luckadjusted
NHL Week 1 · 2026-27

2025-26 regular season · March 18, 2026 · SO

No betting line for this game, so there is no before-puck-drop or fresh-game number.

Stars at Avalanche, Mar 18, 2026: who deserved to win?

The Stars beat the Avalanche 2–1, but the replays lean toward the Avalanche: they take 6 of 10 with the same day's form. Luck ran their way: shootout alone was worth 50 points of win probability to the Stars.

The game turned on a luck play (SO): the Stars won the shootout, worth 50 points of win probability to the Stars.

Final (SO)

Dallas Stars 2
Colorado Avalanche 1

March 18, 2026 · Ball Arena

Luck index +64 Stars very lucky · −64 Avalanche

Each team’s own shots from this game, resampled into new games 4,000 times, each a goal at its chance, with the power plays the penalties gave and the pulled goalie at the end: with the way both teams actually played, how often does this performance win? How the replay works

Ten replays. Stars win 4, Avalanche win 6.

  1. 1 DAL 4 COL 1
  2. 2 DAL 4 COL 2
  3. 3 DAL 3 COL 2
  4. 4 DAL 4 COL 3
  5. 5 DAL 2 COL 3
  6. 6 DAL 0 COL 1
  7. 7 DAL 2 COL 4
  8. 8 DAL 2 COL 5
  9. 9 DAL 0 COL 3
  10. 10 DAL 3 COL 8

Three other ways to ask

Each answers a different question, so none of them replaces the replay above.

No closing line for this game.

Needs the closing line, which this game has not got.

Every shot redrawn at its chance, overtime and the shootout replayed as near coin flips; the shots themselves and every call fixed. The narrowest view of luck.

How 4,000 replays ended

Stars win 36% of them, Avalanche 64%. The line is what actually happened: Stars by 1.

Show as a table
Final margin in the replaysShare of replays
Stars by 3+9.6%
Stars by 27.2%
Stars by 118.9%
Avalanche by 121.1%
Avalanche by 212.0%
Avalanche by 3+31.1%

Where the luck was

What the luck was worth

Every lucky and unlucky play, weighted by the moment it happened: its swing in win probability, added up by kind. A post hit in a blowout counts for nothing; the same post with a minute left in a tie game can be most of a win.

Swings add up across a game, so in one that went back and forth they can pass 100%. On the scoreboard the same luck was worth +1.8 goals to the Stars: the final margin was Stars by 1, without the luck Avalanche by 0.8. Luck shares are the fitted ones (how much of each kind is luck).
ShootingUnblocked attemptsGoalsExpected goalsAbove expected
Stars2912.04−1.04
Avalanche4913.15−2.15

Regulation with both goalies in. Expected goals: our model of each shot’s chance from where and how it was taken (distance, angle, shot type, rebound, rush, strength), adjusted for the shooter’s and the goalie’s records. A team above expected finished better or met a weaker night in goal; 75% of the gap counts as luck (why that share).

Calls

Every penalty that put a team on the power play, and every goal taken off the board on review, with what it did to the helped team’s chance of winning at that moment. Calls don’t count toward luck by default: who draws penalties repeats from game to game like a skill (how calls are valued). Switch them on below.

WhenCallWin probability for
P1 17:54 Tripping on the Avalanche (Zakhar Bardakov): 2 minutes on the power play +4% DAL
P1 2:50 Holding on the Stars (Jamie Benn): 2 minutes on the power play +4% COL
P3 11:00 Hooking on the Stars (Tyler Myers): 2 minutes on the power play +7% COL
P3 7:33 Too many men on the ice on the Avalanche (COL): 2 minutes on the power play +8% DAL

When it happened

Avalanche’ chance of winning

As low as 0% and as high as 73%. Numbers mark the biggest luck moments, listed below.

Our in-game win probability: the score and the time left, starting from the betting line (or the home edge alone when there is none), with goals at the league’s rates for the score. Overtime is drawn at a fixed width. Shaded above the middle while the Avalanche were ahead in it, below while the Stars were.
#WhenWhat happenedWin probability forNext drive
1 P1 1:27 Cale Makar scored on a snap shot from 24 feetShooting · 21% chance +14% COL –
2 P2 13:42 Jason Robertson scored on a wrist shot from 23 feetShooting · 6% chance +19% DAL –
3 P2 2:17 Brock Nelson’s snap shot from 9 feet was savedShooting · 22% chance +5% DAL –
4 P3 11:00 Hooking on the Stars (Tyler Myers): 2 minutes on the power playPenalty calls +7% COL –
5 P3 9:50 Martin Necas’s snap shot from 24 feet was savedShooting · 20% chance +6% DAL –
6 P3 7:33 Too many men on the ice on the Avalanche (COL): 2 minutes on the power playPenalty calls +8% DAL –
7 P3 6:17 Wyatt Johnston’s snap shot from 19 feet was savedShooting · 19% chance +7% COL –
8 SO The game turned hereThe Stars won the shootoutShootout · shootouts are close to coin flips +49% DAL –

Adjust the assumptions

Think finishing is more skill than we do, or that the calls were luck? Change it and the coin-flip replay reruns in your browser.

Only coin-flips re-rolled

Stars 31% · Avalanche 69%

Final margin
Stars by 1
Luck in goals
1.8 for the Stars
Luck-adjusted margin
Avalanche by 0.8
Replay with same day's form
Stars 36% · Avalanche 64% unchanged: why
How much of each is luck?

0% treats it as skill, 100% as pure luck. The tick is our fit.

What counts
Flip a play

Force the other outcome, in this game and in every replay: a goal becomes a save, a post goes in, a penalty goes the other way.

About these numbers

Play-by-play from the NHL via fastRhockey; expected goals, in-game win probability and the replays are our own models, fitted on every game since 2018-19 (methodology and validation). Blocked shots aren’t replayed (their location is the block’s, not the shot’s). No betting line for this game, so there is no before-puck-drop or fresh-game number.

All week 24 2025-26 games · Stars · Avalanche

More from week 24, 2025-26

Every game of week 24, 2025-26