PDO numbers by NHL team - Apr 15

Cam Charron
April 15 2013 05:39AM

PDO doesn't stand for anything, but that doesn't mean we can't learn anything from adding up the overall shooting and save percentages for a team at even strength. A layman's explanation for 'PDO' and why we use it can be found here over at the Backhand Shelf. Basically, if a team is playing with a PDO number way higher than 1.000, they're producing above their expected output. If a team is playing with a PDO number below 1.000, they're producing below their expected output. Over the course of a long season, the number will generally correct itself. 

For any daily updates you may be so inclined to find, Hockey Analysis compiles these numbersbehindthenet.ca has a page that offers a team's shooting percentage and a team's save percentage numbers, and we'll use those for weekly rankings here at NHLNumbers. Shooting percentage is the 17th column from the left on BTN's team shots page—the first one to say SPCT. It's cousin, team save percentage, is three columns to the right also saying SPCT. The team shooting percentage needs to be subtracted from 1000 to get the actual number.

I've also included team Fenwick Close % numbers, pulled from behindthenet.ca as well. Treat it as a proxy for the amount of time each team spends with the puck. Any team with a rate over 50% is pretty good and should make the playoffs in an 82-game season unless something goes horribly wrong.   

Here are the team PDO standings through games played Sunday night: 

TEAM Fenwick Close % Team Shot % Team Save % Team PDO
Toronto Maple Leafs 45.16% 11.20% 0.920 1.032
Anaheim Ducks 47.22% 10.20% 0.928 1.030
Pittsburgh Penguins 50.79% 9.70% 0.929 1.026
Chicago Blackhawks 55.57% 9.40% 0.930 1.024
Vancouver Canucks 52.92% 9.10% 0.926 1.017
Tampa Bay Lightning 44.78% 10.70% 0.909 1.016
Montreal Canadiens 53.57% 9.10% 0.923 1.014
Columbus Blue Jackets 44.95% 8.90% 0.924 1.013
Buffalo Sabres 43.88% 8.60% 0.926 1.012
Washington Capitals 47.53% 8.80% 0.924 1.012
Dallas Stars 48.83% 9.90% 0.912 1.011
Boston Bruins 53.45% 8.20% 0.927 1.009
Phoenix Coyotes 50.24% 8.30% 0.922 1.005
Ottawa Senators 51.05% 6.70% 0.936 1.003
Detroit Red Wings 51.75% 7.00% 0.928 0.998
New York Rangers 53.09% 6.70% 0.931 0.998
Winnipeg Jets 50.13% 8.70% 0.911 0.998
Edmonton Oilers 44.28% 7.40% 0.922 0.996
Minnesota Wild 48.41% 7.80% 0.918 0.996
Los Angeles Kings 58.47% 8.30% 0.912 0.995
Nashville Predators 45.97% 7.70% 0.918 0.995
St. Louis Blues 53.20% 8.10% 0.907 0.988
Carolina Hurricanes 52.31% 7.50% 0.911 0.986
San Jose Sharks 52.84% 6.30% 0.922 0.985
New York Islanders 51.17% 8.20% 0.901 0.983
Colorado Avalanche 46.63% 7.30% 0.909 0.982
Philadelphia Flyers 48.03% 7.40% 0.898 0.972
New Jersey Devils 54.08% 6.60% 0.904 0.970
Florida Panthers 50.35% 6.60% 0.903 0.969
Calgary Flames 49.27% 8.20% 0.886 0.968

As always, here are the numbers from last week.

  • Then again, if PDO does regress to the mean, nobody told the Toronto Maple Leafs or Anaheim Ducks. After 40 games or so games, we should expect just 5% of the teams in the league to be outside 1.025 or .975 (great work here by Snark SD). The actual number is 23%, as this is officially a silly season and has made an absolute mess of things.
     
  • It's interesting that the Calgary Flames can't seem to buy a save, particularly since they tried to sell some at the trading deadline. Isn't that what happened to the stock market in 2008? Traders were selling assets they didn't necessarily have?
     
  • The New Jersey Devils could be the best team to miss the playoffs in quite some time. That .904 is not particularly flattering on Martin Brodeur or Johan Hedberg.
     
  • Aside from seven outlying teams... everybody else is following the normal dance. The New York Islanders finally got some bounces their way and they're in playoff contention. The Minnesota Wild are doing whatever it is the Minnesota Wild do. The Columbus Blue Jackets have gone on this super duper hot streak and find themselves in 10th place.
     
  • Curious to see what happens to Boston's possession numbers in the absence of Patrice Bergeron. That could dramatically affect Stanley Cup picks. If I had to choose the best all-around player in hockey today, it would probably be Patrice Bergeron. With him out, I think I'll go with Marc-Andre Bergeron.
     
  • San Jose and St. Louis are good "trap teams" this year and one could land in the "not your standard 8-seed" role. Same thing with the New York Rangers out East. Last five years, One seeds have had much tougher competition than Two seeds. Seems to me that isn't real fair.
63811cbf517d2d685ea09e103488ea3a
Cam Charron is a BC hockey fan that writes about hockey on many different websites including this one.
Avatar
#1 JAM
April 16 2013, 09:44AM
Trash it!
0
trashes
+1
0
props

I don't agree with the expectation that PDO to regress to 1.000. I understand that the league-wide average is 1, since sv% and sh% are linked, but by assuming the PDO will regress to 1, you make the claim that all goalies in the league are effectively of equal caliber, or that a team with a better goalie will somehow have a below-average shooting percentage to compensate.

It makes more sense to me to individually compare team sh% and goalie sv% to their career totals and search for discrepancies that way.

Avatar
#2 mattyc
April 16 2013, 12:23PM
Trash it!
0
trashes
+1
0
props

@JAM

I agree - for instance we can look at the Leafs and say "Reimer is has either hit a new level, or he's getting the bounces because he's way above his career SV%" but you can't really say the Flames are unlucky because Joey MacDonald hasn't been saving 92% of his shots, given his career SV% is 90.3%.

Plus you have the notion on SH% that (some, a few) players can actually boost on ice SH% (Stamkos, Tanguay, Crosby, Sedin, Datsyuk). Teams should really only regress to their mean, or individual players to their mean.

Avatar
#3 Kent Wilson
April 16 2013, 01:01PM
Trash it!
0
trashes
+1
0
props

@JAM

Not 100% regression, but 87%...which is high enough to bet on in most instances:

http://www.arcticicehockey.com/2011/12/20/2648333/pdo-regression-to-the-mean-or-why-you-should-ignore-shooting

especially when it comes to teams in 40-odd game samples. Chances are a high PDO is in 9/10ths luck.

Avatar
#4 mattyc
April 16 2013, 01:12PM
Trash it!
0
trashes
+1
0
props

@Kent Wilson

but that 13% is really important!

It seems to me there is still a little bit of subjectivity in using these numbers (as with anything I guess). But we kind of intuitively know Toronto isn't going to shoot 11.5% forever, but we're less inclined to say that Chicago and Pittsburg are in for some regression - is that fair?

My suspicion is that there are players/teams that are in that 13% and even a 1.01 over 2000 shots (25 shots/night * 80 games) is a 20 goal differential.

Avatar
#5 Matt
April 16 2013, 08:40PM
Trash it!
0
trashes
+1
0
props
mattyc wrote:

but that 13% is really important!

It seems to me there is still a little bit of subjectivity in using these numbers (as with anything I guess). But we kind of intuitively know Toronto isn't going to shoot 11.5% forever, but we're less inclined to say that Chicago and Pittsburg are in for some regression - is that fair?

My suspicion is that there are players/teams that are in that 13% and even a 1.01 over 2000 shots (25 shots/night * 80 games) is a 20 goal differential.

I'd argue that both Chicago and Pittsburgh are due for some regression. Chicago's on pace to beat Detroit's 1996 regular season record (62 wins/131 points). Chicago is good but they're not *that* good. Corey Crawford and Ray Emery are both fine goalies, but are they goalies who can sustain a .929 team save percentage? Probably not. Likewise for Pittsburgh. Pittsburgh's top two lines are fantastic and the rest of the team is solid, but Fleury's not a particularly great goalie and Vokoun is getting old - they're probably not going to maintain a .928 SV% over a full season.

Avatar
#6 Patrick D. (SnarkSD)
April 17 2013, 10:26PM
Trash it!
0
trashes
+1
0
props

Good discussion here;

I would say that we expect regression towards 1, not to 1. Which is a huge difference both statistically and in terms of wins.

Secondly, just a minor clarification. The figure linked above is PDO accounted for by chance. It turns out the expected SD for 40 games is about +/- 0.016. Therefore we only expect 5% above 1.032 and below 0.968. There are exaclty 2 teams at those numbers so apprx 2/30 = 6.7%. As expected.

As Cam suggested above, we expect just 5% above 1.25 or below 0.975 by chance alone, and we have 23%. So those teams are probably doing something outside of chance to inflate/deflate their PDOs (in addition to get some ridiculous luck).

Avatar
#7 Jason
April 19 2013, 10:37AM
Trash it!
0
trashes
+1
0
props

The "Super Duper Hot Streak" by the CBJ corresponds to a one week/3 game PDO of 1.112 by my calculation, with Bob saving at .971 and the team shooting at .141.

Extrapolating from SnarkSD's graph, that would be right around the 3sd boundary for a three game stretch, and the one tailed probability of hitting that is one in a thousand. 30 * 12 weeks of play means that there are 360 weekly PDOs for teams, which suggests to me that the CBJ weekly PDO might be the highest the league has seen this year... Any data or intuition on this?

Comments are closed for this article.