How Empty-Net Goals Change NHL Final Scores and Margins
Across 17,921 NHL regular-season games, the empty net turns one-goal games into two-goal wins, and it now does so far more often than it used to.
Hockey bettors often price margins as if the scoreboard tells the whole story. In the final minutes it rarely does. When a trailing coach pulls the goalie, the leader often scores into an empty net, and a tense one-goal game goes into the records as a comfortable two-goal win. I wanted to know how often that happens and whether it is changing, using 17,921 regular-season games from 2010-11 to 2025-26, with 2014-15 not included.
The headline answer is that 56.6% of two-goal regulation wins included an empty-net goal by the winner, and in 2025-26 that rose to 74.3%. Before the winner's empty-net goals were added, 42.6% of regulation games were one-goal affairs. At the final horn only 26.3% were. This study feeds directly into my ice hockey score predictions page, where the puck line and correct-score markets depend on exactly this late-game swing.
Period 1Empty nets drain the one-goal bucket into bigger margins
This dumbbell compares 13,820 games decided in regulation at two points: the final score, and the score with the winner's own empty-net goals taken away. The one-goal row shows the biggest gap, falling from 42.6% to 26.3%. The weight moves down the chart. Three-goal wins jump from 15.5% to 28.8%, and wins by 3 or more goals overall go from 30.7% to 47.2%.
The two-goal row barely moves, 25.8% against 26.6%, and that stability is deceptive. Two-goal wins are not untouched by the empty net. One-goal games are being promoted into that bucket while existing two-goal games are promoted out of it into three-goal finals.
Removing the winner's empty-net goals shows the margin the game had before the net was empty.
View the data: regulation wins: margin before and after empty-net goals
| Without winner's EN goals | Final score | |
|---|---|---|
| 1 goal | 42.6% | 26.3% |
| 2 goals | 25.8% | 26.6% |
| 3 goals | 15.5% | 28.8% |
| 4+ goals | 15.2% | 18.4% |
One honest limit applies throughout. The adjusted margin is a description, not a counterfactual, because some of these games would have ended differently with a goalie in net. I read it as a picture of the game state late on, not an alternative history. When I price a favourite to cover a two-goal handicap, I am really pricing two events: the lead holding late, and the empty-net goal arriving.
Period 2Wins that needed the empty net have grown every era
This stacked chart splits all regulation wins into three groups by era. From 2010-11 to 2013-14, 13.0% of regulation wins were 2+ goal wins that only got there through an empty-net goal. From 2022-23 to 2025-26 it was 19.9%. Over the same span, one-goal wins shrank from 30.6% to 22.8% of regulation results.
What strikes me is the third band. Wins by two or more that did not need the empty net hardly move, from 56.4% to 57.3%. Hockey has not become more lopsided in open play. The extra margin is almost entirely added after the goalie is pulled.
View the data: where 2+ goal regulation wins come from
| Won by 1 | Won by 2+ only via EN goal | Won by 2+ without needing EN | |
|---|---|---|---|
| 2010-11 to 2013-14 (3,343) | 30.6% | 13.0% | 56.4% |
| 2015-16 to 2018-19 (3,871) | 27.6% | 16.6% | 55.8% |
| 2019-20 to 2021-22 (2,529) | 24.3% | 19.1% | 56.6% |
| 2022-23 to 2025-26 (4,077) | 22.8% | 19.9% | 57.3% |
Across the whole sample, 23.3% of regulation wins by two or more goals would have been one-goal wins without the empty-net goal. From 2022-23 on, that figure was 25.8%. Many of the results that look like routs in the records were one-goal games deep into the third period.
Period 3Empty-net goals have become far more common since 2010-11
The line chart tracks two series. Games with an empty-net goal rose from 17.8% in 2010-11 to 35.1% in 2025-26, with a peak of 37.0% in 2024-25. The step across the missing 2014-15 season is sharp, from 17.7% to 28.2%. That fits the move towards earlier and more aggressive goalie pulls.
The second line is the one I watch most closely. It shows the share of two-goal regulation wins with a winner empty-net goal, which went from 48.6% in 2010-11 to 74.3% in 2025-26. The path is noisy, with 58.1% in 2023-24 sitting between two higher seasons, and 2012-13 and 2020-21 were shortened campaigns. Even so, the direction is clear.
2014-15 is not in the goal-event data. 2012-13 and 2020-21 were shortened seasons.
View the data: empty-net goals by season
| Games with an empty-net goal | 2-goal regulation wins with a winner EN goal | |
|---|---|---|
| 2010-11 | 17.8% | 48.6% |
| 2011-12 | 19.0% | 40.6% |
| 2012-13 | 18.8% | 48.1% |
| 2013-14 | 17.7% | 40.0% |
| 2015-16 | 28.2% | 56.2% |
| 2016-17 | 22.7% | 47.2% |
| 2017-18 | 25.9% | 55.5% |
| 2018-19 | 30.1% | 59.2% |
| 2019-20 | 27.8% | 63.3% |
| 2020-21 | 28.3% | 59.9% |
| 2021-22 | 32.3% | 66.1% |
| 2022-23 | 30.7% | 66.1% |
| 2023-24 | 31.1% | 58.1% |
| 2024-25 | 37.0% | 68.3% |
| 2025-26 | 35.1% | 74.3% |
When I see a two-goal final in the modern game, my default assumption is that the net was empty when the margin got there.
Period 4Common scores split cleanly by whether the net was empty
The table lists the 10 most common regulation results and the share of each in which the winner scored into an empty net. The one-goal scores are almost clean: 3-2 at 2.4%, 2-1 at 0.7% and 4-3 at 3.9%. Add a goal to the margin and the picture flips. 4-2 and 5-2 both sit at 59.1%, and 5-3 reaches 65.4%.
4-1 is the most common regulation-win score at 10.1% of regulation wins, and 52.3% of those included a winner empty-net goal. 5-1 is the outlier among the bigger margins at 24.4%. That makes sense, because a coach trailing by that much has little reason to pull the goalie.
| Score | Games | Share of reg. wins | Winner scored EN |
|---|---|---|---|
| 4-1 | 1,390 | 10.1% | 52.3% |
| 3-2 | 1,271 | 9.2% | 2.4% |
| 4-2 | 1,241 | 9.0% | 59.1% |
| 5-2 | 1,206 | 8.7% | 59.1% |
| 3-1 | 1,100 | 8.0% | 51.0% |
| 2-1 | 967 | 7.0% | 0.7% |
| 4-3 | 773 | 5.6% | 3.9% |
| 5-3 | 708 | 5.1% | 65.4% |
| 3-0 | 629 | 4.6% | 49.0% |
| 5-1 | 620 | 4.5% | 24.4% |
For correct-score bettors, 3-2 and 4-2 are closely linked outcomes, not separate stories. A large share of 4-2 finals were 3-2 games with the trailing goalie on the bench.
Full-time verdict
With 56.6% of two-goal regulation wins including a winner empty-net goal, treat a two-goal handicap as a bet on a late lead plus an empty-net finish, not on dominance.
4-2 and 5-3 carry a winner empty-net goal in 59.1% and 65.4% of cases. Group them with 3-2 and 4-3 when you judge a narrow-lead scenario.
Games with an empty-net goal reached 35.1% in 2025-26 against 17.8% in 2010-11. Older margin samples understate how often leads now stretch.
Questions bettors ask
How often do NHL games have an empty-net goal?
Across 17,921 regular-season games from 2010-11 to 2025-26, 27.2% had at least one empty-net goal. The rate peaked at 37.0% in 2024-25 and was 35.1% in 2025-26.
How many NHL two-goal wins include an empty-net goal?
56.6% of two-goal regulation wins included at least one empty-net goal by the winner. In 2025-26 that share reached 74.3%.
What is the most common NHL regulation-win score?
4-1 is the most common, at 10.1% of regulation wins, and 52.3% of those included a winner empty-net goal. 3-2 follows at 9.2%, with an empty-net goal in only 2.4% of them.
I'm Min-jun Park, and I cover ice hockey score predictions at correct-score.vip โ focusing on special teams, goaltending and the metrics that actually drive results.
View profile โ
