Three Unanswered Goals in the NHL: How Often Does That Team Win?

Every hockey fan knows the feeling of a building going quiet after three unanswered goals. I wanted to know how often that momentum actually converts into two points, and how much the final score of the run matters. This study covers 17,921 NHL regular-season games, and in 46.4% of them, 8,308 games, one team scored three regulation goals in a row with empty-netters stripped out.

The headline answer: the team that completed the first three-goal run won 92.4% of the time, 87.9% in regulation and 4.5% in overtime or a shootout. That average hides a wide spread, though, and the spread is where the betting value sits. If you want my wider thinking on game-by-game angles, my free ice hockey tips page is where I post it.

Branch AThe size of the lead decides how the game ends

This chart splits the run team's games by its lead at the moment of the third goal. A run that builds a cushion of three or more is close to decisive. One that only takes a team from behind to a one-goal lead is a very different situation, because the run has mostly repaired damage rather than created a margin.

Look at the overtime slices as well as the losses. In the one-goal group a large share of games drift into overtime or a shootout, and those finishes split almost evenly between winning and losing. Across all runs, the team lost in regulation only 2.9% of the time but lost after regulation 4.7%.

My reading: the run itself is not the signal. The scoreboard after the run is. Two teams can both score three straight and be in completely different games.

Branch BFair odds shorten sharply as the cushion grows

Here the win rates are converted into fair prices with no margin. A team leading by 3+ after its run (5,076 games) won 95.9%, a fair price of 1.04, or −2339 in American odds. At a two-goal lead (2,283 games) the rate falls to 90.2% and the fair price to 1.11.

The one-goal state (913 games) is where I pay most attention. A 79.6% win rate gives a fair price of 1.26, or −390. That is a genuine favourite, not a formality, and live prices sometimes treat the momentum as if it were worth more than the scoreboard.

These are averages over every type of matchup, so treat them as a baseline to compare a live price with, not as a verdict on any single game.

Branch CThird-period runs convert at the highest rate

Runs completed in the 1st period (1,265 games) led to a win 90.6% of the time, and in the 2nd period (4,054) 90.7%. Those two are practically identical. In the 3rd period (2,989) the rate climbs to 95.5%.

The logic is clock rather than momentum. An early run leaves plenty of hockey for the opponent to answer, and after the run the opponent replied first 42.2% of the time, slightly more often than the run team scored next at 38.7%. Late in a game there is simply less time for that reply to matter.

Branch DUnderdogs and favourites both jump, but not equally

This chart compares each group's pre-game win chance, taken from a public Elo rating model rather than bookmaker prices, with what happened after a three-goal run. Pre-game underdogs averaged 38.7% before puck drop and won 88.0% after scoring three straight. Near-even teams went from 50.1% to 92.4%, and favourites from 62.8% to 95.3%.

Once a run has happened, the pre-game line can no longer be bet; the comparison only shows how much information the run adds. Team quality still leaks through, and favourites hold on more reliably than underdogs.

The main caveat sits here. Teams that build runs were often the stronger side to begin with, and the rating split only partly separates that. Part of the high win rate reflects who tends to score three in a row, not only what the run does to the game.

Written byDmitri Volkov

I'm Dmitri Volkov, based in Moscow, and I cover ice hockey predictions at cricketscoretips.com — focused on goaltending, special teams and the metrics that move results.

View profile →