Power Play Percentage in the NHL (PP%)

The power play got its modern shape from one Canadiens hat trick. On Nov. 5, 1955, Jean Béliveau scored three power-play goals against Boston in 44 seconds, all on the same penalty, because a penalized player back then served his full two minutes no matter how many goals went in. The NHL changed the rule for the next season so a penalized player could return once his team gave up a goal, and today's rulebook still ends a minor the moment a short-handed team is scored on.

The percentage came later. The league's stats database carries power-play opportunities for every team from 1977-78 on, and that first season produced a record that stood until 2022-23, when Montreal's 31.9% finally fell to an Edmonton power play that converted 32.4% of its chances. Only five of the 1,282 team seasons since 1977-78 have cleared 30%, and Edmonton's 30.6% last season was one of them.

The formula hasn't moved in all that time. NHL.com's glossary defines it as total power-play goals divided by total power-play opportunities, Hockey-Reference uses the same division, and the league's stats site adds a net version that takes away the shorthanded goals a team gives up while it's on the power play. That's the whole family.

The arithmetic is one division, and the judgment is in what counts as a power play and which goals count toward it, which the rulebook settles for most cases. These are the inputs:

• Power-play goals, scored while a team has more skaters on the ice than its opponent
• Power-play opportunities, one for each man advantage an opponent's penalty hands a team
• Shorthanded goals allowed, for the net version
• Power-play time on ice, for rate versions like goals per 60

Here's a worked example, with invented round numbers. A team draws 250 power plays and scores on 55 of them, so its PP% is 22.0%. It gives up five shorthanded goals along the way, which drops the net rate to 20.0%. Give the same team 200 power plays at the same rate and it scores 44, eleven fewer goals with a percentage that hasn't changed at all.

The rulebook shapes those numbers. A power play on a single minor ends at the first goal, because a goal against a short-handed team wipes out the minor with the least time left (Rule 16.2). A major doesn't end on a goal (Rule 20.2), so a five-minute power play can produce more than one. Coincidental minors don't put either side short-handed, so matching penalties don't hand anyone a power play.

PP% measures conversion. The league as a whole converted 21.1% last season, and the median team sat a hair lower. Edmonton led at 30.6% and Philadelphia finished last at 15.7%, a gap that's worth about 35 goals over an average team's season of opportunities.

Why does a few points of conversion matter that much? Because goals come much faster with the extra skater. MoneyPuck's numbers for last season have the league scoring about three times as often per minute at five-on-four as at five-on-five (7.32 goals per 60 against 2.48), even though the average team spent less than five minutes a night with the man advantage.

Detroit converted 22.6% of its chances, 12th in the league. Volume helped the Wings more than the rate did. Their 248 opportunities tied for the seventh-most in the league, so their 56 power-play goals tied for seventh as well, five spots better than the percentage. MoneyPuck graded Detroit's five-on-four chances higher still, seventh in the league in expected goals per minute at five-on-four.

Does a good power play stay good? Some of it carries over, and a lot of it doesn't. Over the last ten seasons a team's PP% correlated at 0.37 with its rate the following season, which leaves plenty of room for a unit to rise or fall on shooting luck and personnel. Detroit's the nearest example, fourth in the league two seasons ago and 12th last season.

Splitting conversion from volume

PP% ignores how often a team gets the chance, so analysts pair it with opportunities per game. Toronto drew the fewest power plays in the league last season and converted 21.3%, while Florida drew the most, converted 19.5% and scored 52 power-play goals to Toronto's 42. The team with the worse percentage scored ten more goals. If you're judging what a power play is worth to a team, you need both numbers, and the second one depends on how often its skaters draw penalties.

Checking the chances behind the rate

Expected goals at five-on-four show you whether a power play is creating chances or riding a hot stretch of shooting. Across the league last season, PP% and five-on-four expected goals per 60 correlated at 0.72, so the two mostly agree. Colorado is the outlier worth knowing. The league's best five-on-five team by expected goals converted only 17.1% (27th), and MoneyPuck had Colorado's five-on-four unit scoring 40 goals on 50.1 expected.

Netting out shorthanded goals

A power play that gives up shorthanded goals hands some of its work back, which is what net PP% measures. Colorado allowed 13 shorthanded goals, the most in the league, and its net rate fell to 12.2%. Detroit gave up five, and its 22.6% netted out at 20.6%. Washington's 17.8% fell to 13.3%, which turned a below-average power play into one of the league's worst.

It ignores volume

Two teams with identical percentages end up about 15 goals apart if one draws last season's high of 266 power plays and the other draws last season's low of 197. The percentage can't see that gap, and it's the gap that shows up on the scoreboard.

Shorthanded goals don't count against it

Gross PP% never subtracts the shorthanded goals a unit gives up, which is why the NHL publishes a net version beside it and why Colorado's two rates sat almost five points apart last season.

The sample is small

A team gets about 240 power plays a season, so two or three goals either way moves its rate by a full percentage point. That's enough to jump several places in a league where the middle third sat within about four points.

It can't separate finishing from chances

A unit can shoot well above or below the quality of its looks for a whole season. Edmonton's league-best power play scored 63 times on 52.0 expected goals at five-on-four, and the percentage alone can't tell you that 11 of those goals came from finishing above what the model expected.

How long does a power play last?

Two minutes for a minor, unless the power play scores first, since a goal ends the short-handed team's earliest-expiring minor. A double minor is four minutes, and a major runs its full five no matter how many goals go in. By the league's own time-on-ice totals, the average NHL power play lasted about a minute and 39 seconds last season.