What Is Points Percentage in Hockey? The P% Stat

Points percentage exists because NHL standings are kept in points, and points aren't handed out evenly. A win in regulation, overtime or a shootout is worth two, an overtime or shootout loss is worth one and a regulation loss is worth nothing, a split you can check against any team's line in the standings. The NHL's own season-by-season standings settings date the loser's point to 1999-2000 and put the last ties in 2004-05, so every season since has been decided without them.

Hockey-Reference defines PTS% as points divided by maximum possible points, for teams and for goalies, and the NHL labels it P%. The league builds it into its tie-breaking procedure, too. The first step is "the fewer number of games played (i.e., superior points percentage)," which only matters while teams have played different numbers of games, so NHL points percentage does most of its work in the middle of a season. By the last night, when everyone's schedule is complete, it's just points in a different coat.

Its biggest moment came during the pandemic. When play stopped in the 2019-20 season, the NHL declared the regular season over and picked its 24-team return-to-play field "on the basis of points percentage at the pause." Detroit's .275 was the worst in the league, and the Wings entered the draft lottery with the best odds of anybody, 18.5%.

It takes two numbers from any standings table, and a third for older seasons. That's the whole recipe:

• Standings points (two for any win, one for an overtime or shootout loss)
• Games played
• Ties, for seasons before 2005-06, worth a point each

Here's a worked example, with invented round numbers. A team with 24 points from its first 20 games has a .600 points percentage. A team with 25 points from 22 games sits a point ahead in the standings, but its rate is .568. The first team's ahead where it counts, and it passes the second on points if it wins either of its two games in hand.

It's printed as a three-digit decimal, the way a batting average is. The standings themselves still sort on points, and percentage steps in only when two teams are level.

Points percentage measures how much of the available credit a team collected, and the catch is the baseline. Every game that goes past regulation hands out three points where a regulation game hands out two, and last season 326 of the 1,312 games did, so the league average came out to .562. Winnipeg finished at exactly .500 and ranked 26th. Break-even is below average now. It's worth remembering every time somebody calls a club a .500 team as if that's a compliment.

Colorado led the league at .738 and Vancouver finished last at .354. Detroit's .561 was the league median exactly, and the Wings were one of four teams tied at 92 points with Utah, Columbus and Anaheim, sorted by the tiebreakers that come after percentage once everyone's played the same number of games.

The playoff line depends on the conference. Ottawa took the East's last wild card at .604 (99 points), seven points clear of Detroit, while Los Angeles got in out West at .549. The Wings' .561 was better than the West's final qualifier and nowhere near the East's. Geography mattered.

Reading the standings before the schedule evens out

When the 2019-20 season stopped, teams had played different numbers of games, so the NHL ranked them on points percentage. That put the Islanders (.588) ahead of Toronto and Columbus (.579 each) even though the Islanders had a point fewer, because they'd played two fewer games. You can run the same check on any midseason table where the games-played column doesn't match.

Comparing seasons of different lengths

Detroit's rebuild reads cleanly in one column. The Wings were at .275 when the 2019-20 season stopped and .429 in the 56-game season that followed, and they've climbed in four of the five seasons since, to .561 last season. Raw points can't make that comparison. The schedule's length changed from season to season along the way, and a team's points total depends as much on how many games it got to play as on how well it played them.

Separating points from wins

How did Los Angeles make the playoffs while winning fewer than half its games? The Kings won 35, a .427 win percentage, and their 20 overtime and shootout losses (the league's most) paid out a point each. Points percentage counted them at .549. Put the two numbers next to each other and you can see how much of a team's standing came from games it lost.

It counts the loser point

An overtime or shootout loss counts as half a win, so teams that reach overtime a lot look better than their win totals. Los Angeles' 20 loser points last season were more than a fifth of its total.

It doesn't care how a team won

A shootout win is worth the same two points as a lopsided win in regulation, and the standings sort that out only in tiebreaks, through regulation wins, when two teams finish a season level on points and have played the same number of games.

Eras don't line up

Before the loser point arrived an overtime loss paid nothing, and until the league stopped allowing ties a game could end level, so older points percentages came out of different point systems. Compare a team with its own era's league average first.

It's a record of results

Points percentage tells you what happened, every bounce included. It doesn't separate a team that won on its goaltending from one that won on its shot share, so pair it with goal differential before you decide a record is real.