing

Dallas Cowboys is at the 91th league percentile in blended passing efficiency entering Week 3.

Defense profile · DAL

efficiency

Dallas Cowboys is at the 3th league percentile in blended defensive efficiency; higher values reflect fewer EPA allowed.

Statistical context only. Picks Terminal publishes matchup analysis, not wagering advice.

Keys to the Game

  • Blended model gives Baltimore Ravens 57% and Dallas Cowboys 43%.
  • Elo puts Baltimore Ravens at 65.0% and the efficiency model at 44.2%.
  • The two components disagree by 20.8 points, so the blend is less confident.

What Could Flip It

  • Rating and efficiency views point in opposite directions on this game.
  • Ratings carry into Week 3 from a two-game sample, so single-game variance still dominates.

Model Analysis

pt-nfl-ensemble-v0.3 rates Baltimore Ravens (1-1) at 57% to win at Dallas Cowboys (1-1). Elo puts Baltimore Ravens at 65.0% and the efficiency model at 44.2%; the blend lands at 57%. The components disagree, which pulls the blend toward the middle.

Forecast

Baltimore Ravens projected to win, 57% of the time in the model.

Sources & Methodology

Win probability: pt-nfl-ensemble-v0.3 — a deterministic ensemble of a margin-aware Elo component and a regularized efficiency component built from pregame team differentials (EPA per play, dropback and rush EPA, success rate, explosive rate, sack performance, home field and rest). The blended probability is calibrated on training seasons only and validated by walk-forward backtest against completed results. All inputs come from public play-by-play and schedule data. No external market data of any kind is used anywhere on this page.