NHL picks
NHL Capital One Arena
Away

Dallas Stars

9-5
Final
6 – 5
Pregame proj 3.12–3.15
Home

Washington Capitals

8-6

LiveIn-game

Loading the action…

PMill verdict

🔒 PMill pick is locked

Buy credits or subscribe for $39.99/mo. Every stat, projection and price on this page stays free.

Start 7-day free trial

PMill Pro unlocks this verdict and every game on the board, $39.99/mo or $299/yr. Cancel anytime.

Odds & market Consensus

Moneyline 4.8% hold
Side Consensus No-vig Best
DAL -108 open -105 49.6% —
WSH -112 open -116 50.4% —
Puck line 4.3% hold
Side Consensus No-vig Best
DAL +1.5 -278 73.5% w/ vig 70.5% —
WSH -1.5 +225 30.8% w/ vig 29.5% —
Total 4.3% hold
Side Consensus No-vig Best
Over o5.5 -125 55.6% w/ vig 53.2% —
Under u5.5 +105 48.8% w/ vig 46.8% —

No-vig = the consensus price with the bookmaker margin removed — the market's own win probability. Best-price shopping appears when our odds feed has more than one book on this game.

Tap a price to add it to your slip

Tap a cell to add to your slip

Puck Line
ML
O/U
DAL
+1.5 -278
-108
O 5.5 -125
WSH
-1.5 +225
-112
U 5.5 +105

Model vs market

Projected score
DAL 3.12 – WSH 3.15
Projected total
6.48
Market 5.5 · model +1.0
Projected margin
WSH 0.0
Selection Market Model Blend Fair EV
Over 5.5 Total · -110 (std) 50.0% 59.5% — — 🔒
Under 5.5 Total · -110 (std) 50.0% 40.5% — — 🔒

Market = no-vig probability from the consensus price. Model = PMill's raw probability. Blend, fair odds and EV unlock with the pick. “(std)” = standard −110 assumed where the book's price isn't in our feed; “n/a” = no book has posted this line yet.

Most likely scorelines DAL–WSH

3–3
6.5%
2–2
6.0%
2–3
4.5%
3–2
4.5%
4–4
4.0%
3–4
3.7%
Over 5.5: 60%Over 6.5: 49%

Why the model leans this way

🔒 See which of the model's 11 inputs push this game and how hard, plus the category leans and the AI write-up.

Start 7-day free trial

PMill Pro unlocks this breakdown plus every game on the board, $39.99/mo or $299/yr. Cancel anytime.

Team stat comparison Brighter bar = better · #rank in league

Team profile Season team stats
3.01 #18/29
Goals / game
3.29 #11/29
2.50 #1/29
Goals against / game
3.04 #11/29
25.6 #21/29
Shots for / game
27.8 #13/29
22.3 #1/29
Shots against / game
28.1 #21/29
21.9% #12/29
Power play %
17.6% #20/29
93.6% #1/29
Penalty kill %
77.3% #15/29
55.6% #1/29
Faceoff %
49.4% #18/29
0.97 #18/29
Attack strength 1.00 = league average
1.04 #11/29
0.83 #1/29
Defense strength below 1.00 = stingier than average
0.98 #11/29
1.21 #11/29
Form points / game
0.99 #15/29
1571 #6/29
Elo
1559 #9/29
Shot quality · 5-on-5 Shot-quality model
50.4%
Expected goals share (xGF%)
51.3%
2.21
xGF / 60
2.71
2.18
xGA / 60
2.56
49.6%
Shot attempt share (Corsi%)
49.3%
51.3%
High-danger shot share
51.4%
8.71
Power play xGF / 60
6.15
5.72
Penalty kill xGA / 60
7.50

Early in the season last year's numbers are blended in (this season weighs games ÷ (games + 15)).

Goaltending Shot-quality model · season stats
Jake Oettinger Goalie priced Logan Thompson
+0.20
GSAx / 60 (regressed)
+0.30
+9.8
GSAx this season
+3.5
+5.7
GSAx last season
+29.3
91.6%
Save %
89.4%
1.78
Goals-against avg
3.02
11
Starts
9

GSAx = goals saved above expected; regressed toward average.

Starting goalies & rest

Dallas Stars
Goalie Jake Oettinger Confirmed
2 days rest
Washington Capitals
Goalie Logan Thompson Confirmed
1 day rest

Form & trends

DAL · last 6
WWWLLW
WSH · last 6
WLWWWL
Home / road splits (per game)
WSH at home
Scores 3.42 · allows 2.92
DAL on the road
Scores 3.17 · allows 2.50

AI Analysis Experimental

This is a meeting between Dallas and Washington with both teams on similar records early in the season. The projected scoring environment is on the higher side, with the teams essentially level in the model's overall rating.

Both clubs have confirmed their starting goaltenders, Jake Oettinger for Dallas and Logan Thompson for Washington. Dallas arrives with two days of rest and Washington with one. Those are the only concrete pieces of team information available.

The remainder of the news feed consists of pregame and projected-lineup headlines without accompanying detail, and a couple of items that appear to reference other fixtures. As a result, late lineup changes, scratches or special-teams adjustments cannot be verified from this feed, and no recorded line movement is available to gauge how the market has reacted.

An AI research pass over our model and recent headlines (). It only uses the model's numbers and the linked sources, is tracked like a model, and does not change the pick. Unlock the game to see which side the research leans to.

Model run Season team statsShot-quality modelSeason statsLeague statsOdds: sportsbook consensus

Community

Loading community picks…

Community picks —
Live thread
Real-time chat during the game.
Open

Discussion 0

Be respectful. Bad takes are fine, bad manners are not.