Do NHL Winning Streaks Continue? Next-Game Win % by Streak
Twenty seasons of NHL results show that hot and cold streaks carry almost no information beyond team strength, with one exception: venue.
Every hockey bettor has felt the pull of a hot team. Five straight wins, the goalie dialled in, the power play humming, and the instinct says ride it. I wanted to know whether that instinct holds up across 23,990 regular-season games from 2005-06 to 2024-25, or whether it simply restates what we already know about the team. It is the first question I check before a streak influences anything on my NHL predictions page.
The headline answer is blunt. Teams on a 5+ game winning streak won their next game 54.5% of the time, while a public rating model expected 55.6%. That is a gap of โ1.1 pts, so the streak did not beat its own rating. Losing streaks behave the same way: 43.1% actual against 43.8% expected. Streaks describe good and bad teams. They do not add momentum on top.
1Streak teams win roughly as often as ratings predict
Each row pairs what the rating model expected with what actually happened, from long losing runs at the top to long winning runs at the bottom. The dots climb steadily as streaks lengthen, which looks like momentum until you see that the expectation dots climb right alongside them. After a single win teams won 50.9% of the time, and after a single loss 49.3%, which is barely separable.
The middle of the chart is where the myth weakens most. Teams on three straight wins won 51.3% against 52.4% expected, so they slightly underdelivered. Teams on three straight losses won 49.5% against 47.6% expected, so they slightly overdelivered. If anything, a mild pull back towards the middle shows up, not a push away from it.
Streak = consecutive wins or losses (OT and shootout losses count as losses) within the season.
View the data: next-game win % by current streak
| Rating expected | Actually won | 95% range | |
|---|---|---|---|
| Lost 5+ (1,778 games) | 43.8% | 43.1% | ยฑ2.3 |
| Lost 4 (1,467 games) | 46.5% | 45.9% | ยฑ2.6 |
| Lost 3 (2,938 games) | 47.6% | 49.5% | ยฑ1.8 |
| Lost 2 (5,833 games) | 48.4% | 49.1% | ยฑ1.3 |
| Lost 1 (11,665 games) | 50.0% | 49.3% | ยฑ0.9 |
| Won 1 (11,667 games) | 50.2% | 50.9% | ยฑ0.9 |
| Won 2 (5,865 games) | 51.4% | 51.1% | ยฑ1.3 |
| Won 3 (2,955 games) | 52.4% | 51.3% | ยฑ1.8 |
| Won 4 (1,498 games) | 53.4% | 53.2% | ยฑ2.5 |
| Won 5+ (1,702 games) | 55.6% | 54.5% | ยฑ2.4 |
My main caveat sits here. The rating model moves with recent results, so part of every streak is already baked into the expectation, and a model probability is not a market price. If a bookmaker's line overreacts to a streak more than the rating does, the edge would be against the streak, but this data cannot prove that directly.
2Long losing runs end with more regulation defeats
This chart splits the next game into regulation wins, overtime or shootout results, and regulation losses. The shape matters for anyone betting three-way or regulation-only markets. Teams on a 5+ game winning streak won in regulation 41.9% of the time and lost in regulation 34.4%.
At the other end, teams on a 5+ game losing streak won in regulation only 32.7% and lost in regulation 46.2%. The overtime slice stays fairly stable across every streak band, so the swing happens almost entirely between the two regulation outcomes.
View the data: how streak teams win or lose the next game
| Regulation win | OT or shootout | Regulation loss | |
|---|---|---|---|
| Lost 5+ | 32.7% | 21.1% | 46.2% |
| Lost 3 | 38.1% | 21.9% | 40.0% |
| Lost 1 | 38.3% | 22.6% | 39.1% |
| Won 1 | 39.0% | 23.3% | 37.7% |
| Won 3 | 39.2% | 23.6% | 37.2% |
| Won 5+ | 41.9% | 23.7% | 34.4% |
I read that as strength showing through, not momentum. Weak teams lose cleanly to better ones, and a long losing run is usually a weak team being exposed. That logic belongs in the price, not in a narrative.
3Home ice decides how much a streak is worth
Split by venue, the picture sharpens. A team riding a 5+ game winning streak won 61.4% at home across 830 games, but only 47.8% on the road across 871 games. The same hot side becomes close to a coin flip once it travels.
Losing streaks follow the same pattern. Teams on a 5+ game skid won 48.4% at home and 37.5% on the road. Venue, travel and last change move these numbers far more than the length of the run.
View the data: streak teams at home and on the road
| At home | On the road | |
|---|---|---|
| Lost 5+ | 48.4% | 37.5% |
| Lost 3-4 | 53.7% | 42.8% |
| Lost 1-2 | 53.7% | 44.8% |
| Won 1-2 | 55.2% | 46.7% |
| Won 3-4 | 56.4% | 47.6% |
| Won 5+ | 61.4% | 47.8% |
This is the part I actually use. When a streaking team is on the road, the narrative around them is loudest while the underlying win rate is at its most ordinary.
4Long streaks are rare, so samples thin quickly
The table counts how often runs reach each length. Winning streaks of 3+ games happened 2,999 times and losing streaks 2,969 times. By five games that drops to 797 winning and 794 losing streaks, and only 41 winning streaks and 53 losing streaks reached 10 or more.
That thinning is why the 5+ bands carry wider uncertainty ranges. I would not draw conclusions about ten-game runs specifically, because the sample simply is not there.
| Length | Win streaks | Loss streaks |
|---|---|---|
| 3+ games | 2,999 | 2,969 |
| 5+ games | 797 | 794 |
| 7+ games | 222 | 235 |
| 10+ games | 41 | 53 |
Betting notes
- Price the team, not the run. A 5+ game winning streak produced 54.5% next-game wins against 55.6% expected. Judge the line on team quality and goaltending, and treat the streak as already counted.
- Fade travelling hot teams carefully. Hot teams won 61.4% at home but 47.8% on the road. A road favourite priced up on a streak deserves extra scrutiny.
- Respect regulation risk on skids. Teams on 5+ game losing streaks lost in regulation 46.2% of the time. In regulation-only markets, a cold side's slump tends to show up in clean defeats.
Quick answers
Do NHL winning streaks continue?
Only at the rate team strength suggests. Teams on 5+ game winning streaks won their next game 54.5% of the time, slightly below the 55.6% a rating model expected.
Should I bet on an NHL team to snap a losing streak?
Not on the streak alone. Teams on 5+ game losing streaks won 43.1% of the time against 43.8% expected, and on the road that fell to 37.5%.
How rare are long NHL winning streaks?
Across twenty seasons there were 797 winning streaks of 5+ games, but only 41 reached 10 or more. Samples that small carry wide uncertainty.
I'm Nathan Brooks, based in New York City, and I cover the best NHL bet of the day at bettips.vip โ focused on goaltending, special teams and shot quality.
All of Nathan Brooks's work โ
