Mandatory Credit: Russell LaBounty-USA TODAY Sports
In our first introduction to NHL advanced stats, we went over Corsi, Relative Corsi, CorsiRelQoT, and CorsiRelQoC and how they can be used to better understand what is going on out on the ice. This week I am going to introduce another advanced stat: PDO. Before we break down what PDO is, understand that unlike baseball advanced stats like WAR, OPS, and BABIP, hockey stats are not acronyms. PDO is simply derived from the message board username of its creator.
Backhand Shelf has a great post explaining some of the basics of PDO, as well as the background story on its creation. In the post Cam Charron defines PDO as:
"PDO, at a team-level, is the simple addition of shooting percentage and save percentage at even strength."
That seems simple enough, but what does PDO really measure? It is often described as a stat to measure team’s overall “luck”, and in a way it does. More accurately, PDO measures regression and sustainability. The theory behind PDO and regression is that baseline of play is a 1.00 PDO. If a team has a PDO that is >1.00, then they are over performing and if their PDO is <1.00, they are under performing. Please note that the stat can be expressed with or without the decimal.
The other part of PDO is the theory of sustainability, in that a team can only be so good or so bad for so long before they start advancing or regressing towards the 1.00 mark.
PDO can then be used to follow several trends on many different scales in hockey. For example, you can track a player’s progress throughout the season and follow their hot and cold streaks. (One tool that comes in handy for this is the Chart Builder over at SportingCharts.com). PDO can show which players traditionally start slow and then increase their performance in the second half of the season. PDO on the individual player scale can also be a useful tool for fantasy hockey (more on that in this week’s fantasy fix).
PDO can also be tracked on a team scale as well. Let’s take a look at the Blue Jackets Team PDO over their entire history:
Data Courtesy of SportingCharts.com
This chart overall shows the idea of regression and the inevitable return to 1.00. In their first season (in which they finished last in the Central Division) the Jackets performed close to the baseline with a PDO of .997. They regressed greatly in their second season to .977, but returned to what is effectively 1.000 in 2002.
Mandatory Credit: Brace Hemmelgarn-USA TODAY Sports
What I did find surprising was the 2008-09 playoff season, in which the Jackets finished with a .996 PDO. History would have you believe that the Jackets over performed that season on the back of stellar play from goalie Steve Mason. However, PDO would say they actually slightly under performed. Even more surprising is the fact that the disappointing 2009-10 season was not that big of a drop, and instead the Jackets slowly regressed to their second lowest PDO over the next two seasons.
Somewhat alarming is the drastic jump in PDO for the 2012-13 season, where we saw the Jackets nearly make the playoffs. Some of this has to do with the smaller sample size due to the lockout. However, it does indicate that the Jackets were over performing in the final month or two of the season. This comes as no real surprise, as Sergei Bobrovsky turned into a Vezina winner, and it seemed everyone on the offense got that needed lucky bounce or critical goal.
When you look at the individual PDOs for the 2012-13 season for the Jackets (minimum 20 games played), you notice that 18 out of 22 players had a PDO over 1.000 . Again, this supports that the majority of the roster was performing above the norm last season and some regression could be expected for 2013-14.
Be sure to check out tomorrow’s Fantasy Fix on how PDO can assist you in making better decisions in your fantasy hockey league!