A team’s ability to control the puck at even strength is the single most important determinant of team quality. Special teams, goaltenders, and even shooting talent all have some effect, but since the vast majority of the game is at even strength and the spread of goaltender and shooter talent is relatively small, even strength puck possession is the key skill. Using the proxy of shot attempts to measure this skill, the Kings blow the Ducks out of the water – Los Angeles is first in the NHL by a significant margin, Anaheim a mediocre 20th.
The Ducks, at 116 points, are also 16 points ahead of the Kings in the standings. That’s a huge gap – I mean, the Kings are 16 points ahead of the Toronto Maple Leafs and Winnipeg Jets and those guys suck. Given what I just wrote about puck possession, how did this happen? As it turns out, there’s a lot of ways a team can get lucky.
Converting Goals into Wins
One major area in luck is how goals are distributed; teams that win a lot of close games will get more standings points than one would expect from their goal differential, but this never lasts. Take out empty netters and Anaheim is at a +60 goal differential, LA at +30. Obviously, 0 is the league average goal differential. Using the six goals = two points heuristic, we’d expect Anaheim to be at 112 points, 20 points above the league average point total (92), and LA to be at 102, 10 points above it. Suddenly the 16-point standings gap is down to 10.
Anaheim has simply been lucky in close games (2nd best win % in the NHL in one-goal games), whereas LA has a nasty habit of losing one goal games in regulation. But these things are simply variance and won’t last. Lest you think record in close games is indicative of a team’s heart, remember that LA was 27th in one-goal games the year they won the cup (a big reason they were an undervalued 8 seed). It doesn’t say anything about team quality.
Converting Shots into Goals
PDO (even strength shooting percentage plus even strength save percentage) is mostly out of a team’s control and will regress over time to roughly 100. The Ducks boast a 102.3 PDO this year, good for 2nd in the league. They’ve shot 9.7% at even strength, nearly a full percentage point higher than the second best team.
Had the Ducks had league-average puck luck at even strength (7.86 sh% and .9214 sv%), they’d have scored 38 less goals and allowed 7 more. Six goals = two points, so this adjustment would cost the Ducks a whopping 15 points in the standings. Now, looking at all even-strength shots in this way is a very crude method, completely discounting score effects, but factoring that in doesn’t make much difference.
Perhaps more problematic is assuming that the Ducks are a league average shooting team. Is that fair? On the one hand, we know that even over a full season team sh% is subject to tremendous variance and shot quality varies little from team to team. On the other, Perry and Getzlaf have historically sustained rather high 5v5 on-ice shooting percentages (note, by the way, that Marian Gaborik is fifth on that list). But both Perry and Getzlaf are way exceeding even their career averages this year, and the Ducks are getting extremely high shooting percentages even from their bottom six. I don’t think either of those things are sustainable, so the Ducks shooting percentage can safely be regressed most, but perhaps not all, the way to league average.
Meanwhile the Kings basically had league average puck luck this year (100.1 PDO), compensating for a terrible shooting percentage with a fantastic save percentage. In all likelihood neither of those things are sustainable (remember, the Kings did the opposite last year), but they cancel out.
Give Anaheim league average shooting luck and they drop another 15 points to 95. If we are generous and say Anaheim’s true shooting talent at evens is 8.5%, well above league average, that would put them around 99 points.
Converting Shot Attempts into Shots
The Ducks have gotten 50% of the Corsis at all even strength situations and 50.3% of the Fenwicks, but somehow 51.6% of the shots. The Ducks get 72% of their unblocked shot attempts on net, but Ducks opponents manage just 68.7%. League average is almost exactly 72%. So it’s not that the Ducks’ shooters are particularly good at hitting the net, but that the Ducks’ opponents seem to be inaccurate.
This is a bizarre effect, and not one the advanced stats community has thoroughly studied. It strikes me as unlikely that the Ducks are forcing their opponents to miss the net; this just doesn’t seem plausible. Then again, the Ducks produced the same effect last year. If this is a talent on their part, then they are slightly better than their Corsi or Fenwick numbers indicate. If it is randomness (as I would guess), they’ve given up seven less goals (and pocketed another two standings points) because of it.
Special Teams
Just a brief note here, as both teams have average special teams units that have produced average results. The Ducks are a net -1 goal in all non-even situations, the Kings -6. That difference is mostly due to LA taking way more penalties than Anaheim.
Conclusion
Please take none of the estimated point totals in this article as absolute indicators of team quality. They are very rough measures. I merely hope to show how random variance influences every part of hockey, from converting goals into wins, shots into goals, and even shot attempts into shots. Anaheim was an average team that happened to get very lucky in a number of ways. It shouldn’t come as a surprise that luck can vault a club like Anaheim to the top of the standings; luck is as important a factor in the final standings as skill.
It would be incorrect to say that, after being so fortunate throughout the year (and in their first round playoff series), Anaheim is due for bad luck. But neither can the Ducks count on their luck continuing. And if the good bounces stop, Anaheim has very little chance against a much better Kings team.