nhl power rankings and projections: oct. 22

Two weeks into the season, it’s finally time to do some power rankings apparently. It’s still stupid early in the season to judge or predict anything, but everyone else was doing it so here’s mine.

Most power rankings are done with subjectivity and gut feel. This is not one of those. I don’t think you care who anyone thinks is the best team, because no one really knows, it’s just an opinion. I want to change that and rank the teams as objectively as possible, but even that won’t be perfect either, especially this early in the season.

If you’ve been following me for a while, you probably know I’ve developed a stat that can do exactly what I’ve described, objectively rank teams to a closer approximation of their true ability. It’s certainly not a magic formula by any means, but in the past it’s performed intuitively better than anything else.

This early into the season, POP (or playoff output projection) is prone to a lot of variance considering the volatility of the stats used, meaning it probably won’t be predictive of much. A lot of things will correct themselves with more data throughout the year, especially in the goals and special teams categories. But the ranking below still seems like a decent representation of which team has played best so far. (You can probably tell that I’ve been inspired a bit by 538’s presentation).

The four stats to the side are the things POP uses to generate a ranking. Even if you don’t care about what it spits out, it’s interesting to look at where teams rank league-wide in each individual category. Green is good. Red is not so good. At a glance, it gives you a great indication of strengths (if any) and weaknesses (if any) of every club. It’ll also show you why a certain team is “too high” or “too low” overall.

PR Oct. 22

10 Thoughts On Rankings

1. What is going on in California? The Ducks look legit so far, but the Kings and Sharks are struggling with their possession game, usually a hallmark of both clubs. Both teams have spent more time leading, so that’s obviously a factor, but still surprising to see.

2. There’s something eternally wrong with Florida’s special teams. League worst powerplay and penalty kill last year doesn’t look like it’s been fixed at all.

3. I have my doubts Buffalo continues to be a top 10 penalty killing team, or that Carolina remains a top 10 powerplay, especially without the Staals. Philadelphia’s powerplay is real though. A shame that’s all they have.

4. Dear Colorado, regression hits hard and fast doesn’t it? I’m sure it’s because the team isn’t trying hard enough though.

5. Edmonton and Toronto have really improved their possession game so far, a point of failure in recent years. The goals will even out with it, so that’s not a huge concern. The Leafs have good special teams, so getting break-even possession means they could easily find their way into the playoffs. The same can’t be said about Edmonton though.

6. If you’ve watched any Pittsburgh games, you get the sense that the Penguins can definitely maintain a 47 percent powerplay efficiency. They are lethal.

7. St. Louis hasn’t been getting the bounces, but they’ve looked great so far aside from the spanking they received by Anaheim Sunday. It was the tail-end of back-to-back so we’ll let it slide.

8. Chicago is an unstoppable force at 5-on-5. At one point yesterday against Philadelphia, the shot count was 17-1. Their exceptionally high possession rate probably has something to do with their ability to absolutely murder bottom feeders.

9. Montreal is 6-1-0 somehow. They’re a good team deep down, but I get the feeling Michel Therrien isn’t getting the most out of their talent. That’s weird to say about a team that’s lost one game, but they can be a true elite team and I think they fall back down to earth soon.

10. The Metro division is a scary place to be right now. If the sixth best team was better than the fourth place team in the Atlantic this year, it would not even remotely surprise me. Washington is an interesting sight among the upper echelon of teams, and I think they can stay there too.

Bonus Thought: How does Elliote Friedman at Sportsnet, who I’ve blatantly ripped off here, do 30 of these every week?


In the past, POP has had a pretty high correlation with regular season points. So aside from the power rankings, I will also use it to project where teams will finish at the end of the year. Between games 1-20 and 60-82 it’ll be a fools errand thanks to the small samples of games played and games left respectively.

For a few teams, you’ll probably notice there’s little difference between the projection and their actual point pace. Those teams are playing to their ability and since points and POP correlate, the two won’t be far off. It’s teams that have been getting the bounces (or not) that it corrects the pace for.

Because most teams have only played about five to seven games, making projections is beyond stupid. This is more to show you what this space will be in the future, especially when POP evens itself out.

Screen Shot 2014-10-22 at 9.01.53 AMRight now the spread of teams is between 44 and 122. Usually it’s between 65-95 (graph on the left to illustrate the difference). That’ll correct itself over time, but for the sake of projections I’ll have to adjust that. A team at 122 ends up being an almost undefeated team, which is obviously wrong. No NHL team will ever sustain that for long (the highest was the 2007-08 Wings at 98) so instead of putting the Penguins down for 122, I put them down at 95.5, which is the average for the best team in the league. Ditto for Buffalo, who gets bumped up from 44 to 63 (though that might be too generous for them). I did that for all the teams to get a more balanced projection.

Proj. Oct 22

No this doesn’t mean I actually believe Boston or San Jose don’t make the playoffs. Or that Los Angeles and St. Louis will just barely squeak in. Or even that Philly is worse than Carolina.

The projections, like the power rankings will only get better with more data. Right now there’s no need to panic or celebrate just yet.

It’s still early, there’s a lot of hockey left.



One thought on “nhl power rankings and projections: oct. 22”

say something

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 )

Google+ photo

You are commenting using your Google+ 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 )

Connecting to %s