The shootout is just a lottery, Toronto is the league's luckiest team, New Jersey has the highest shot quality, and the league's most valuable goalie for the third year running is Tomas Vokoun - those are just a few of the fascinating insights explained in the 7th edition of Alan Ryder's Player Contributions annual.
Every year I eagerly anticipate the release of Alan Ryder’s analysis, and with very good reason. When most sports journalists say something - for example that Jay Bouwmeester is a great defenseman - why should we believe them? Have they played alongside Bouwmeester? Have they coached him, or professionally scouted him over several seasons and hundreds of games? Generally the answer is no, so I struggle to understand why their opinion is worth more than some random guy sitting next to me in a sweaty Oilers jersey.
By contrast, when Alan Ryder asserts that Jay Bouwmeester is the most overhyped defenseman in the NHL, even more so than Zdeno Chara, he's basing that on objective analysis. These aren't opinions, they're conclusions based on an interpretation of the facts, making his insights far more reliable. Let's take a look at what insights Ryder has found for us this year, after a quick recap of PC.
What is PC?
For a quick refresher on PC (Player Contributions, take advantage of our archives. Simply put, PC is an all-encompassing metric, very similar in approach and methodology to GVT. In fact, the “notion of threshold performance is critical to the analysis of individual performance,” so you can think of PC as GVT's more complex brother.
Though Ryder confesses that the nature of the sport makes it very difficult to separate offense from defense, PC can be broken down into its various components just like GVT. While they're closely correlated both offensively (0.91) and overall (0.89) the more detailed approach yields different results for PC defensively (0.77).
While GVT is measured in goals, which can be easily converted to wins, PC is stand-alone (though you can divide it by 8 or 9 to get an equivalent goal-based measurement, depending on the season). This year Ryder introduces the idea of multiplying it by $56,800 to get a measure of a player's value. For example, Crosby's 138 makes him worth about $7.8 million.
Also in this year's edition, Ryder continues to defend the principle that every goal scored or prevented has the same impact on winning percentage by explaining that “the marginal impact of more/fewer goals on win/points is virtually linear,” with the slope being the average number of goals per game. But enough on the methodology, the guy in the sweaty Oilers jersey is getting impatient.
New to this year's version is the introduction of Ryder Points: 5 points for a regulation win, 4 for an OT win, 3 for a shootout win, 2 for a shootout loss, and 1 for an overtime loss and 0 for a regulation loss. Ryder points are an attempt to strip away the impact the NHL's goofy scoring system can have, where 3 total points are awarded in some games, and 2 in others - especially since that third point is essentially awarded randomly.
In fact, if you learn nothing else from the objective analysis hockey, learn that "there is very little evidence that overtime and, especially, the shootout are anything other than elaborate coin tosses." Ryder observed that, in general, poor defensive teams missed the playoffs (exception Montreal), and strong defensive teams were in the playoffs (exception Rangers). It's the randomness in how ties are resolved that explains those exceptions - the Rangers were unlucky and the Canadiens lucky.
For promoters of objective statistical analysis, we sure talk about luck a lot. There's the OT/shootout, the role of luck in +/- is well-established, Gabriel Desjardins has solidified our understanding of luck in shooting percentage, and now Ryder uses luck to explain how teams can have records that vary significantly from their goal differentials.
We may not think of the Toronto Maple Leafs as a lucky team, except for the fact that they can charge the highest ticket prices in the league with a consistently non-competitive team, but for the third year in a row they were the league's luckiest team, based on marginal goals per point. Equally surprising is Chicago's three-peat as the league's unluckiest team.
Ryder concluded that "wins are about 94% predicted by goals for and against, marginal goal totals or goal differentials." In an area where we can probably expect future study, Ryder can only conclude that teams like "Nashville and Ottawa are either very skilled at winning close games or just plain lucky." If it's luck, we should see that regress to the mean soon enough.
As far back as the 1972 Canada-Soviet Summit Series there were scorekeepers recording from where shots were being attempted. Using that information you can determine not only how many shots a goalie is facing, but the overall quality of those shots - something that Ryder uses to significantly improve our understanding of today's players.
Unfortunately Ryder agrees with Tom Awad that "there are serious issues with something as simple at shot counts - the only objective event around the net is a goal. Saves, misses, blocked shots are all subjective events." Unfortunately the accurate recording of attempted shots is not much better today than it was back in 1972.
Ryder studied the severity of the problem by comparing the shot differentials between a team's regular scorekeeping team at home, and the various scorekeepers when on the road. Any difference between the two could only be explained by a difference in the subjective scoring of these events, or in significant changes in styles of play. For example, Anaheim has a 9.4% inflation, and New Jersey an 11.6% deflation.
The classic example of a clear shot recording bias is Madison Square Garden, a claim for which Ryder would receive plenty of backup from Gabriel Desjardins. It would appear that the next quantum leap in the analysis of shot quality will unfortunately have to wait until the NHL addresses this. (For more on Shot Quality, click here)
PC is well-known for placing a high value on a goaltender's contributions. Ryder's conviction that "It is, absolutely, the key position on a hockey team" is a controversial stance in our field. There are even those at the opposite end of the spectrum that conclude that while there are occasionally observed differences between goalies over small sample sizes, that there is no provable difference in their actual underlying ability. Given the huge correction in free agent goaltender salaries this offseason, NHL teams may be drifting away from Ryder's position.
Perhaps the impasse is based on our inability to accurately measure a goalie's performance. When trying to meaningfully compare one goalie to another, we were previously armed with only even-strength save percentage. (Note: We eliminate save percentage on special teams both to avoid punishing goalies on undisciplined teams, and because we see so little consistency or correlation between special teams save percentage and even-strength). With the addition of shot quality, we can use PC, much as we did in this article last season.
Even with Ryder's help you need to be careful when making judgments on goalies - he's very clear that a higher PC score means a player was more valuable, not better, and the distinction is clear: "You won't find impactful goaltending behind a great defense; it just doesn't get the opportunity to shine." For the 3rd year in a row the most valuable goalie in the league was Tomas Vokoun, followed by Jimmy Howard (top rookie), Evgeni Nabokov, Ryan Miller, Henrik Lundqvist and Ilya Bryzgalov.
According to Ryder, the selection of Henrik Sedin for the Hart Trophy was woefully incorrect. The top forward was Sidney Crosby with a PC of 138, followed distantly by Zach Parise and Alexander Ovechkin at 119, then 7 other players, and finally Sedin with 91. Crosby's achievement is all the more impressive because he suffered from "Mats Sundin syndrome" - a great player without anyone comparable to play alongside.
The best defensive forward was Sami Pahlsson. Of the Selke finalists, neither Pavel Datsyuk nor Ryan Kesler made the list, but Jordan Staal squeezed in at #10.
On the blue line, Drew Doughty was ranked as the top defenseman in a tight race over Mike Green, Chris Pronger, Matt Carle and Nicklas Lidstrom - Norris Trophy winner Duncan Keith was 6th. Ryder uncharacteristically overruled PC and awarded Pronger the best defenseman - given how difficult it is to separate the contributions of a common pairing, Ryder felt that it was only reasonable to assume that Pronger was more to credit than his partner Carle.
Both Carle and Pronger ranked at the top defensively, ahead of Andy Greene, Chris Phillips, Lidstrom (who is clearly a cyborg) and Greg Zanon.
Whenever hockey stat analysts offer something up that challenges conventional wisdom, it is often dismissed by the common fan with a wave of a hand and a "pshaw!" Eventually they come around and the conventional hockey wisdom is changed, but not until the challenge is backed up with concrete evidence, and proven to have predictive merit. That's exactly the reason why Alan Ryder is such a valuable contributor to the field, and why everyone with even a passing interest should read the Player Contributions annual.
Robert Vollman is an author of Hockey Prospectus.
You can contact Robert by clicking here or click here to see Robert's other articles.