Monday, November 8, 2010

NHL: Points-Per-Game Standings 11/8/10


I like the numbers.

The challenge for me is coming up with numbers beyond the obvious (winning percentage, save percentage) but that also mean something.

I have grown to hate “Power Rankings.” In the first place, what a dumb phrase. B) it’s so subjective, primarily because i. no one wants to do the work to get beyond the obvious numbers and 2. people don’t think it sells unless it’s got something remarkable in it.


Especially early in the season, I look at points-per-game (PPG) as a measure of comparative standings. When you’ve got as much as a 6-game differential in games played (CHI has played 17, BOS 11) it’s neigh impossible to derive any value simply from the points column. Does the Blackhawks’ 17 points really look better than the Bruins’ 15? Don’t you have to standardize to account for that differential in games played? I think you do.

If nothing else, PPG is a much better way to project future standings than just a snapshot of points. Now, let me be clear: neither points nor PPG is a GOOD way to look into the future. But, the nature of being a fan is to try to foretell the future for your team when it’s simply impossible to do so. Why not try to do it as well as you can?

As I said, this is a more interesting exercise earlier in the season. But not too early. There’s just not much in the way of predictive value after two or three games played.

However, with the average number of games played at 13 across the league right now, it’s a good time to start taking a look.

Here’s how the conference standings look right now:

East (team, GP-Pts-PPG)
1. WAS 14-20-1.43
2. PHI 15-20-1.33
3. MTL 14-17-1.21
4. TB 13-16-1.23
5. BOS 11-15-1.36
6. NYR 14-15-1.07
7. OTT 14-15-1.07
8. ATL 14-15-1.07
9. PIT 15-15-1.00
10. CAR 14-14-1.00
11. TOR 13-13-1.00
12. FLA 12-10-0.83
13. NYI 14-10-0.71
14. BUF 15-10-0.67
15. NJ 15-9-0.60

1. STL 12-20-1.67
2. LA 13-20-1.54
3. VAN 13-18-1.38
4. DET 12-17-14.2
5. CHI 17-17-1.00
6. MIN 13-16-1.23
7. CBJ 13-16-1.23
8. DAL 13-16-1.23
9. COL 13-15-1.15
10. ANH 15-15-1.00
11. SJ 12-13-1.08
12. NAS 13-13-1.00
13. CGY 13-12-0.92
14. PHX 13-12-0.92
15. EDM 12-10-0.83

Then, if you take the PPG and multiply it by 82 and then re-rank, you get the following:

East (team PPG*82)
1. WAS 117.14
2. BOS 111.82
3. PHI 109.33
4. TB 100.92
5. MTL 99.57
6. NYR 87.86
7. OTT 87.86
8. ATL 87.86
9. PIT 82.00
10. CAR 82.00
11. TOR 82.00
12. FLA 68.33
13. NYI 58.57
14. BUF 54.67
15. NJ 49.20

1. STL 136.67
2. LA 126.15
3. VAN 113.54
4. DET 116.17
5. MIN 100.92
6. CBJ 100.92
7. DAL 100.92
8. COL 94.62
9. SJ 88.83
10. CHI 82.00
11. ANH 82.00
12. NAS 82.00
13. CGY 75.69
14. PHX 75.69
15. EDM 68.33

Notice the differences. Looking at the East first, Boston’s efficiency in PPG so far translates to a division title, jumping up from 5th place currently. Montreal takes the brunt of the Bruins ascension, dropping from 3rd to 5th.

In the West, Chicago feels a lot of pain, as their inefficiency in PPG so far really catches up to them, dropping from 5th to all the way out of the playoffs. I flip flopped Vancouver and Detroit to match up with the league’s playoff formatting.

So, not a lot of big changes, but what changes there are were interesting. I’ll keep an eye on this as the season progresses.


Unknown said...

I like this method, at least this early in the season. I was noticing too that Chicago had way more games than everyone else and it was really skewing the standings. Nice work. Let's hope this prediction works out in the Wild's favor.

Anonymous said...

I've always said points per game should be used instead of just points. I don't know why you say "early in the season" though. To my mind, any time there are games in hand it should be used. And that is ANY time teams have not played an equal number of games. Thus it should be used all season until it's over.

I wish there was a web site that did points per game standings on a daily basis :)