PDO And What It Means

When players have career years and perform beyond expectations, a very important question arises: is it sustainable? This is especially true during the summer months when the free agent frenzy is in full swing. While the statistical analysis to this question isn't perfect, PDO gives us an opportunity to make a strong prediction as to whether a player will be able to replicate one year's success.The recent Matt Cooke signing provides a nice example to examine through the lens of PDO.

Definition and analysis after the jump...

The creator of PDO is Vic Ferrari, a prominent hockey blogger and statistical analyst. He defined PDO as the the sum of a team's even strength shooting percentage and even strength save percentage (this analysis is easily applied to players as well). The idea is that if a team or player has a high PDO, it's going to come down in the future; if it's a low PDO, you can expect an increase in performance. Now that's great and all, but do the predictions PDO makes actually come true?

Regarding team PDO's, Tyler Dellow over at mc79hockey put together a ton of information in this article (I encourage everyone to check it out). He collected quarterly PDO numbers for the 20 best and 20 worst 1st quarter PDO teams since the 2003-2004 season. He then looked at how the PDO for each team changed from quarter 1 to quarters 2-4. The results are pretty startling, and summarized in his table below:

The Top 20 teams saw an average PDO decrease of 2.6% over the following three quarters of hockey, while the bottom 20 teams saw a 2.8% increase in PDO over the same time period. More specifically, 19 of the 20 top teams saw a drop in PDO and all but two of the bottom 20 teams saw a rise in PDO. The correlation wouldn't be as strong if you added the rest of the teams because you'd be adding randomness when looking at the average teams. But the two points to take away from this are that PDO is a great predictor of future performance, and that it works even better for those at the extremes.

But how does PDO do when analyzing individual players? I'm going to do a similar analysis by looking at the top 40 and bottom 40 players in terms of PDO in 2007-2008. I'll then compare their PDO's from that season to their average PDO over the next two years. I'm also going to limit my sample to players who played at least 50 games and saw at least 10 minutes of ice time at even strength for the last three seasons. The reason for this is that PDO needs a lot of games to work, and since a cutoff needed to be chosen, I picked those numbers. The results:

As you can see from the table, the change in PDO for the top and bottom players is even greater than the change in team PDO in the table above. This makes sense given the wider variation of talent among individual players as opposed to whole teams. The correlation is also stronger, as only two players out of the 80 surveyed failed to move in the PDO-predicted direction. There is, however, one cautionary point: when dealing with individual PDO's, one needs to take the context of the player into account. If Eric Godard finishes with a 99% PDO and Sidney Crosby has a 101%, does that mean Godard will get only get better and Crosby will get worse? No, because the talent differential between these two players means that they have different average PDO's. Since Crosby is one of the best players in the world, he'll have an above average PDO. That should be the benchmark by which you judge him, not the general 100% average. Yet discovering each player's "true" average PDO is very difficult, and the sample sizes only become big enough once a player is about 30 years old.

What does Matt Cooke have to do with this? Well, it was just announced that he signed a 3 year, $5.4 million deal with the Penguins, and what the Penguins will be getting from him in the future just became pretty significant. In terms of PDO, Matt Cooke led the Penguins with a 102.4 PDO this year, which put him at 55th in the league. He finished each of the last two year's with a PDO of 101%. Considering that Matt Cooke will be 32 to start next season, and that he doesn't have the elite level of talent that would put his benchmark PDO above 100%, every indication is that his PDO will drop. And since it's unlikely that he'll increase his shot volume that much, his numbers will very likely decrease as well. But, hopefully he continues to do the little things right and proves that Ray Shero knew what he was doing.