Shot charts and playoff teams

yea, he's good

yea, he’s good

There was a pretty interesting graphic that circulated online after the conclusion of the Boston and Ottawa game. It was a chart simply looking at shot differential standings, highlighting the teams that made the playoffs.

I didn’t have access to the raw numbers used by Frank Dumais, who put the chart together. However I was able to copy and sort of duplicate every season since 2008 from stats.hockeyanalysis.com and do a similar thing. These are only even strength statistics, but I’ve highlighted all the playoff teams:

shots diff 1

shots diff 2

shots diff 3

shots diff 4

shots diff 5

shots diff 6

The obvious question: “Why would we look at these over goal differential statistics?” and the easy response is simply that goal differentials are more random. I’ve made a couple of bucks just in side bets with friends betting whether a team would maintain its current record or not.

From Gabriel Desjardins:

Basically, whatever you think you know about your team’s supposed ability to maintain high shooting and save percentages, they are very likely to crash back to league average regardless of how many shots you’ve observed. Internalize this chart and you can make a lot of money betting against people who are convinced there’s mysticism in scoring goals.

There’s a chart in his post that I haven’t copied but even over 3000 shots, the general story is that a team’s PDO (save percentage plus shooting percentage, explained in further detail here) isn’t observed to be higher in the long run.

New Jersey did a bit better on special teams than at even strength this season, but they still had the fourth highest shot differential at evens and missed the postseason, hilariously. Some pretty awful luck out there on the rock.

It is interesting that over 48 games, neither New Jersey, Toronto or Anaheim “regressed” enough to balance out the playoff teams. The Leafs and the Ducks certainly made some noise in the opposite direction in the last three weeks of the season but ultimately there wasn’t enough of a signal to accurately forecast impending doom. We may have to wait until next season to see those team’s shooting rates balance out.

One thing that did strike me was that Vancouver is third worst in shot differential this season, although they do much better in Corsi, so they’re having the puck even if they’re not getting it to the inside for continuous shots.

Another thing that absolutely caught my eye was this, from Aaron Portzline’s post-mortem on the Lumbus Blue Jackets:

“The one thing we really emphasized (in our exit meetings) was that, as a young team, the one area we can really grown in is the summer time,” Blue Jackets GM Jarmo Kekalainen said. “We can put in the work in the summer to get stronger and bigger and faster and quicker … all those things. The growth from our team can really come from there, start from there.

“We’re going to look at everything else as a management team to see where we can improve for next season as well. Certainly nobody thinks we’ve arrived somewhere now that we were so close. Nobody thinks this is great, we almost made the playoffs. Our goal is to make the playoffs and be able to compete for the Stanley Cup.”

Lumbus certainly stole everybody’s hearts during their run down the stretch, but I like the bolded section. Kekalainen has blurbed for Hockey Prospectus’ Annual in the past and while the team has some good pieces, it was also a pretty poor shot differential team that banked entirely on Sergei Bobrovsky down the stretch. I’d hope Kekalainen realizes this and doesn’t bank too much of his future based on the outcome of a few games down the stretch.

I opined on the possibility back when we figured we’d have a 40-game season, that the results from a small sample of games could kill the way teams build for next season. I’m interested to see how the Devils, Ducks and Maple Leafs react in the off-season to results that didn’t match their performance.

Also, while you’ll notice that the three teams on the top of that chart, the 2010 Blackhawks, 2008 Red Wings and 2009 Red Wings all made the Stanley Cup Final, I don’t think there’s any proof that possession stats are better than dumb luck at this point in predicting playoff success. The next three teams, the 2012 Red Wings, 2012 Penguins and 2009 Sharks all lost in the first round of the post-season.

Comments (4)

  1. Having watched a handful of NJ games this year these numbers dont surprise me. The Devils would often control games for minutes at a time with their strong cycle and throw tonnes of pucks on the net, but they just arent a skilled enough team to bury on the great chances they were getting. They play a playoff style game all year and just werent able to grind out enough victories to make it into the playoffs, otherwise they could have gone on another run this year.

    • You’re right, the Devils totally controlled a ton of games that they lost. Not only did they really miss Parise and Sykora’s goals this year, but they just had some amazingly bad runs of luck. They probably hit the post more than they scored for about a 20 game stretch there. It was just insane.

  2. I have a couple of observations and questions:

    1) Toronto missed playoff three times with positive shot differential and then makes it with one of the worst shot differential in recent memory. If shot differential is really a good indicator, then Toronto is just bucking the trend entirely.

    2) 2009 Penguins and 2011 Bruins have the vastly inferior shot differential in comparison to the 2008 Redwings, 2010 Blackhawks, and the 2012 Kings. Does that mean there is more than one way to win in hockey? Or are Pens and Bruins just lucky. It would be interesting to see team save percentage as another column as well to see if goaltending can explain 1) and 2).

    3) Where can I download these data in excel or some database format? I would love to play with them myself.

    • 1) The 2009 and 2010 Maple Leafs had historically terrible goaltending from Toskala and Gustavsson. The 2013 Maple Leafs had the best year in Leafs goaltending ever. There’s your reason. It’s not bucking the trend at all. As Cam says, shot differential isn’t the end all be all, it’s just one big indicator and you have to look at these numbers in context. In the Leafs’ case, that context is goaltending.

      2) Again, as Cam said this isn’t the end all be all. It’s just one very very predictive indicator of team success in the regular season. This data does not include playoffs. If you’re looking for one “superstat” that explains everything, you’d better not hold your breath.

      3) The NHL tracks EV shot data. Cam got his data from NHL.com.

Leave a Reply

Your email address will not be published. Required fields are marked *