- Oops!Something went wrong.Please try again later.
- Oops!Something went wrong.Please try again later.
The concept of luck in hockey is a contentious one.
Fans generally don’t like to hear that a goal was a result of a lucky bounce, or that an entire season’s worth of results — in terms of wins and losses — came because a team simply had the puck go their way for the majority of the year. After all, this somewhat assumes that all things in hockey are created equal, when we know that is not the case.
For instance, a team with a lot of skilled players — say, an actual All-Star Team — could take the same number of shots as a team of fourth-liners, from the same exact positions on the ice and expect better results. That stands to reason. A shot from Alex Ovechkin from x location in a high-danger area has a greater chance of going in than one from Zac Rinaldo or even Lee Stempniak. There is significant stratification of all kinds of talent in hockey, but the two that manifest themselves the most when talking solely about statistics at the end of the year are shooting and goaltending.
With this in mind, we can reasonably extrapolate that guys like Patrick Kane, who led the league in points and finished second in goals only behind Ovechkin are generally going to appear to be more “lucky” than the average NHLer when their stats are viewed without context. What that basically means is that Kane had a shooting percentage of 16 percent in all situations last year, nearly doubling the NHL average of 8.5 percent. So, he was super-lucky, wasn’t he? Well, yes and no. Because while he blew the NHL average out of the water, he came in more marginally ahead of his career average to that point (11.9 percent). So even by his standards, Kane had a “lucky” year, but he’s actually not going to be due for some sort of massive regression that sees him go from 106 points to, say, 60 if he stays healthy. He’s been a point-a-game player or better for four straight seasons, and that’s not likely to change even if his shooting percentage regresses hard back to his career norm.
With that having been said, though, it is outlandish to actually predict him clearing the century mark again. He’s good and it’s certainly within the range of his talents and that of his teammates to get him there, but if that’s where your expectations are starting, you’re going to be disappointed. He exceeded his career high in goals by more than 50 percent this season. He almost certainly won’t hit 46 next year. But then again, he might. Because he is that talented.
But it’s also worth noting that because Kane is well outside the bounds of normal NHLers in terms of talent, there are guys who enjoyed the kind of shooting success he did this year. And conversely, there are guys whose goalscoring numbers fell well short of what they “should have” scored.
Corsica’s “expected goals” stat is an interesting one because it shows the average chance a shot will go in based on a number of factors (shot type, distance from net, angle, whether it was a rebound, or on the rush). So we know that for the number of shot attempts a player took from where he was on the ice, given a number of other factors, how many pucks “should” go in. Obviously that doesn’t account for the ability of Ovechkin to snap a wrist shot of a higher quality than that of Rinaldo, but you see the point.
Here, then, is the expected versus actual goal totals for every non-goalie who got on the ice for more than 500 minutes at 5-on-5 this year (nearly 550 of them):
You can see that most players land within a few decimal places per-60 in terms of the number of goals they score versus what they “should have,” but that there are also some pretty significant outliers in there as well. The guy who led the league in goals per 60 was, not surprisingly, Alex Ovechkin, but his expected goals was “only” 17th. Can you expect Ovechkin to consistently outperform his expected goals? You sure can: over the last five seasons, his 5-on-5 actual goals per 60 (1.15) exceeds his expected per 60 (0.95) by a significant percentage. You can chalk that up to talent.
But there are plenty of guys who don’t have Ovechkin’s unique goalscoring talent, led the league in expected goals per 60 (1.19) but only scored 0.88. Taylor Hall was bang-on with Gallagher in terms of goals scored, but only 0.01 behind on expected scoring. That, to me, implies that both should see their numbers return to form next season. (And that’s more bad news for Peter Chiarelli: Seems like he sold low on an elite winger.)
There are, for the most part, a lot of guys you’d expect to see kicking around the top-20 in terms of expected goals per 60. Saad, Skinner, Tavares, Hertl, Ovechkin, etc. But they don’t often show up in the top-20 in terms of actual goals scored. For example, Eric Staal’s actual scoring here was less than half the expected number, which was 16th in the NHL.
Here are the top 20 in expected goals per 60 from last year, but you’ll note that 13 actually underperformed that number. There may be a lot of factors contributing to that, but in general these appear to be guys whose goals numbers will improve next season. The expected number shows they were among the top players in the league in terms of generating good chances in terms of both quality and quantity.
However, when you compare that list with actual goalscoring leaders in the NHL last season, you don’t get a lot of overlap, as you might expect. Ovechkin, Saad, and Toffoli are the only ones on both lists, indicating that even as they outperformed their expected totals, they were putting themselves in good enough positions to keep themselves somewhere near the top of the league in terms of scoring regardless. And the other guys in the top-20 in actual goals per 60 are typically high-skill players whose talent allows them to outperform the league average when shooting from x part of the ice with y type of shot, and so on.
There are, however, a few guys on this list where you look at say, “Oh, yeah, he’s not doing that again.” Devante Smith-Pelly is probably chief among them, but there are others (Jannik Hansen, Brock Nelson, and Vincent Trochek) who don’t seem likely to come close to replicating their number, based on what we know about them.
Actually, that Hansen number is bananas. He scored almost 120 percent more goals per 60 than he “should have” and no one’s talent level is remotely high enough to sustain a disparity like that. There are guys like Marek Zidlicky and Brooks Orpik whose goalscoring numbers were almost 190 percent of expected totals, and in fact almost all guys at the top of that divide are defenders. That makes sense, because the parts of the ice where those guys are shooting come with low shooting percentages in general, and it therefore becomes relativel easy to exceed them.
Hansen, Pastrnak, Kane, Cal Clutterbuck, and Mats Zuccarello are the only forwards whose difference between expected and actual goals was in the top 20. That, to me, says there’s a likely decline coming for all their goalscoring. I would say Pastrnak and Kane probably have the talent level to keep themselves fairly high up in the rankings, but it’s also fair to say they got a lot of bounces to go their way.
And as far as guys who both significantly underperformed their expected goals and have a talent level that therefore indicates more good fortune may be on the way, I would say Colin Wilson (almost minus-68 percent) and Ryan Getzlaf (minus-56 percent) are good candidates for rebound seasons in addition to guys like Staal, Kadri, and Hornqvist.
These are not, however, hard and fast issues. Injuries to players and their teammates, among a million other outside factors, can hamper anyone’s goalscoring performance. It’s less clear what can accelerate it. But the fact that more information is being developed to give us a clearer picture of the “why” behind surprisingly great or subpar seasons is very helpful to getting a better understanding of the sport itself.
(All statistics via Corsica unless otherwise noted.)