What If the NHL Had Played 82 Games in 2012-2013?

The lockout that wiped out nearly half of the 2012-13 NHL season cost us more than 510 games and the trust of fans who have yearned for the league to learn how to avoid regular labor crises. Truncating the season to just 48 games also put an asterisk of sorts against the achievements of teams like the Anaheim Ducks and the Toronto Maple Leafs: in vastly exceeding expectations, many wondered whether these teams simply caught a hot streak and were able to luck themselves into a playoff spot they couldn’t have earned over the grind of an 82-game season. This analysis over at Fear the Fin provided a great overview of this problem on the eve of the season. On one hand, hockey is such a random sport from game to game that this line of thinking makes sense: over a longer schedule, outcomes attributable to chance are more likely to even out, making game results (and playoff position) more dependent on chance in a short season. On the other hand, both the Ducks and Leafs have carried their high-percentage shooting and strong goaltending over through three-quarters of the 2013-14 season. Much as hockey analysts are fond of predicting that even-strength shooting numbers over 9% are bound to tumble downward, for example, Anaheim has been able to keep it up for nearly 120 games now. In other words, not every team riding a high PDO is the 2011-12 Wild. If everyone had played 82 games last season, would these teams have fallen off a cliff like some have argued?

Photo credit: Flickr user Elsie  esq. Use of this image does not imply endorsement

Fortunately, I’ve developed a method for my Playoff Forecasts that’s well-suited to answering this question: using standard methods of regressing team statistics to the mean, I used 2011-12 season data to estimate values of even-strength goals and shots for and against, even-strength Fenwick events in close-game situations (for and against), and power play and penalty killing data, for an 82-game 2012-13 season. After that, I took 34 games worth of these values and added them to the 48-game data. The table below depicts each team’s estimated Sh%, Sv%, Fenwick Close %, PP% and PK% for a full season, next to the numbers they achieved over 48 games.

82gm_tb1

Some notes:

  • The extrapolation predicts that Toronto’s team Sh% would have fallen from 10.6% to 9.5%, and their goaltending would have dropped from 0.924 to 0.921, had 82 games been played, dropping their PDO from a stratospheric 1030 to a still-very-high 1016. Even with a slight uptick in possession numbers, the model still pegs the Maple Leafs as the second-worst possession team in the NHL last season.
  • Anaheim, on the other hand, would have seen little change in their Sh% (which, of course, has played out over the current season). Their goaltending and power play, though, would have suffered over the additional games.
  • New Jersey’s abysmal luck would have shifted somewhat, but possibly not dramatically enough to affect their playoff fortunes.
  • The projected declines in shooting percentage and goaltending performance for Pittsburgh seem ominous in light of the way both deserted them at times in the playoffs.

But what does all of this mean for the standings at the end of a hypothetical 82-game 2012-13? I explored this in a couple ways: using even-strength and power-play goals for and against (assuming short-handed goals and empty-netters wouldn’t be relevant for this purpose) to calculate a Pythagorean Expectation of each team’s 82-game Pt%; and using the regression equation for season point totals that I’ve been using in my Playoff Forecasts. The final standings estimated by both methods are depicted below.

82gm_tb2

Some things to note:

  • Pythagorean Expectation is an interesting tool, but this is why I don’t like using it with partial-season data: many of the season point totals estimated by this method are implausibly low.
  • That’s right, even with an expected regression in PDO, the Maple Leafs would have likely made last season’s playoffs, if not as the fifth seed. If the Pythagorean estimate is accurate, they would’ve had a much better first-round match-up in the Capitals. If any teams look to have been lucky to make the post-season, they would be the Islanders and the Wild.
  • If the Pythagorean model is correct, we would have had another Pittsburgh-Philadelphia first-round series. If my regression estimate is correct, the Devils would have indeed squeaked into the playoffs.
  • Two teams whose records seemed to sell short their underlying quality: the Rangers and the Kings. Also worth noting: Phoenix was probably unlucky to miss last year’s playoffs.
  • Anaheim probably would have made the playoffs in a longer season, but not with the second-best record in the West.

About Nick Emptage

Nicholas Emptage is the blogger behind puckprediction.com. He is an economist by trade and a Sharks fan by choice.
This entry was posted in Original Analysis and tagged , , , , , , , , , , , . Bookmark the permalink.

2 Responses to What If the NHL Had Played 82 Games in 2012-2013?

  1. Pingback: Can Team Systems Influence PDO? | Puck Prediction

  2. Pingback: UHN│NHL│Montreal: A Look Back on the Canadiens ’13-14 Season | Ultimate Hockey Network

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s