Recrutes

Recrutes: Hockey scouting from a professional perspective from Grant McCagg and Habs coverage from a 33 year veteran covering the NHL from Brian Wilde.

Proudly Sponsored By

Call Of The Wilde: Habs PDO Foretells Turnaround

October 10, 2017

By Brian Wilde

Even if you are old school, there is one advanced stat you should not ignore. It practically predicts the future. Not always in the short term, but always in the long term. It’s called PDO and if your favourite team’s PDO is out of whack in the short term, expect it to come back to the norm in the long term.

The PDO is the sum of a shooting percentage and a save percentage, so naturally the mean average of this is 100 – there are only two results for any shot on net: a save or a goal. For example, if a team has an 8 percent shooting percentage, which is the league norm, then the save percentage of the opposition that they have faced is obviously at 92 percent. So a PDO of 100 is, of course, the league average. Side note: PDO doesn’t actually stand for anything. It’s the handle used by Brian King, who was the first to propose it.

I discovered the PDO one year when the Toronto Maple Leafs were making a run to the playoffs and everyone was very excited that the Leafs were legit. Their PDO at the time was totally extended and it was easy to predict that .940 goaltending from James Reimer and Jonathan Bernier was not going to continue, nor was their 13 percent shooting percentage. Sure enough, they collapsed down the stretch in the last 25 games and missed the playoffs as those elevated PDO numbers were unsustainable over a full season.

So on to the Montreal Canadiens at the moment to show you what they are experiencing now can not continue in the long run.

Here early in the season are some of the league PDO numbers, and we have to admit here that the sample size is small. The larger the sample size with the PDO being extended, the better the chance that a move back to the average will be found.

It doesn’t have to turn around tonight for the Habs, but it will not stay like this over the season. This is a 100 percent guarantee. That’s right. Right here… Brian Wilde is giving you a solemn promise that the Montreal Canadiens will not continue with this PDO number and that it will go up from here. That’s a 100 percent promise.

Here are the league leaders in the PDO stat – the most important analytics tool in all of hockey by a wide margin, in my estimation.

TEAM SAVE % SHOOTING % PDO
CHICAGO .943 19.1 114.5
VANCOUVER .943 11.5 110.4
LOS ANGELES .983 9.2 107.5
WASHINGTON .919 16.9 108.0
COLORADO .944 10.0 106.9

* Note: PDO looks at the 5vs5 save percentage and shooting percentage

The Blackhawks lead the league in PDO with a 114.5 on the back of a save percentage that is elevated at .943 and a shooting percentage that is completely unsustainable of 19.1 percent. Remember that the league average of shooting percentage is 8 percent and the save percentage league average is .920. What this means when you break it down is that if the Hawks were winning with a poor save percentage and a poor shooting percentage then you could expect that the wins would continue as their PDO normalized. With these Hawks numbers elevated, especially a shooting percentage that is seriously not even possible by the 10 game mark, never mind over a season, one can easily surmise that the Hawks will not continue on this path.

In fact, the path of all five of these league leaders are not sustainable. Vancouver will not do 11.5 percent shooting as this is a very high number. Los Angeles will obviously not get .983 goalkeeping this season. Washington is not going to continue to score at a 16.9 percent shooting clip. Colorado will not continue to get .944, nor 10 percent as both will drop to averages.

As a real example for Canadiens fans, when the Habs got off to their start of 9 wins in the first 10 games, their PDO was through the roof with a shooting percentage of 14, which is of course unsustainable. So when the Habs began to falter, it was no surprise to the analytics community who follow PDO. The analytics community can practically predict the future using this statistic because over the long run extended numbers will come back to earth. The 14 percent shooting percentage for ten games couldn’t have lasted; nor could the Habs, and as we know they didn’t.

Here are the bottom five in PDO right now in the NHL. Take a close look at the Habs fortunes and how they can not possibly continue this badly over a larger sample:

TEAM SAVE % SHOOTING % PDO
SAN JOSE .870 6.7 94.3
BUFFALO .865 7.1 93.6
DALLAS .882 3.5 91.7
NASHVILLE .879 5.5 91.6
MONTREAL .892 2.7 90.2

Montreal has the worst PDO in the NHL. The Habs also have the worst shooting percentage in the entire NHL at an unsustainable 2.7 percent. When the Habs went through the collapse in December and January when they missed the playoffs with Carey Price injured, all the talk was about the goaltending then, but the shooting percentage during that two month collapse when they won only three hockey games over 58 days was 4.5 percent. That percentage was half of the league average, and it was just as responsible for the collapse as was their abysmal save percentage.

Looking at right now, as bad as that 4.5 percent collapse was then in that horrific year, this year’s shooting percentage so far is much worse at 2.7 percent. Even reverting back to the league average of 8 percent changes the Habs fortunes drastically. Why, even a move from 2.7 to 4.5 would practically double the output, never mind these Habs players finding the league average at least.

Also, does anyone think that Carey Price is going to have a save percentage this season of .892? This abysmal stat is also not sustainable for the Canadiens.

I 100 percent guarantee you that both of those numbers, meaning the PDO, will be stronger as the season progresses. The .892 will turn into a .925 likely and the 2.7 will turn into something much higher.

Even if you hate analytics, the bottom line of those numbers finding a normal and logical path is this: The Habs will score more goals and allow fewer. That always means more wins too.

I, Brian Wilde, 100 percent guarantee it.

 

 

About Brian Wilde

Brian Wilde has worked in hockey since he was 20. He was the rink side host for the Edmonton Oilers at CTV and Ottawa Senators for Sportsnet. He was also lead reporter on the Montreal Canadiens for 17 years at CTV Montreal.

View all posts by Brian Wilde

Reader Interactions

  • Another great Article Brian! Love the content from Grant and yourself. I Truley love it as a habs fan!
    Your content is always honest and true to realism as possible.
    So much content out there today is either one side or the other. Tear the team down or put them on a pedestal.
    I can always count on Recrutes to be in the middle where the gravy is 🙂

    Cheers from Halifax

  • That was a really great explanation on stats and how they work. I don’t care for stats at all, but this was really informative and thoroughly explained. Thanks for that guys!

  • Very true, Brian. What has Montreal’s shooting % been the last couple of years? The current2.7% is silly low but maybe they’re not an 8% team. Maybe they’re a 7% team. Or 9%.

    • 8.5 percent the last 2 years and I would expect they’ll be close to that again considering it’s close to the same line up.

  • Excellent article Brian, this type of insight is exactly what I was hoping for when I signed up. Would you be able to direct me to where to find last year’s PDO per team per 20 game segment. I am interested to see if it can predict the team’s next 20 win record based on reversion to the mean.

    • Well if you can figure out how to get to nhl dot com and find that category, they let you break it down in 20 game segments. They call it something else and not PDO but a four letter designation. But the website is so bad. It’s hard to use it as a guide exactly when the out of whack numbers come back to the mean ….. but they always do.

  • Totally agree with your article. In real life I am really analytic, sometimes maybe even too much. But in hockey, I usually don’t like the analytics. But your PDO theory is something I really like. It make so much sense. Great COTW. Continue the work 🙂

  • This is fantastic. You explain PDO so well. You really are working hard to give us great content! 👍

    Thanks!

  • Some of the Sv% + Sh% add up to the PDO result. Eg LA .983+9.2=107.5
    Some don’t: Eg Col .944+10.0=106.9 (should be 104.4)

    Does the difference have to do with stats from PK and PP?

  • If you could somehow make 23 of me using cloning, my team would have a PDO of 0.000. Things need not regress if you’re truly below average on offense or in goal. They lwill in this case, but some teams are not destined to reach 1.000 due to sheer lack of talent.

  • Best COTW ever. This is something that I will be watching for all season. Thank you once again for your hard work Brian. Recrutes.ca was worth the investment.

  • Herve here, I do love your analysis with these numbers. Let’s hope the habs can actually convert those shooting % up. Let’s make an hypothesis: Paccioretty, Hudon, Gallagher, Galchenyuk, even Hemsky and Pleckanec all ends up scoring a few goals, that might help out. The problem now is will they all “unblock”? In the same time? how?