For a better browsing experience, please upgrade your browser.

FiveThirtyEight

Sports

If you read the North American sports media — Sportsnet, CBC, THN, USA Today, CBS, ESPN — you’ve heard that Henrik Lundqvist may as well go on vacation; this year’s Stanley Cup already belongs to the Los Angeles Kings. My quick scan showed 12 of 14 hockey media types picking the Kings to beat the New York Rangers in the NHL playoffs’ final round, which begin on Wednesday, and it’s easy to see why. The Kings have a recent track record of success (a Cup in 2012 and a conference finals appearance in 2013). They come from the stronger conference — the West won 246 games and lost 202 against the East this year — and to get to the finals they had to beat teams that had 111, 116 and 107 points this season. Quite different from the Rangers’ playoff run, which included struggles to beat flawed teams and scrapping against backup goaltenders.

Except it isn’t that simple, and not just because hockey is a sport disproportionately fueled by luck. The Rangers have a case to make — even on paper. The stats give them a real shot.

Let’s start with shooting percentage, where the teams are evenly matched. Both New York and LA struggled this year: The Rangers’ 6.7 percent at 5-on-5 ranked 28th in the league and the Kings’ 6.6 percent was 29th. That’s not a big enough gap to make a difference, because shooting (and save percentages) in hockey are prone to large fluctuations. Given that the teams took about 2,000 shots, that 0.1 percentage point difference represents just two goals, and it’s easy to see how some random bounces could explain it.

That’s not to say that shooting percentage is completely meaningless. Pulling our estimates of a team’s shooting skill two-thirds of the way towards the mean helps account for the impact of random chance. If the Rangers and Kings had huge differences that might tell us something about their differing skill. But they only had a margin of 0.1 percentage points this year and 0.8 over the last three years. Between the change in personnel and systems over time and the limitations of multiyear analysis, the Kings and Rangers are close enough that it’s hard to be confident that either team has an edge in shooting percentage.

But there are differences to be found among the less top-level stats. Much of today’s advanced stat analysis begins with studying teams’ shot differential as an indicator of their ability to control play. In this regard, the Kings do have a clear edge; indeed, over the last few years they’ve been the best puck-possession team in the league.

The Kings outshot their opponents 57 percent to 43 percent during 5-on-5 play this year,1 excluding situations where the score was close enough that teams sat back to protect a lead.2 They led the NHL, but the Rangers weren’t too far behind, outshooting their opponents 54 percent to 46 percent. From these two figures, we might expect the Kings to get something like 51.5 percent of the shots against the Rangers; when we include the somewhat tougher opponents they faced this year, we might revise our estimate upwards a bit to something closer to 52 percent.3

But that was the regular season, and it’s worth testing whether anything has changed in the playoffs. That means looking at a smaller sample of data — 20 games instead of 82 — which makes it important not to let any stat go to waste. So instead of outright excluding the lead-protecting situations from our analysis (the common way of doing it), let’s include them and correct for the impact of score effects.4 It’s a small but important difference, especially when dealing with a sample size this small.

By this method, the Kings’ adjusted shot differential in the playoffs was about 52 percent to 48 percent, very similar to the Rangers’ 51-49. However, the Kings were dominant against much tougher competition; they held their opponents about 5 points below those teams’ season averages, whereas the Rangers held their opponents just a fraction of a point below theirs. Once we correct for that, we again end up estimating that the Kings will get about 52 percent of the shots over the series, or maybe as high as 53 percent. That represents a clear edge, if not an overwhelming one.

So far, so Kings. But there’s also special teams play to take into account. The Rangers drew 32 more penalties than they took in the regular season, whereas the Kings took 12 more penalties than they drew. The Rangers had a higher power play conversion percentage and a better penalty kill percentage,5 so we should expect the them to have more and better power plays than the Kings in the long run — even if the actual results of this short series will be dominated by random chance.

Finally, it’s possible that it’s all going to come down to goaltending — this is hockey, after all — and the Rangers have a clear advantage there. This was the fifth straight year that Henrik Lundqvist posted a save percentage higher than 92 percent, and his save percentage has been higher than Jonathan Quick’s in every year of Quick’s career. Obviously, over a short series either goalie can get hot and turn the tide, but goalie streaks are almost entirely unpredictable and all we can do in advance is note which goalie is more talented. In this case, it’s clearly Lundqvist. The question is just how big of an advantage he gives the Rangers.

In other words, Lundqvist is the fulcrum. If we expect the Kings to get 52 to 53 percent of the shots and expect Lundqvist and Quick to match their average save percentages over the last three years, that leads to a draw at even strength. Other components — special teams, shooting and perhaps fatigue — are all pretty small factors, but also seem to work in the Rangers’ favor.

Ignore the pundits — this thing’s closer to a toss-up than a blowout.

Footnotes

  1. In this piece, “shots” will be taken to include both shots on goal and shots that miss the net, the measure proposed by Matt Fenwick. ^
  2. Focusing on situations where the score is close was first popularized by hockey stat pioneer Tore Purdy, more commonly known as JLikens. Purdy recently died at the age of only 28, a tragic loss. He wrote the piece about estimating team shooting talent that I linked above. He’ll be missed. ^
  3. The Kings’ average opponent got 49.91 percent of the shots, just a little bit higher than the Rangers’ average opponent (49.77 percent). That simple 0.14 point difference probably underestimates the competition — just as the Kings’ shot differential underrates them by not factoring in the strength of the opponents they faced, this metric also underrates their opponents slightly for the same reason. Since there are lots of things we can’t account for (specific matchups, who’s currently nursing an injury, etc.), our projected matchup can never be accurate to three decimal places. I’ve been rounding these figures off in most places, which means that we don’t need to plow through the arithmetic of exactly how much of an effect it has; the Kings’ likely share of shots against the Rangers will round to 52 percent in the end. ^
  4. To do that, I used a methodology I developed a couple of years ago. ^
  5. They also had better shot rates in both situations, which is an important component of predicting future performance. ^

Filed under , , , , ,

Comments Add Comment

Never miss the best of FiveThirtyEight.

Subscribe to the FiveThirtyEight Newsletter
×

Sign up for our newsletters to keep up with our favorite articles, charts and regressions. We have two on offer: The first is a curated digest of the best of what’s run on FiveThirtyEight that week. The other is Ctrl + , our weekly look at the best data journalism from around the web. Enter your email below, and we’ll be in touch.




By clicking subscribe, you agree to the FanBridge Privacy Policy

Powered by WordPress.com VIP