Saturday, November 14, 2009

Even Strength Data for Games 1-250

For those that have been following the scoring chance postings, I realize that I've fallen behind a tad and that the last five games have yet to be posted. I apologize for that. I'm going to make an effort to catch up this week.

In the meantime, I figured that I'd post some data showing how each team has performed at even strength thus far. I've also included information concerning each team's performance at EV when the score is tied. The teams are ranked according to their EV Corsi ratio with the score tied.

I should note that the data is based only on the first 250 games played in 09-10, and is therefore not entirely up to date. If anyone is experiencing any difficulty in viewing the document, please let me know so that I can make the necessary adjustments.

Some points of note:
  • At this point in time, the Blackhawks are far and away the best team in the league in terms of outshooting the opposition at EV with the score tied
  • To the surprise of many, the Coyotes are the second best team in this regard
  • The Penguins are surprisingly mediocre according to this metric; I'm not sure how much of this can be attributed to their recent injury woes
  • It's unlikely that the Avalanche are as bad of an EV team as their Corsi ratio with the score tied would suggest; their EV scoring chance ratio, while still below average, isn't nearly as bad
  • The data suggest that the Leafs are better than their record would indicate; likewise for Carolina
  • As lucky as the Bruins were last season, the reverse has been true so far in 09-10; I suspect that they'll turn it around sooner or later


Olivier said...

Having watched the leafs a few times, I wonder if (was it SunnyMetha on IOF who got that data out this summer? Or you?) there isn't something along the lines of an over-reliance on defensemen to shoot at the net. Which means they do have time of possession in the O-Zone, but mostly burn it lobbing pucks from the blue trying to finish on a rebound or tip-in.

I guess Kessel changes things tough...

I'm kinda throwing this out there tough.

How do you collect those data btw? Makes me realize I have start poking again at that ugly python script I once hoped would allow me to bathe in an ocean of csv delimited NHL data. Hope makes you waste a ton of time, sometimes ;).

Vic Ferrari said...

Odd that Colorado's Corsi ratio while tied is worse than the overall.

I expect that over the next tean games COL will see a small bump in scoring chance % (They've been around 47.4% in the first 10 and the same in the most recent 9 iirc.) I'll guess 48.5%, assuming no serious injuries to key guys.

Overall corsi% will be about that too, I'll guess.

And as I'm specualting ... they'll have a losing record. Maybe 8 points in the next 10 games.

JLikens said...


That might have some relevance in terms of their overall EV SH%, given that they've (presumably) been playing from behind more often than not thus far.

In terms of their score-tied EV SH%, though, I think that it's mostly, or perhaps even entirely, simply bad luck.

Again, shot quality effects at EV when the score is tied are non-existent or virtually non-existent.

Also, the Leafs score-tied EV SH% last season was above the league average, and the team hasn't changed that much in terms of style or roster composition.


Yeah, I also expect them to experience an improvement in their underlying numbers over the next while, partly because their schedule has been fairly road-heavy up to this point.

I also agree that they're bound to see their record deteriorate sooner or later.

JLikens said...


I collected the data through excel, by using the 'web query' function.

I used the function to scrape the play by play data from and import it into excel.

I then designed a formula in excel in order to obtain the desired information.

Because I have very little programming experience, the whole process involved a lot of trial and error.

If you have any questions, I'd be glad to help (although, again, I'm far from an expert in the area).

Scott Reynolds said...

Thanks for doing all this. Much appreciated.

The Avs are nearly sure to regress significantly in terms of their record both by shots data and by scoring chances. They just aren't that good. A fantastic hot streak for them though. Assuming they come back some it will be interesting to see how they act at the trade deadline. I expect they'll be on the bubble and then they'll need to make choices on guys like Tucker, Svatos and Clark. Do they sell them for futures or hold on to them and give up some futures to make a run. I'm interested to see how they handle it.

JLikens said...

It'll be interesting to see where Colorado ends up in the standings.

As you say, the smart money is on them having below average results between now and the end of the season. The real question is whether or not their hot start will be enough to get them in the playoffs.

Kenny Smith said...

This is my first time seeing the column, and thank you for the numerical analysis. It certainly proves what I thought about my Hurricanes, that we have simply not gotten the bounces this season. We've only looked out matched a few times this season, Staal was hurt from day one, shutting him down was the right call. Of course, I almost want them to throw the white flag and draft in the top 2.

Could someone explain what the CORSI stat is? Thanks

Love Hockey!

Go Canes!

JLikens said...


Corsi is defined as:

[Even strengh shots on goal for + Even strength missed shots for + Even strength blocked shots for] -
[Even strengh shots on goal against + Even strength missed shots against + Even strength blocked shots against]

It's basically how many shots a team directs at the other team's goal at even strength, minus how many shots are directed at it own goal at even strength. A measure of territorial dominance, in other words.

As for Carolina, you're absolutely correct that they've been more unlucky than anything this year. Once the bounces start evening out, they should be alright.