The ultimate stat

(This post is part of a fundraiser for 826 Boston, a non-profit tutoring and writing center. For $50, readers can get me to write anything about hockey they want, so donate today to make me say stuff by clicking here. I probably don’t believe the nonsense below, which was requested by Matt Krawczyk.)

Any proponent of the predictive power of things like corsi and PDO will tell you that you tend to get a lot of pushback from idiots who don’t know anything about how these statistics actually work. Such was the case over the weekend as I wrote on Friday, correctly I might add, that the Ducks were steering themselves toward some very rocky shores given the way they’ve played since about November or so: getting outshot but continuing to win at an improbable clip because their even-strength shooting percentage is, to this day, tops in the league.

The thing you hear most from these people is, as explained to me in an email this morning, that wins are “the ultimate stat.” This is, in some respects, true, because at the end of the season all you’re really judged on is how many games you won. But that is of course not the point of corsi or PDO or fenwick or any of the other things tracked by the people who like being able to make more educated guesses about the sport. These, ultimately, boil down to winning in a number of ways, and can be largely used to guess at the likelihood of future wins.

For example, the Ducks having won each of their two games since my article was published does not invalidate my concerns about their ability to be actually competitive in the Western Conference. In fact, it lends credence to them, as Anaheim was 76-53 across those two games, and managed to outscore their opponents 8-5. This is more or less exactly what I’m talking about with why Ducks fans should be so concerned: they got 41.1 percent of the shots on goal in those games, but scored 61.5 percent of the goals. Their shooting percentage was 15.1 percent. Their save percentage was .934. These were in back-to-back games. There is no way this kind of performance is sustainable, and they’ve been extraordinarily lucky.

One Ducks fan who was real cheesed by my post wrote that I did not understand how the concept of “luck” applied in hockey. They argued that luck essentially means a bounce here or there — which I’d agree with — but that it would only impact maybe two or three wins per team per season, and that it would probably even out during that time. So essentially the argument was “between zero and two points are added or subtracted from a team’s score due to luck,” which is of course ludicrous.

The average NHL shooting percentage this season is 8.94, and Anaheim’s is currently 10.16. That’s 216 goals on 2,120 shots. If the Ducks shot at the league average, they’d have about 190 goals, or 26 less than they currently do. That means that shooting percentage, all by itself, has added an additional five wins or so for the Ducks this season alone (the latest data showing that about 5.5 goals of differential equals two points in the standings this year, down a little bit from the six or so that has been the norm for most of the BTN era). The team’s save percentage being .916, three points above the league average of .913, has saved them roughly five goals, close to another full win. So let’s say it’s 5.5 wins added on by luck this year. That’s 11 standings points. All things being equal, they should be roughly fifth in the West, rather than second. That’s how important luck is.

Again, not that it really matters. Wins are wins and they don’t take them away from you because you shot 15 percent. But we’ve seen time and again that shooting even 10 percent over a long period of time is not a thing that teams can do. It’s not controllable. And thus the point of the article — that the Ducks are going to crash and burn in the playoffs — remains unchanged by two regular-season wins on consecutive days, one of which was against the Avalanche.

Not that I expect Ducks fans to get that. Like most others, they don’t understand a thing about these stats. And they don’t bother to learn because of what they’re afraid they might find.

Thank you for reading and supporting 826 Boston. Don’t forget to donate!

One thought on “The ultimate stat

  1. Eric H.

    Ryan: How do those statistics factor in the shooting ability of the various players on each team? For instance, Craig Simpson (extreme outlier obviously) shot at a 23.7% clip for his entire career. Steven Stamkos is in the high teens for his career and at 20% for the season. Is is plausible that the guys who consume a lot of the Ducks’ shots each game are just really good shooters, or do most teams regress to the mean over time, with good shooters being washed out by subpar shooters? My question aside, I do agree that if the Ducks continue to get outshot, they will start to come back to the pack and look like a beatable team.


Leave a Reply

Your email address will not be published. Required fields are marked *