Pittsburgh Penguins vs Dallas Stars: Prediction, Analysis & Stats — NHL
This preview was published 9h before kickoff — the pick is only revealed after settlement.
The matchup
Pittsburgh Penguins take on Dallas Stars in the NHL. Both sides arrive with their own story — here is our model's full, data-driven read: win probability, expected goals or points, recent form and the head-to-head record. And at the end: where the model sees betting value — without giving away the actual pick.
What our model predicts
Our model favours Dallas Stars at 62% to win, Pittsburgh Penguins at 38%.
Find the bets where soft books misprice the odds.
See today's value bets — free accountA probability is not a certainty: even a strong favourite drops matches. That is exactly why our edge lives in the precision of the numbers versus the market, not in the direction anyone can see.
Goaltending & scoring
In goal: A. Silovs (SV% 88.9, GAA 3.04) for Pittsburgh Penguins vs C. DeSmith (SV% 90.7, GAA 2.43) for Dallas Stars. Starting-goalie form is one of the highest-variance inputs in the NHL model.
Expected goals: 7.1 (3.16 Pittsburgh Penguins + 3.98 Dallas Stars).
PDO signal (regresses to ~100 over time): Pittsburgh Penguins PDO 107.059 (running hot) · Dallas Stars PDO 97.301 (running hot).
Form & head-to-head
Pittsburgh Penguins have won 2 of their last 5 (L-L-L-W-W, most recent first). Dallas Stars have won 2 of their last 5 (W-W-L-L-L, most recent first).
The last 5 meetings (Pittsburgh Penguins–Dallas Stars): 3–6, 3–2, 3–5, 1–7, 4–2. On a small sample the head-to-head is context, not a dominant signal — the model's depth carries the probability, not a handful of individual results.
What to watch
The model makes Dallas Stars the favourite at 62%, but at 38% Pittsburgh Penguins is far from a write-off. The early stages often decide it: whoever sets the tempo first can swing the probabilities quickly. Watch whether the favourite asserts itself early or the game stays open — that is where the model's lean is won or lost.
Where the value is
Our model has flagged a value bet on this match. We don't publish the selection, the odds, or the stake — the edge is in the precision of the numbers, not the direction anyone can see. A free account unlocks the full pick and live odds; the wider picture is in the NHL hub. New to value betting? See our guide and our free daily pick.
Result & model verdict
Final score: Pittsburgh Penguins 2–1 Dallas Stars. Here is how our model's value bets settled:
- ❌ Over 6.5 @ 2.20 — lost (-2.7u)
Reading Between the Numbers
The final score of 2–1 landed nowhere near the model's 7.1 expected-goals total, and that gap is worth sitting with rather than dismissing. Hockey scoring is inherently noisy — a hot goaltender, a few blocked shots, or a stretch of disciplined penalty-killing can turn a projected shootout into a tight, low-event game. The model's job isn't to call every final score; it's to price the range of outcomes correctly over time.
Both teams carrying an above-100 PDO into the matchup is a subtle but telling detail. When shooting percentage and save percentage are elevated on both sides simultaneously, it usually signals two teams riding variance rather than sustainable process — and variance, by definition, doesn't care which direction it breaks on a given night. That's a reminder to treat hot streaks on either roster as fragile rather than predictive.
The goaltending gap — DeSmith's 90.7 SV% against Silovs' 88.9 — pointed toward Dallas controlling the game between the pipes, yet Pittsburgh still won. That outcome doesn't invalidate the signal; it illustrates how a single strong period or a few timely saves can override a modest percentage-point edge over just one game. Single-game sample size remains the central humility check in any predictive hockey model.
Included with every account — even the free one.