Available in USContent
The goal that decides 9% of all NHL margins
The most consequential goal in NHL betting is the one that happens after the game is effectively decided. About 9% of all goals scored in 2024-25 came with the trailing team’s net empty. Those goals don’t change the moneyline. They change everything else — the puck-line cover, the over/under, the team total, the player goal props. Underdogs covered +1.5 in roughly 60% of games last season, and the empty-net goal is the silent mechanism behind a large share of those covers.
The two-way logic plays out in every late-game scenario. A home favourite leads 3-2 with two minutes left. The empty net opens. If a goal is scored, the cover hits at -1.5 and the over typically lands. If no goal is scored, the cover fails. The punter’s outcome depends on a 90-second window that has its own logic — coach tendencies, line decisions, even ice conditions — most of which doesn’t reach the pre-game line.
What actually counts as an empty-net goal
The technical definition is narrower than most punters realise. An empty-net goal in NHL records is a goal scored when the net is unattended because the goalie has been pulled for an extra attacker — typically in the final 90-180 seconds of regulation. Goals scored when a goalie is briefly absent for other reasons (delayed penalty, mid-shift switch) don’t count under the same statistical category.
The distinction matters because the pulled-goalie scenario is predictable, while delayed-penalty empties are situational and short. About 95% of empty-net goals in any given season come from the late-game pulled-goalie scenario. The remaining 5% are concentrated in the rare cases where a team pulls early — sometimes two or three minutes out, occasionally even four — and gets caught.
For betting purposes, the question reduces to two variables: how often will the goalie be pulled in this matchup, and once pulled, how likely is the empty net to be scored on? The first variable depends on coach tendency and game state. The second depends on the leading team’s offensive zone awareness, the trailing team’s six-on-five structure, and the time available.
One quirk worth knowing: an empty-net goal that puts the leading team up by two doesn’t change a moneyline, but it does change the over/under if the total line was within the new margin, and it directly tips the puck-line cover. A late goal at 17:50 of the third period from the leading team scoring into the open net flips a one-goal moneyline win into a two-goal puck-line cover.
Empty-net frequency and the puck-line math
Home favourites covered -1.5 only 41.8% of the time in 2024-25. Underdogs covered +1.5 about 60%. Those numbers are inversely linked, and the empty-net mechanic is responsible for a measurable chunk of both.
The math: in a typical NHL game, a one-goal margin entering the final two minutes happens about 28% of the time. In those games, the trailing team pulls the goalie roughly 85% of the time (some don’t pull when the lead is at one but the trailing team is on a power play, or when very late penalties make pulling impractical). Of the games where the goalie is pulled, the leading team scores into the empty net about 35-40% of the time. Multiply through: 28% × 85% × 37.5% ≈ 8.9% — close to the 9% league-wide empty-net goal rate.
For the puck-line cover specifically, here’s the chain. A home favourite leads by one entering the final two minutes. There’s roughly a 32% chance an empty-net goal is scored, flipping the one-goal win into a two-goal cover. There’s roughly a 13% chance the trailing team scores the equaliser instead. The remaining 55% sees the score stay at one goal — the favourite wins ML but loses ATS. That breakdown is why the -1.5 home favourite price is structurally tough — the empty-net upside only flips about a third of close games.
The flip side is worth more attention. The +1.5 road underdog wins ATS in three scenarios: the underdog wins outright (about 36% of road dog games), the underdog loses by exactly one with no empty net scored against them (the 55% case above scaled to underdog perspective), or the underdog loses by exactly one even after conceding an empty-net goal (which happens when the empty-net scoring rate is below 50% — and it is). The cumulative cover rate is roughly 60%, and that’s where the structural value lives.
Coach tendencies on pulling the goalie
NHL coaches don’t all pull at the same time. Some coaches pull early — at 2:30 or 3:00 remaining, occasionally even further out — believing the marginal extra possession outweighs the risk of the empty-net goal. Others pull conservatively — at 1:30 or 1:15, accepting fewer offensive zone seconds for less defensive exposure. The decision depends on the game state, opponent quality, and the coach’s prior tendencies.
The aggressive pullers — Tampa Bay under Jon Cooper, Boston under recent regimes, Toronto across multiple coaches — tend to pull at 2:30 or earlier when down by one. Aggressive pulling correlates with slightly higher comeback rates (about 14-15% versus the 13% league average) but also slightly higher empty-net concession rates (around 40% versus 35%). The trade-off favours the trailing team marginally on the moneyline but increases puck-line cover variance for the leading team.
Conservative pullers — coaches who wait until 1:30 or later, more common in the Western Conference and in defensive systems — produce lower comeback rates (around 11%) but also lower empty-net concession rates (around 32%). For the leading team, this slightly reduces the chance of the puck-line cover on a one-goal lead but slightly increases the chance of the straight-up win.
The market doesn’t fully price the coach-specific tendency. A home favourite leading by one with two minutes left against an aggressive pulling coach is more likely to cover -1.5 than the same situation against a conservative pulling coach. The line treats the situations identically. Punters who track coach tendencies — by simple game logs over the past 20-30 fixtures per coach — find small but persistent edges on late in-play markets that account for the empty-net flip.
The cash-out decision before the empty net opens
Live cash-out offers tighten dramatically once a goalie is pulled. A +1.5 road underdog ticket that was offering reasonable cash-out value at the start of the third period might suddenly show a discount of 15-20% once the trailing team pulls their goalie. The operator is hedging against the high-frequency empty-net goal that would resolve the bet at a loss.
The decision framework is straightforward. If you hold the +1.5 underdog ticket and the trailing team you backed has the goalie pulled, holding through the empty-net window is usually +EV unless the operator’s offer is within 5% of fair. The 32-40% empty-net concession rate means there’s a 60-68% chance no goal goes in, and your +1.5 cover holds. The operator’s cash-out is typically pricing the empty-net scenario at 50-55% concession risk, which is over-pricing the danger.
If you hold the -1.5 favourite ticket and the trailing team has just pulled the goalie, the cash-out offer typically jumps to within 5-10% of full value. The operator is now offering closer to fair because they know the empty-net goal would settle your bet. Cash-out at that point is purely a behavioural choice about variance tolerance, not a structural value play. The detailed walk-through of the cash-out math and the price spread mechanic sits in the broader piece on cash-out timing tied to empty-net risk, which covers when the operator’s offer crosses the fair-value line.
EIHL and IIHF differences worth knowing
EIHL coaches pull goalies earlier on average than NHL coaches — often at 2:00 or 2:15 — because the league average for one-goal-back comebacks is slightly higher (around 16-17% versus 13% in the NHL). The earlier pull produces a higher empty-net concession rate against EIHL trailing teams, around 42-45%, which makes EIHL puck-line covers for home favourites slightly more frequent than NHL equivalents.
IIHF Worlds 2025 produced 489,450 spectators across 64 matches at an average of 5.83 goals per game. That high goal rate compresses the empty-net relevance — many IIHF games are decided by 2+ goals before the final two minutes, so the empty-net mechanic kicks in less often. When it does kick in, IIHF coaches tend to pull aggressively (often at 2:30+), reflecting the win-or-go-home tournament structure that values offensive risk over defensive caution.
Frequently asked questions
Two questions punters bring to me about the empty-net mechanic show up regularly. Both touch the practical decision points around late-game tickets.
At what score margin do most NHL coaches pull the goalie?
Almost all NHL coaches pull the goalie when trailing by one or two goals in the final two minutes of regulation. The split: trailing by one, roughly 85% of coaches pull, typically at 1:30 to 2:30 remaining; trailing by two, about 75% pull, usually at 2:30 to 3:30 because they need more time to score twice. Trailing by three or more in the final two minutes, the pull rate drops to around 30% — most coaches accept the loss rather than risk a four-goal margin that compounds losses on subsequent puck-line and total markets. Aggressive coaches like Jon Cooper tend to pull earlier in all situations than the league average.
Should I cash out a +1.5 underdog before the goalie is pulled?
Usually no, if the operator’s cash-out offer is more than 5% below the implied fair value. Once the trailing team has pulled the goalie, the empty-net concession rate of 35-40% means there’s a 60-65% chance no empty-net goal will be scored. Operators typically price cash-out offers as if the concession risk were 50-55%, which over-prices the danger. Holding the ticket through the empty-net window is the higher expected-value choice in most cases. The exception: if your variance tolerance is low or the cash-out offer is within 5% of fair value, taking the offer protects against the moderately worse outcomes without losing significant EV.
