Defensive Metrics and NHL Penalty Kill Betting

Evaluating Top NHL PK Units and Shorthanded Props

Dallas at home against Edmonton in a midweek slot last March. The total was 6, the books had taken some moneyline action on the Stars, and the public was lining up for the over because Edmonton’s offence looked terrifying on paper. I bet the under at +100 and watched a 2-1 final. The story was not the goalies, though both played well. It was the Stars’ penalty kill — 88.4 % at home across the season, the best in the league — refusing Edmonton’s elite power play four times in five attempts.

Power plays get the headlines. Penalty kills win the bets that matter. The asymmetry is small but consistent: a strong PK depresses opposition goal totals by a meaningful margin, and the totals market is usually slower to price it than it is to price the offensive equivalent.

PK and game totals

The link between penalty-kill rate and game totals is more direct than people credit. Dallas’s 88.4 % home PK in 2024-25 saved roughly twelve goals across their 36 home games compared to a league-average kill — a third of a goal per game suppressed from opposition scoring. That alone should shift the totals line a quarter of a goal lower on Dallas home games, and in many matchups it does. In others, particularly against elite power-play offences, the line does not move enough.

The cleanest under angle on the board is a top-five penalty kill at home against an opponent whose offence relies heavily on power-play production. Edmonton, Tampa and Winnipeg all fall into that latter category — clubs whose five-on-five offence is good but whose scoring profile leans noticeably on special-teams output. When those teams visit Dallas or another top kill, the under is usually well-priced and the public is on the over.

The under does not always cash. PK rates regress, hot goalies behind a strong kill bring noise into the sample, and a single power-play goal in regulation followed by an empty-net late in the third can flip a 5 total above 6 in a hurry. The angle is profitable across the season, not on any given night.

The top PK teams

Dallas’s 88.4 % home PK in 2024-25 was the league benchmark. Their structure was conservative — four-man boxes that collapsed quickly toward the slot, smart sticks in the lanes, and a confident netminder who did not need to overcommit on the cross-ice pass. The result was a kill that gave up perimeter chances rather than high-danger ones, which is exactly the profile that holds up against elite PP units.

Beyond Dallas, the structurally sound kills share a couple of traits worth noting. They have at least one penalty-killing forward who reads passing lanes well — the Patrice Bergeron archetype, even if Bergeron himself is no longer playing. They have a number-one defenceman who can clear the zone under pressure, which is how kills end. And they have a goalie who is comfortable with traffic and can track cross-ice passes without overcommitting.

Teams without those structural pieces tend to post middling kill rates regardless of effort, and their numbers swing wildly from month to month. Those are not the teams to bet under on, because the kill numbers carry more variance than the totals line accounts for.

Shorthanded props and goals

The shorthanded-goal market is thin but occasionally interesting. Across a full season, NHL teams average about 8 to 10 shorthanded goals each — roughly one every ten games. That low base rate makes “team to score a shorthanded goal” a longshot at most books, typically priced around +400 to +600. The value, when it exists, is on penalty-killing teams with aggressive forecheckers who actively pressure the puck-carrier on the PK rather than collapsing.

Those teams — typically the same ones with strong overall PK percentages — produce shorthanded chances at a meaningfully higher rate than the average. A Dallas or Carolina backing the SHG market on a busy power-play night against an undisciplined visitor is occasionally worth a small stake. The juice is the obstacle: at +400, you need to hit better than 20 % to break even, and the actual rate even on the best PK teams is closer to 10 to 12 %.

The associated under on “no shorthanded goals” is usually overpriced because the books load juice on the favourite side of a binary market. Worth tracking, rarely worth betting.

Faceoffs and zone time

A piece of the PK puzzle that nobody talks about is faceoff-win percentage on penalty kills. Winning the defensive-zone faceoff to start a kill cuts off roughly twenty seconds of opposition offensive-zone pressure, which translates into noticeably fewer high-danger chances. The top PK teams almost always have at least one centre with a 55 %-plus faceoff win rate in his own end on the kill.

For bettors this is mostly a tiebreaker, not a primary input. When two teams have similar PK percentages on paper but one has a clearly stronger defensive-zone faceoff centre, the under is slightly safer in matchups where penalties are expected to drive the totals story. It is a small edge, but it is the kind that compounds when stacked with the more obvious inputs above.

One last note. Penalty kills, like power plays, regress. A team running at 88 % through November is more likely to finish at 83 to 84 % by April. The angles described here work best when the underlying structure supports the number — the right personnel, the right scheme, the right goalie. When a high PK number sits on top of mediocre underlying inputs, expect mean reversion, and stop hammering the under once the schedule moves into February.

The totals-side framework that integrates PK with the broader save-percentage story sits in the ice hockey over/under betting guide, which connects this kind of special-teams suppression to the wider totals environment.

Published by the Ice Hockey Betting Tips team.