How Hockey ChangedRules
The Overtime Faceoff’s 30-Second Head Start
In nine seasons of 3-on-3, the team that won the opening draw won 52 overtimes in 100. Its edge was gone after half a minute, and only one overtime in twelve ended that fast. A power play carried over from regulation was worth far more: about seven wins in ten in the overtimes that ended before a shootout.
By Edgehalla ResearchPublished 11 min read
Chandler Stephenson won the faceoff. The puck went to Brandon Montour, who “streaked in on goalie Jakub Dobes and beat him with a shot to the upper-right corner,” as the Associated Press put it. Overtime in Seattle had lasted four seconds.
It was March 12, 2025, and the AP called it the fastest goal in NHL overtime history.
Three-on-three overtime starts with one draw at centre ice, and the team that wins it gets the puck first. That night in Seattle, Stephenson’s draw won the game. We went back through nine seasons and more than 1,500 overtimes to see how often the draw does that.
On this page
The draw winner wins about 52 overtimes in 100
Montour’s goal is the fastest in our records too. We have every regular-season overtime from 2017-18 to 2025-26, and no other overtime goal in them came that fast. Hardly any draw decides an overtime the way Stephenson’s did.
We took every game in those nine seasons that opened overtime at 3-on-3, 2,286 of them. Of those, 1,526 were settled before a shootout. The team that won the opening draw won 789 of them, or 51.7%. That is 1.7 percentage points better than a coin flip, and every “point” in this story means a percentage point like that. Suppose the draw meant nothing, and every one of those overtimes was a coin flip. Flip a coin 1,526 times, and heads will almost always land within about 2.5 points of 50%, one way or the other. The draw winner’s 51.7% is inside that, so it could be luck. The home team, for comparison, won 51.6%.
Winning the opening draw is worth about as much as home ice. Starting on a power play is worth far more
Win rate in regular-season overtimes decided before a shootout, 2017-18 to 2025-26
The opening-draw winner won 51.7% of overtimes, close to the home team’s 51.6%; better teams, moneyline favourites and teams starting on a power play won more.
Every bar beats a coin flip by a little, except the last one, which beats it by a lot. The draw sits near the bottom, next to home ice. The better team did more. We rated teams by points percentage. The 2024-25 Jets, for one, took 116 points from 82 games, 70.7% of the points they could have had. The betting favourite, the team the odds favoured at puck drop, did more too. The team that started overtime on a power play did far more than any of them.
Shootouts tell the same story. In the 760 games that went the full five minutes and on to one, the team that won the opening overtime draw went on to win 52.6%. Five minutes after the draw, that could be luck too. Shootouts have their own story, and how often games reach overtime is in The Point for Losing.
The edge lasts about 30 seconds
The draw does show up, for a few seconds. On Jan. 3, 2018, in Detroit, Dylan Larkin won the opening draw from Jean-Gabriel Pageau. The puck went forward, and Andreas Athanasiou and Ottawa’s Erik Karlsson both reached for it. “Athanasiou was able to control it and skated in behind Karlsson on a breakaway,” the AP wrote. He scored on a backhand six seconds in. Two months later in Chicago, Jonathan Toews won the draw from Nathan MacKinnon and tipped in the winner himself, nine seconds in.
The fast goals don’t always go to the team with the puck. On Dec. 1, 2022, in Seattle, Washington’s Lars Eller won the draw from Matty Beniers. Seven seconds later Beniers scored. Of the 125 overtime goals scored inside 30 seconds, the team that lost the draw scored 44.
The draw winner scores most of the first-30-second goals, then half of the rest
Share of overtime goals scored by the team that won the opening draw, by time of the goal, 3-on-3 openings, 2017-18 to 2025-26
The draw winner scored 64.8% of goals in the first 30 seconds of overtime and about 50% of every later goal.
The two green bars are the head start. After half a minute, every bar sits at about 50%, and the draw is forgotten. Only 30 overtimes ended inside 15 seconds, and the draw winner scored 21 of those goals. In the first 30 seconds the draw winner scored 64.8% of the goals. From the 31st second on, it scored about half the goals in every window.
Those fast goals are rare. Only 125 of 1,526 overtimes ended inside 30 seconds, about one in twelve. The table scales that down to 100 overtimes.
| In every 100 overtimes | Overtimes |
|---|---|
| Ended inside 30 seconds | 8.2 |
| Draw winner scored the goal | 5.3 |
| Draw loser scored the goal | 2.9 |
| Each team, if the draw meant nothing | 4.1 |
Go down the table one row at a time. About 8.2 overtimes in 100 end inside 30 seconds. If the draw meant nothing, each team would end 4.1 of them. The draw winner ends 5.3 instead, and the loser 2.9. That puts the winner 2.4 quick wins ahead of the loser, and 1.2 ahead of an even split. Those 1.2 extra wins in 100 add 1.2 points to the draw winner’s record. Its whole edge over a coin flip, all five minutes counted, is the 1.7 points from the start of this story. So most of the draw’s edge, about seven tenths of it, comes from those few quick goals.
Most overtimes run far longer. Half of all overtime goals came before 2:09 and half after. After the first minute came 1,169 goals.
The draw buys the first shot, and the rest of the five minutes evens it out
The head start dies fast because 3-on-3 teams don’t shoot right away. They circle, regroup and wait. “I think possession is huge,” Cory Schneider, then the New Jersey Devils’ goalie, said in 2019. “You may not get a shot (on goal) for 45 seconds, but if you tire the other team out and get fresh legs on – that’s when you can attack.” His coach, John Hynes, agreed: “Often the team that has the puck doesn’t want to give it up.”
Schneider’s “shot (on goal)” is one the goalie has to stop. A shot attempt counts more: it also takes in the shots that miss the net or get blocked by a defender. The draw winner did take the first shot attempt of overtime in 64.3% of games, about two in three. In half of all overtimes, nobody had tried a shot until 33.5 seconds in. Some nights nobody shoots for minutes. On Dec. 20, 2024, Aleksander Barkov won the opening draw for Florida against St. Louis, and nobody on either team took a shot attempt for 3:20.
Most overtimes go more than 30 seconds without a shot attempt, and the draw winner usually takes the first one
Share of the 2,286 overtimes with a shot attempt by each mark, 3-on-3 openings, 2017-18 to 2025-26
Seconds into overtime
Show data
| Seconds into overtime | Either team | Draw winner shot first |
|---|---|---|
| 10 | 5% | 3% |
| 20 | 24% | 20% |
| 30 | 44% | 34% |
| 45 | 66% | 46% |
| 60 | 79% | 53% |
| 90 | 92% | 60% |
| 120 | 97% | 62% |
| 180 | 100% | 64% |
| 300 | 100% | 64% |
The top line, either team, shows how long overtime waits for its first shot: under half the games have one by 30 seconds, about eight in ten by a minute, and nine in ten by a minute and a half. The lower line counts only the games where the draw winner shot first. Wherever the top line runs above it, the draw loser got the first shot away.
The team that took the first attempt won 59.8% of overtimes. Part of that is automatic. In 231 games the first attempt was the winning goal, like Montour’s, so shooting first and winning were the same thing. Leave those out, and 1,295 games remain. The team that shot first won 52.7% of them. Luck alone could just about produce that.
Over the whole five minutes, the draw winner took 53.0% of the shot attempts. Some attempts are better than others. Montour’s four-second wrist shot was the kind that goes in about one time in fourteen. Give every shot odds like that, add them up, and you get a team’s expected goals. The draw winner had 52.5% of them.
The league has noticed the waiting. At the general managers’ meetings in November 2023, NHL.com reported “a growing concern about the regroup in 3-on-3 overtime” from Colin Campbell, the league’s senior executive vice president of hockey operations. The regroup is a team carrying the puck into the zone, then taking it back out to keep it. The fix was not obvious. “We don’t want to stop play,” Campbell said. “We don’t want to have face-offs.”
The best faceoff men don’t win more overtimes
The draw itself buys little. The men who take it might still matter, if teams with great faceoff men are the teams that win overtime. Dylan Larkin is the place to start. He took 109 opening overtime draws in our nine seasons and won 66 of them, 60.6%. That is the fourth-best rate of the 34 players who took at least 50, and nobody ahead of him took as many. Detroit went 34-42 in the overtimes he started that were settled before a shootout.
Athanasiou’s goal off Larkin’s draw in January 2018 left Detroit “1-6 in games decided in overtime,” the AP noted. The Wings won 10 of their 18 opening draws that season. They went 3-11 in 3-on-3 overtimes that ended before a shootout. They finished with 73 points. Over all nine seasons, Detroit won 59.1% of its opening draws, the most of any team that played every season, and lost 59 of its 104 overtimes.
Winnipeg lost draws and won anyway. In 2024-25 the Jets won four of sixteen opening draws and went 10-3 in overtimes settled before a shootout. They also had the league’s best record, 116 points. Over the nine seasons Winnipeg won the fewest opening draws of any team that played all nine seasons, 40.7%, and had the best overtime record, 61.5%. Mark Scheifele won 41.4% of the 70 opening draws he took. He is tied for sixth in overtime goals in our sample with 14, and only five came after his team won the draw. New Jersey, where Schneider and Hynes worked in 2019, won 42.1% of its opening draws and went 48-34.
Kirill Kaprizov has 12 overtime goals in our sample. His team had won the opening draw before only two of them.
| Player | Opening draws taken | Won | Team’s overtime record |
|---|---|---|---|
| Jordan Staal | 110 | 66 (60.0%) | 45-28 |
| Dylan Larkin | 109 | 66 (60.6%) | 34-42 |
| Leon Draisaitl | 108 | 64 (59.3%) | 46-35 |
| Ryan O'Reilly | 106 | 59 (55.7%) | 35-43 |
| Aleksander Barkov | 98 | 52 (53.1%) | 41-24 |
| Anze Kopitar | 97 | 62 (63.9%) | 29-39 |
| Joel Eriksson Ek | 97 | 53 (54.6%) | 39-21 |
| Sidney Crosby | 97 | 50 (51.5%) | 34-32 |
| Bo Horvat | 89 | 37 (41.6%) | 31-27 |
| Nico Hischier | 83 | 35 (42.2%) | 31-21 |
Read down the third column and then the fourth, and they don’t move together. Jordan Staal won 60.0% of his draws, and his teams went 45-28. Larkin won 60.6%, and Detroit went 34-42. Nico Hischier won 42.2% and his Devils went 31-21. Even Larkin, one of the best at it, won about six opening draws in ten. Any one of these lines could be luck. Larkin took 109 opening draws in nine seasons, and luck alone can push a record that size about ten points either way. So look at everyone at once. Across all 34 players with 50 or more opening draws, and across the 32 teams with 50 or more, a better draw record came with no better overtime record at all.
Better teams win a few more draws, and their record does the rest
Team strength is the last place the draw’s value could hide. Good teams might win more draws and more overtimes, so the draw would only look useful because good teams win it. They do win slightly more draws. Teams with the better season faceoff record won 53.4% of opening draws. Teams with the better points percentage won 52.4%. The faster team won 49.6%, a coin flip. Faster here means the team whose skaters hit 20 mph more times a game that season, as clocked by the NHL’s EDGE player tracking.
The better faceoff team wins about 53 opening draws in 100, and the faster team wins half
Share of opening 3-on-3 draws won, 2017-18 to 2025-26
The better faceoff team won 53.4% of opening overtime draws and the faster team 49.6%.
Every team here wins the draw a bit more than half the time except the faster one. The better faceoff team gets about 53 draws in 100, a small tilt.
You would expect that tilt to grow when the faceoff gap is big. A team that wins far more draws all season should win this one more often. It doesn’t. The table splits the games by how far apart the two teams’ season faceoff records were.
| Season faceoff % gap | Games | Better team won the draw | Likely range |
|---|---|---|---|
| Under 1 point | 511 | 53.6% | 49.3-57.9% |
| 1 to 2 points | 504 | 53.8% | 49.4-58.1% |
| 2 to 4 points | 753 | 54.2% | 50.6-57.7% |
| 4 points or more | 518 | 51.5% | 47.2-55.8% |
Start at the top row. When the two teams were less than a point apart on faceoffs all season, the better one won the opening draw 53.6% of the time. Now the bottom row. When they were 4 points or more apart, it won 51.5%. The last column is the likely range. The top row rests on 511 games, so its true rate could sit anywhere from 49.3% to 57.9%, a coin flip included.
Now match the teams up. Take two teams with the same season record. The one that wins the draw wins about 51 overtimes in 100, 1.3 points better than a coin flip. We matched teams two other ways too. By their record in games played before that night, the draw was worth 1.4 points. By the betting odds at puck drop, for the 381 overtimes of 2024-25 and 2025-26, it was worth 1.1. So three ways of matching teams give the draw winner 1.1 to 1.4 points, and each could easily be luck.
The teams’ records are another matter. When one team has the better record, it wins more overtimes, by too much to be luck. A team’s season faceoff record adds nothing. Once you know who won tonight’s draw, how many draws a team won all season tells you nothing more about who wins overtime.
Season by season, the draw winner’s record wanders around 50%
A real edge, even a small one, should turn up season after season. This one comes and goes. The draw winner’s overtime record ran from 43.6% in 2018-19 to 59.3% in 2021-22. Each season has only about 120 to 200 overtimes settled before a shootout, and seasons that small bounce around.
The draw winner’s overtime record wanders around 50% from season to season
Opening-draw winner’s win rate in regular-season overtimes decided before a shootout, 3-on-3 openings. The shaded band is how far luck alone could plausibly move each season.
Show data
| Season | Draw winner wins | Draw winner wins (range) |
|---|---|---|
| 2017-18 | 47.2% | 39.9% to 54.5% |
| 2018-19 | 43.6% | 36.1% to 51.2% |
| 2019-20 | 55.4% | 47.1% to 63.7% |
| 2020-21 | 56.6% | 47.8% to 65.4% |
| 2021-22 | 59.3% | 52.1% to 66.6% |
| 2022-23 | 49.0% | 41.9% to 56.0% |
| 2023-24 | 53.6% | 46.0% to 61.1% |
| 2024-25 | 48.1% | 40.8% to 55.3% |
| 2025-26 | 54.5% | 47.6% to 61.5% |
- Reference: Coin flip (50.0%)
The line crosses 50% back and forth. Only in 2021-22 did the draw winner finish too far above 50% to put down to one season’s luck. Nine seasons of coin flips would bounce around too. These nine bounce a little more than coin flips usually do, but not by enough to rule luck out.
Split the seasons at 2022-23, and both halves give the same answer. From 2017-18 to 2021-22 the draw winner won 52.1% of 783 overtimes. From 2022-23 to 2025-26 it won 51.3% of 743. Matched on season record, the draw winner was 1.7 points better than a coin flip in the first half and 0.9 points better in the second, both inside what luck can do.
The calendar shows how big luck’s swings are. Nothing about the draw changes between 2021-22 and 2022-23, or between any odd year and the next even one. Yet the seasons starting in odd years give 53.9%, and those starting in even years give 48.8%. A swing that size, with nothing behind it but the year on the schedule, is what luck alone can do to this number.
The early goals point the same way in both halves, though there aren’t many of them. From 2017-18 to 2021-22 the draw winner scored 69.6% of the 69 goals inside 30 seconds. From 2022-23 on it scored 58.9% of 56.
The head start that matters is a power play carried into overtime
Everything so far leaves out 185 overtimes that did not open at 3-on-3. Most of them, 163, opened 4-on-3, because a penalty was still running when regulation ended. In those games one team starts overtime with an extra skater. The other 22 of the 185 stay out of the story altogether: the feed recorded odd or missing skater counts at their opening draws, such as 5-on-5.
On Nov. 1, 2019, the St. Louis Blues went to overtime against Columbus with a few seconds of a power play left. Ryan O'Reilly won the draw from Alexander Wennberg, and David Perron scored eight seconds in, the fastest goal off a 4-on-3 opening in our nine seasons. “We know there’s eight seconds (remaining on the power play entering overtime), and I talked to [O’Reilly] and said, ‘Let’s get it going up the ice if we can,’” Perron said. His coach, Craig Berube, described the shot: “It comes off quick.”
A power play carried into overtime wins about seven in ten of the overtimes that end before a shootout
Teams that opened overtime on a 4-on-3 power play won 69.3% of the games decided in overtime and 62.0% of all such games, shootouts included.
The team with the extra skater usually won, and won fast. When the goal came in the first two minutes, it won more than eight times in ten.
Compare that with the opening draw’s 51.7%. Count only the overtimes that ended before a shootout, as we did for the draw. The power-play team won 69.3% of them, about seven in ten. That is about 19 points better than a coin flip, from one extra skater. Now add the games that went to a shootout. The power-play team won 62.0% of all of them, about six in ten. It won 62.2% from 2017-18 to 2021-22 and 61.7% from 2022-23 to 2025-26. That edge is too big to be luck. With only 163 of these games in nine seasons, though, its true size could be a fair bit bigger or smaller.
A penalty can also arrive late, and then the draw is long forgotten. On Nov. 16, 2025, in St. Paul, Vegas’s Brett Howden won the opening draw from Minnesota’s Joel Eriksson Ek. Nobody on either team took a shot attempt until Mats Zuccarello’s at 3:57. At 4:31, Shea Theodore went off for hooking Matt Boldy, “on a call that angered Vegas coach Bruce Cassidy,” the AP wrote. Joel Eriksson Ek won the 4-on-3 draw from Jack Eichel, and Kaprizov scored “on a one-timer from the right side on a power play at 4:50 of overtime.” Howden had won Vegas the opening draw almost five minutes earlier.
Fast teams don’t win more opening draws, and their overtime edge could be luck
A power play gives one team open ice. Speed should do the same at 3-on-3, so do fast teams get more out of the draw? Take the fastest team and one of the slowest. The NHL’s EDGE tracking counts every time a skater tops 20 mph. Colorado led the league in those bursts in every season from 2021-22 to 2025-26, about 30 to 35 a game. Nashville was among the three slowest teams every one of those seasons, at about 14 to 17.
Over those five seasons Colorado won 36 of its 73 opening overtime draws, 49.3%. Nashville won 43 of 80, 53.8%. The slow team won more draws. Across the league, faster teams won 49.6% of opening draws, a coin flip. Speed doesn’t win the puck off the draw.
Colorado still did better in overtime. It went 25-16 in 3-on-3 overtimes that ended before a shootout, and Nashville went 27-30. In overtimes across the league from 2021-22 on, the faster team won 53.4% of 920. But Colorado also had more points than Nashville in every one of those seasons, and better teams win more overtimes anyway. Compare teams with the same points percentage, and the faster team’s edge could be luck.
Colorado and Nashville are only two teams. Our earlier speed research looked at whole games across the league. It rated every lineup before puck drop, from how often its players topped 20 mph the season before. Then it checked those ratings against each game’s scoring chances, its result and the betting odds. Between teams of equal strength, the faster lineup still got a slightly bigger share of the chances, and that held up in both of its tests. Wins were murkier. Speed’s edge in results failed the first test and passed the second, so it is still unproven. That research never looked at overtime by itself. This story uses a simpler measure: each team’s bursts a game in the same season.
The play-by-play can’t see who has the puck between shots
All of this comes from the NHL’s play-by-play feed, and the feed misses things. Barkov’s overtime in December 2024 shows what. The feed has his draw, then no shot attempt from either team for 3:20. It can’t say who carried the puck in that time.
Two more limits. This is the regular season only. Playoff overtime is 5-on-5 with full 20-minute periods, and none of this applies to it. And all nine seasons use one rule set: 3-on-3 with five minutes of overtime, from the third season of the format on. The 760 games that went on to a shootout count only where we say so.
None of those limits touches the early edge. In the first 30 seconds the draw winner scores nearly two goals for every one by the loser. But 91.8% of overtimes last longer than 30 seconds.
The record draw was one of Stephenson’s coin flips
Back in Seattle, Stephenson’s draw went the Kraken’s way, and four seconds later the game was over. No NHL overtime goal has come faster. In most overtimes, half a minute passes before anyone even tries a shot.
Chandler Stephenson took 49 opening overtime draws in our nine seasons. He won 25.
Share this article
Betting tools
Live Edgehalla tools on the same subject.
- NHL betting history How favourites, puck lines and totals have paid since 2024.
- NHL Prop Lab Betting ideas tested against real results and market odds.
- Shots on goal props Projected shots for tonight’s skaters against the market line.
- Anytime goal scorer Tonight’s most likely goal scorers, by our model.
Show the work
- Tests written down
- Not in advance (exploratory)
- Results computed
Every NHL regular-season game from 2017-18 to 2025-26 that reached overtime (2,471), from our NHL play-by-play, with overtime events ordered by clock time. The opening overtime draw’s winner is the faceoff event’s owning team; its taker is the event’s winning player. Team strength is each team’s season points % with the game itself left out (checked against points % from earlier games only) and season faceoff % with the opening draw left out; closing moneylines for 2024-25 and 2025-26. "Worth X points" = the draw coefficient of a logistic model of the home team winning (draw coded +1/−1, with the points % and faceoff % gaps), turned into a win probability for the draw winner between equal teams. Chance band: 1.96 × √(0.25 ÷ 1,526) = 2.51 points. Season spread: chi-square 14.19 on 8 degrees of freedom around the pooled rate (p 0.08). Player and team lines count opening 3-on-3 draws, grouped by player id; a taker’s "team record" is his team’s record in those overtimes when settled before a shootout; the no-link checks are correlations of draw % with overtime win % across 34 players with 50+ draws (r 0.07) and 32 teams with 50+ (r -0.05). Holdout: 2017-18 to 2021-22 against 2022-23 to 2025-26, plus odd and even seasons.
Continues below
Strength measures: a team’s points % is for the whole season with this game left out, so it includes games played later; the earlier-games-only check is the strictly pre-game version. Faceoff % leaves out only this game’s opening draw. The closing moneyline prices the whole game: in the draw + moneyline model the log-odds slope is 0.58, so only about six tenths of a favourite’s edge carries into overtime. "Likely range" in charts and tables = 95% interval. The quick-finish arithmetic: the draw winner ends 5.3 and the loser 2.9 overtimes in 100 inside 30 s; half the 2.4 difference (1.2) is the draw winner’s gain over an even split, against a total edge of 1.7 points. Speed scene: Colorado and Nashville bursts a game and league ranks come from edge-bursts.csv, written by the build script from its otfo-edge-* cache (edge_team_season, bursts_20 ÷ gp); their draw and overtime records from team-season-draws.csv.
| Test | Result | Verdict |
|---|---|---|
| The opening-draw winner wins 3-on-3 overtime | 51.7% (n 1,526, p 0.18); 52.1% and 51.3% by half | Could be luck |
More tests below
| Test | Result | Verdict |
|---|---|---|
| The draw matters after team strength | +1.3 points, p 0.31; points % gap p < 0.001; faceoff % gap p 0.93 | Could be luck |
| Same, with points % from earlier games only | Better team 54.7% (n 1,478); draw +1.4 points, p 0.27 (n 1,508) | Same answer |
| The draw winner scores the very early goals | 64.8% of goals in the first 30 s (n 125); both halves at 58.9% or more | Holds, rarely matters |
| First shot attempt wins, beyond the built-in wins | 59.8% raw; 52.7% without the 231 games where it was the winning goal (p 0.06) | Luck could only just produce it |
| Better faceoff teams win the opening draw | 53.4% (n 2,286); no rise with the size of the gap | Small |
Continues below
| Test | Result | Verdict |
|---|---|---|
| Better draw-takers or teams win more overtimes | r 0.07 across 34 players; r -0.05 across 32 teams | No link |
| Faster teams win the opening draw | 49.6% (n 1,364) | No |
| Faster teams win overtime once strength is level | p 0.12 (n 920) | Could be luck |
| A carried-over 4-on-3 power play wins overtime | 62.0% of games (n 163); 62.2% and 61.7% by half | Holds |
Frequently asked questions
Does winning the opening faceoff in NHL overtime matter?
Only a little. From 2017-18 to 2025-26 the team that won the opening 3-on-3 draw won 51.7% of the 1,526 overtimes decided before a shootout. That could be luck, and most of the edge, about seven tenths of it, comes from the 8.2% of overtimes that end in the first 30 seconds.
Did a faceoff win set up the fastest overtime goal in NHL history?
Brandon Montour scored four seconds into overtime for Seattle against Montreal on March 12, 2025, after Chandler Stephenson won the opening draw. The Associated Press called it the fastest goal in NHL overtime history, and it is the fastest in our records from 2017-18 on.
How long does the average NHL 3-on-3 overtime last?
Half of all overtime goals came within 129 seconds (about 2:09), across 1,526 overtimes from 2017-18 to 2025-26 decided before a shootout. About a third of overtimes (760 of 2,286 with a 3-on-3 opening) went the full five minutes and on to a shootout.
Do better teams win more 3-on-3 overtimes?
Yes, modestly. The team with the better season points percentage won 54.0% of overtimes decided before a shootout, and the betting favourite at puck drop won 56.2% of 381 overtimes in 2024-25 and 2025-26. Both beat the opening draw.
What happens if a penalty carries into NHL overtime?
The team on the power play starts overtime 4-on-3. Those teams won 69.3% of the overtimes that ended before a shootout, about seven in ten, and 62.0% of all 163 such games from 2017-18 to 2025-26 once shootouts are counted, about six in ten. That is the biggest edge we found in 3-on-3.
More in Rules
- Verdict: Data story
Spearing Hit Zero. Hooking Never Went Away.
- Verdict: Data story
Does an NHL Game Really Stop Less Than It Used To?
- Verdict: Data story
A Shootout Shooter Needs More Than 200 Tries Before Skill Outweighs Luck. The Most Anyone Has Taken Is 135.
- Verdict: Data story
How Much Do NHL Referees Change the Power-Play Count? The Busiest Crews Call About 14% More Than the Quietest
- Verdict: Data story
NHL Scoring Doubled in 1929-30 Under the New Forward-Pass Rules
- Verdict: Data story
NHL Overtime Has Risen From 18% of Games to 23% Since 1999
Sources and method
- NHL stats (NHL.com), every season since 1917-18
- NHL API: game play-by-play (faceoffs, shot attempts, skater counts)
- Scouting the Refs: NHL historical rule changes
- NHL Official Rules 2026-27, Rule 84 (overtime: 3-on-3, and a power play at the end of regulation starts overtime 4-on-3, Table 15)
- Wikipedia: overtime in ice hockey (NHL adopted 3-on-3 on June 24, 2015, for 2015-16)
- AP via CBS Sports: Montour has fastest goal in NHL overtime history at 4 seconds (March 12, 2025)
- AP via FOX 2 Detroit: Athanasiou scores 6 seconds into OT, Detroit tops Ottawa 2-1 (Jan. 3, 2018)
- The Hockey Writers, Dan Rice: NHL overtime, the strategy has changed (Schneider, Hynes; April 30, 2019)
- NHL.com, Dan Rosen: overtime rule changes, face-off clarity discussed at GM meetings (Campbell; Nov. 14, 2023)
- NHL.com, Lou Korac: Blues rally past Blue Jackets in OT (Perron; Nov. 1, 2019)
- STLSportsPage, Lou Korac: Perron’s goal eight seconds into overtime (Berube; Nov. 2, 2019)
- AP via CBS Sports: Kaprizov scores on late OT power play, Wild past Golden Knights 3-2 (Nov. 16, 2025)
- NHL API reference (community documentation of the stats endpoints)
Research and analysis, not betting advice. Past results do not guarantee future returns; bet only what you can afford to lose.