Everything Lasts Forever: Riding the PDO Pony

Well, gang, they did it again.

The Calgary Flames managed their sixth come-from-behind third period win last night with a 4-3 OT win over the Colorado Avalanche. At this point, 27 games into this season, the Flames have banked a lot of points and won games in dramatic fashion.

The two overriding questions now are:

This one’s a reasonably simple answer: probably not.

If the Flames played infinity games and talent wasn’t a factor (e.g., it was evenly distributed), you’d expect everyone’s PDO to be 100. Granted, they don’t play infinity games and talent isn’t evenly distributed, so the big questions are “When is the correction coming?” and “How severe is it going to be?”

Corsi and, to a larger extent, Corsi Close are our main analytics that have proven to have much predictive value – in the sense that you can guesstimate with some degree of accuracy how well a team will do points-wise based upon their performance.

In Corsi terms, the Flames are 29th in the entire NHL at Corsi – 43.5% – with only Buffalo being worse. To put this in real terms, they are SIX AND A HALF PERCENT WORSE than the league’s average teams. That’s pretty damn awful. Their PDO right now is second in the league – 102.1. If you buy into the argument that talented teams are more resistant to PDO corrections – their talent in puck possession gives them more ability to work through things and makes them less prone to wacky corrections – then the fact that Calgary is downright lousy at puck possession should be worrisome.

In Corsi Close terms, the Flames do slightly better – 45.0% – with Buffalo, Colorado and Columbus being worse. Relative to the league’s middle teams, they’re five percent worse – which is still bad. Their PDO in these situations is actually only 98.8 – Edmonton’s is a lousy 96.9, the poor saps.

In other words, the underlyings suggest that the Flames should be (a) worse overall but (b) slightly better in close situations.

Based on last season’s Corsi Close numbers, I looked at the metric’s predictive value and developed a formula over at the Hockey Writers after running the season’s numbers. (The R-squared is 0.34, which is fairly decent but not great; a LOT of the variation in team points is left unexplained.)

Anyhow, here’s the equation:

Based on their 45.0 Corsi Close %, you would expect the Flames to earn roughly 80.5 points for a full season (or about 49.1% of all available points). They’ve earned 66.7% of available points so far, so there’s a pretty big gulf between expectations and reality. Welcome to professional sports.

There are 55 games left and 110 possible points on the table going forward. At a 49.1% clip – which is what the underlyings suggest the Flames should be at – they earn 54 points and go just a shade under .500. Combined with the 36 points they have in the bank, you’re looking at about 90 points – presuming that the Flames (a) correct immediately and consistently going forward and (b) correct specifically to that level.

Dallas made the playoffs with 91 points last season, so it’s not entirely out of the realm of possibility that the Flames could make the playoffs even if the team’s luck runs dry and they perform closer to their Corsi Close level than to the lofty levels they have been.

Back to the questions!

Presuming that everyone in the West performs as their Corsi Close suggests they should for the remainder of their games, here’s the standings: