A shot differential metric developed on Broad Street Hockey is gaining increasingly widespread use. Let's look at what it means and what it says about Flyers West.
Earlier today, I said I'd write something up on what advanced stats are commonly used and what they can tell us. Let's start by talking about team-level predictions and looking at the Score-Adjusted Fenwick standings in the context of recent history.
For most of you, the first question will be "What the heck is Score-Adjusted Fenwick?" I'm glad you asked...
In today's NHL, the differences between teams in shooting percentage and goaltending really aren't that large, and over a few dozen games a team can easily run much hotter or colder than their talent. (I'm sure you can all think of some examples.) It's possible that a team that's shooting 10 percent at 5-on-5 or stopping 94 percent of the opponent's shots has an offense or defense that is unprecedented in recent history, but it's a lot more likely that they've just been on a hot streak -- and that we shouldn't expect said hot streak to continue.
The result is that, although it may be counter-intuitive, if you want to predict a team's future winning percentage, you'll do better by looking at their current shot differential (ignoring shooting percentages) than looking at their current goal differential or their current winning percentage.
Yes, that's strange. Everyone's first instinct is to think that some teams will get higher quality shots than others, and looking at shot differential ignores that. And that's true, but it turns out that the differences in shot quality aren't very large, and so over the course of a season team shooting percentages are driven more by random streaks than by talent. Which means that time after time, the analysis comes away with shot differential being a better predictor of team success than goal-oriented analysis.
Unfortunately, it's not quite that simple. Just using shot differential does OK, but it gets confounded by the way teams adjust their strategy to suit the score. A team that's leading will go into a bit of a defensive shell, allowing the other team to outshoot them. Accounting for that makes our predictions much better.
The simplest way to do that is to narrow things down to just look at shot differential when the score is tied, so there aren't any score effects. But whenever you shrink the sample size, you make it so you need more data to make good predictions; if our goal is to be able to tell really early in the season what the final standings will look like, then we don't want to throw out large swaths of the game if we can help it.
So we take a few steps to expand the sample size.
The first step is to count more events. Instead of just looking at shots on goal, we can add missed shots. Shot differential including missed shots is called Fenwick, named after the person who first suggested it, and it is an even better predictor than simple shot differential.
The next step is to look at more than just tied situations. One way to do this is to add in situations where score effects play only a small role -- where the team is up by a goal or down by a goal in the first two periods, and hasn't changed their strategy much. My preferred method is to include all of the data but correct for score effects. We know that the average team gets 56 percent of the shots when they are down by two goals, so if a certain team has gotten 58% when down by two, we know they were doing 2 percent better than average and can just factor that in.
The result is a formula that I called Score-Adjusted Fenwick, which averages together how much better or worse than average a team did in each game state. This turned out to be a better predictor than Fenwick Tied or Fenwick Close, especially early in the season. I released this stat last year and looked at how it did for predicting based on small sample sizes -- and it sparkled in that role, very quickly flagging the Kings as the best team in the league after the Jeff Carter trade.
I've seen a few people compiling Score-Adjusted Fenwick so far this year, most recently Travis Yost of HockeyBuzz. And while I've seen people talking about the Kings as a sleeper team, I haven't seen many people emphasize just how good they appear to be. Here's how they rank against the teams with the highest Score-Adjusted Fenwick of the last five years:
|2007-08 Red Wings||58.8%||Stanley Cup|
|2009-10 Blackhawks||58.8%||Stanley Cup|
|2012 Kings w/Carter||57.5%||Stanley Cup|
|2008-09 Red Wings||56.6%||Stanley Cup Finals|
The Kings now have close to 50 games under their belt since the Carter trade and have played a dominant brand of possession hockey over that span. Don't let their middling record fool you -- Jonathan Quick won't be a .894 goaltender in the long run, and when things start to go his way, they will be awfully tough to beat.
Oh, and by the way, the Kings are still 14:1 in Vegas. Not as juicy as last year, but still a good bet.