By all accounts, Toronto Maple Leafs forward Nazem Kadri has had a phenomenal season. While he has cooled off lately, he’s still put up 42 points in 46 games this season, good for 26th in the NHL. In his first full year in the league, he has more points than Anze Kopitar, Daniel Sedin, Joe Thornton, Corey Perry and Rick Nash and averages much less ice time than all of them. What makes it more impressive is he put up these numbers spending much of the season with guys like Clarke MacArthur and Colton Orr as his linemates. Only recently has he had the luxury of being paired with Joffrey Lupul and Nikolai Kulemin. Not bad for a “bust”.
The only problem for Kadri and the Leafs is that his numbers have been inflated by luck – or in a more politically correct term for Leafs fans, randomness. PDO is a stat that measures randomness (luck) by looking at even strength shooting percentage and save percentage. The logic behind the stat is that five on five shooting and save percentages fluctuate wildly from year to year. If it’s high one year, it means you were likely getting a lot of lucky bounces (in terms of your team shooting the puck and stopping it) and that will be reflected in inflated points and plus/minus totals. Gabriel Desjardins from behindthenet.ca, the Mecca of analytics in hockey, says if he were to encourage fans to get to know any advanced stat, it would be PDO.
PDO combines five on five save and shooting percentages to come up with a number that is above or below the median of 1000. If a player or team’s PDO falls very far from 1000, you can expect a strong regression to the mean and a turnaround in their fortunes, either good or bad, in the long run. The reason why you shouldn’t expect a repeat performance from Nazem Kadri is that his PDO is 1068, good for 4th in the NHL among skaters with 30+ games played.
If you look at the top 10 players in terms of PDO (luck) in the NHL last year (min. 60 GP), only one of them saw an increase in their points per game or plus/minus totals this year. In general, the more points they put up last year, the bigger their regression was this year. Here’s the list:
1. Chris Kelly – PDO: 1056
Points per game 2011/12: 0.48 Plus/minus 2011/12: +33
Points per game 2012:/13: 0.29 Plus/minus 2012/13: -7
2. Mike Fisher – PDO: 1045
Points per game 2011/12: 0.71 Plus/minus 2011/12: +11
Points per game 2012/13: 0.53 Plus/minus 2012/13: +5
3. Matt Halischuk – PDO: 1044
Points per game 2011/12: 0.38 Plus/minus 2011/12: +9
Points per game 2012/13: 0.31 Plus/minus 2012/13: +2
4. Adam McQuaid – PDO: 1044
Points per game 2011/12: 0.14 Plus/minus 2011/12: +16
Points per game 2012/13: 0.14 Plus/minus 2012/13: 0
5. Radim Vrbata – PDO: 1038
Points per game 2011/12: 0.81 Plus/minus 2011/12: +24
Points per game 2012/13: 0.74 Plus/minus 2012/13: +2
6. Martin Erat – PDO: 1037
Points per game 2011/12: 0.82 Plus/minus 2011/12: +12
Points per game 2012/13: 0.56 Plus/minus 2012/13: -6
7. Ray Whitney – PDO: 1037
Points per game 2011/12: 0.94 Plus/minus 2011/12: +26
Points per game 2012/13: 0.93 Plus/minus 2012/13: +4
8. Todd Bertuzzi – PDO: 1036
Points per game 2011/12: 0.54 Plus/minus 2011/12: +23
Points per game 2012/13: 0.42 (7 games) Plus/minus 12/13: +3
9. Valtteri Filppula – PDO: 1035
Points per game 2011/12: 0.81 Plus/minus 2011/12: +18
Points per game 2012/13: 0.41 Plus/minus 2012:13: -5
10. Benoit Pouliot – PDO: 1035
Points per game 2011/12: 0.43 Plus/minus 2011/12: +18
Points per game 2012/13: 0.63 Plus/minus 2012/13: +9
I do realize the sample size is smaller this year and that also hurts plus/minus numbers, but that still doesn’t take anything away from the trend. You put up a ridiculously high PDO one season, you’re due for a regression the next season. Also, the more points you put up, the bigger the regression seems to be percentage wise. Benoit Pouliot is an exception in this case, but that’s only because his PDO has gone UP from 1035 last year to 1062 this year.
This regression is what we can expect from Nazem Kadri. It’s problematic for the Leafs because Kadri is due for a new contract at the end of the season. If Nonis is smart, he’ll try to strike a short term deal with Kadri like P.K. Subban’s that will buy him time and a larger sample size before he decides how much to pay Kadri on a long term deal. The only problem is, Kadri will surely demand big bucks. After all, look at the names and big contracts behind him in scoring this year. But if Nonis doesn’t look at the PDO numbers, Kadri will be the big contract who has a lot of names in front of him next year.