🏒 NHLForecasts.com
Data-Driven NHL Predictions & Analytics
2025-26 Season Live

2025-10-28 • NHLForecasts

The NHL's overtime format is supposed to be chaos incarnate: three-on-three hockey with wide-open ice and sudden death hanging over every rush. But through the first month of the 2025–26 season, the chaos has been slightly more orderly than you might expect—at least when it comes to who actually wins these bonus-period barnburners.

Nearly a Quarter of Games Reach Overtime

Through October 2025, 39 of the season's first 168 games have needed overtime to settle things—a 23.2% rate that's right in line with recent NHL trends. That's roughly one out of every four games going to the bonus frame, which means plenty of opportunities to test whether the three-on-three format truly levels the playing field.

Favorites Are Holding Their Ground

Here's where things get interesting. When we look at who was favored before puck drop, pregame underdogs have won just 35.9% of those 39 overtime games. Shift the lens to the end of regulation—when we know exactly how the 60 minutes of regulation unfolded—and underdogs at that moment win 41.0% of the time.

Compare that to games that end in regulation, where underdogs win 47.3% of the time based on pregame odds. The pattern is clear: favorites are converting in overtime at a higher rate than they do when games end in regulation. That's the opposite of what you'd expect if three-on-three hockey were truly a coin flip.

What This Tells Us

The early-season numbers suggest that skill still matters in overtime, even with the wild swings that three-on-three can produce. Whether it's better depth allowing teams to roll skilled forwards in the extra frame, or simply the same quality advantages that made them favorites translating to open ice, pregame favorites are slightly better at closing out tight games in overtime than they are at holding leads through 60 minutes.

It's still early—39 games is a decent sample but not definitive—but if this trend holds, it means that betting markets and prediction models might actually be underrating how much pregame favorites maintain their edge when games go to overtime. The chaos is real, but it's not quite the great equalizer it appears to be on the surface.

← Back to Articles