- Now What – Updated
- Oilers Gameday – Vs. Blackhawks
- Oilers Gameday – Vs. Devils
- Oilers Acquire Lain From Canucks
- Oilers Gameday – Vs. Canucks
- Kage Kool-Aid – POSITIVELY Delicious
- Diana in the Autumn Wind: End of an Era, this Season’s Shot/Chance Rates
- The Oilers Must Do Something, Anything
- Yes, The Oilers Should Claim Josh Harding
- Oilers Gameday – Vs. Coyotes
Two: Where to Find NHL Stats, A Novice’s Guide
- Updated: August 18, 2014
In my inaugural post here at the Rig, I wrote about the democratization of the hockey media landscape. Since then, I’ve been meaning to do a follow up post that rides a little closer to practice than to theory.
(Immanuel Kant, “On the Common Saying…” )
A large part of what often makes democratization possible is the structural changes experienced in everyday life when new tools become available, or old tools are re-purposed. It is not uncommon for these changes to be accidental in nature, i.e., some tool designed to enable X, goes on in the course of its use to also enable Y.
The democratization of the hockey media landscape has largely been enabled by the creation of public knowledge tools, primarily websites aggregating raw information and the social media platforms that allow for the contextualization and dissemination of this information.
A recurrent theme found in the wake of this democratization has been a pair of refrains: “what are these new concepts (such as Corsi) and what do they tell us?” and “how do I find this information for myself?” [I’ll leave aside the reactionary themes]. It is my belief that helping people with these questions is vital. As much as possible, those interested in engaging the modern conversation about hockey should be given the tools to do so.
As a community, we’ve done a very good job providing primers on the various new stats. See this popular overview from Sean McIndoe of Grantland (and follow the links therein for more). In this article, I’m less concerned with explaining new concepts as I am with giving people a tour of the online world of information. This article is a primer on the publicly available data merchants and some pointers on how to use them effectively. For the moment, at any rate, the online world of hockey stats is devastatingly diffuse with bits of information scattered to the four winds. The hope, then, is that this article will serve as a resource to those unfamiliar with what’s available publicly. The current in-limbo status of extraskater.com, perhaps makes this article even more relevant.
Despite the growth of various #fancystats websites and tools, nhl.com remains an invaluable source of information. In fact, all the data used to compile shot metrics, zone starts, etc. is scraped from the NHL’s own “event summary” game pages (here’s an example from the Oilers’ last game of the 13-14 season). My assumption is that many, especially those unfamiliar with #fancystats, under-utilize what the NHL makes available.
For an individual game, the first thing you want is to go to the “schedule” listings for any given NHL team, toggle for the season you are after and then hit the “recap” button for the game you are interested in. Now, once you get to an individual game recap, select “boxscore” and you can find all the good stuff is collected to the right in a series of links titled “Official Game Reports.”
(All the photos can be enlarged in a new tab)
There are a variety of uses for these reports. For myself, I find I spend the most time looking at the time on ice/shift reports. Often, I’ll look at them to confirm something I’ve noticed in game. For example, if I believe the coach or a player has started to shorten or lengthen their shifts, I can confirm my suspicions in real time, or after the fact, here. Or, say I believe a player has been benched for a segment of the game. With these reports, I can confirm if so-and-so was/wasn’t given his regular shifts for any given period.
Another great feature of the boxscore page is the video links for all goals scored. If you hit the “play” icon next to a goal a video pops up. Each video has an “share” icon that allows you to post or embed the link on various social media platforms.
Aside, from simply reliving the glory or pain of a goal, this is a great tool for analyzing hockey plays. Don’t trust your memory when it comes to assigning praise or blame. Re-watch the video. Get yourself some proof that Gagner was lollygagging once again.
Most, I presume, are familiar with the NHL’s general stats page. Here you can find everything from straight boxcars, to esoteric bits of trivia (who’s the tallest NHLer? Chara), to more complicated matters. There are a couple of things of interest to the #fancystat inclined here that I’d like to draw attention to.
For general skaters, one of the better tools is the “faceoff” page.
The reason I return to this page is to resolve recurring questions about who is and who isn’t being used as a center. One of the problems with player data collection is that things like vital stats (height, weight) and position (esp. wing or center for forwards and side-alignment for defensemen) can be out-of-date. Rather than trust your memory, or the NHL’s official listing on forward position (which is often wrong), you can use this rule of thumb: a regular forward playing c. 82 games is going to take well north of 700 faceoff attempts in an season if they are being used as a center. If they are taking less, it doesn’t mean that they aren’t what’s often called “a natural center.” It does mean, however, that they aren’t currently being used as one.
Unfortunately, the NHL’s website forces you to go through pages and pages of names. An alternative source for the same information is sportingcharts.com, which offers all the same faceoff data on one single page searchable by year.
Another useful page is the “real time” page.
This page lists all manner of things from hits to shooting %. I typically visit this page to refer people to the lack of context when deploying “give away” stats. As you can see, the league leaders in give aways is basically a who’s who of the best players in the league. The simple explanation for this is that those who give the puck away a lot, have the puck a lot and that, on balance, is a good thing.
Another good page is the “special teams” page. This page let’s you filter boxcars by discipline (even strength, power play, penalty kill). I typically use this page to get a quick look at how dependent a player’s boxcar stats are on the man advantage.
I also use the “time on ice” page to get a quick idea of how coaches are deploying their team. For example, if I want to know which defender Dallas Eakins entrusted with the most even strength time (a good rule of thumb for evaluating coaching decisions is that whomever gets the most even strength ice time is who the coach trusts and relies upon the most), I can toggle for the Edmonton Oilers, select defence and select “ES TOI/G” (even strength time on ice per game):
For goalies, I’m primarily interested in the “save % leaders” page and the “Special Teams.” The latter breaks down save percentage by on-ice situation (even strength, power play, penalty kill). Because save percentage can fluctuate year-over-year on the power play due to the relatively small sample size, I prefer to look at a goalie’s even strength save percentage. It has greater predictive power.
There’s a lot more one can get from the NHL’s own website, but these are the most common things I take from it.
Hockey reference is a great site packed with a ton of information for everyone from old-timey fans interested in some bit of trivia to #fancystats kids looking for possession stats now that extraskater is gone. Here’s some of the things you can do with Hockey Reference.
Here’s Taylor Hall’s page. There’s a few things I’d draw your attention to. Under the “NHL Standard” heading, I find the “Splits” and “Additional Stats” pages the most helpful.
While, the “splits” page offers a variety of information, I use it primarily to get an idea of how consistent a player’s general scoring has been through a season.
I use the “additional stats” page in the same way I’d use any of the other #fancystat websites (albeit a pared down version). Here I can screen for on-ice situations and get an overview of a player’s real time stats and possession numbers through a season.
I use the “NHL Miscellaneous” section primarily for quick references to goals, assists and points scored per game (points per game is a simple way to equalize performance across multiple players who haven’t played the same number of games in a season; curiously the NHL doesn’t offer this valuable stats).
And, finally, I use the “Other Standard” page to get a quick sense of a player’s tenure in various leagues. One of the problems with the NHL (and another popular site, hockeydb) is that they don’t list the “career” stats of a given player in a given league. Thankfully, hockey reference does it for us. This is useful especially for players who (unlike Hall) have spent considerable time in the AHL, or various European leagues.
For goalies, hockey reference has created a couple of interesting stats: “Goals Allowed %” and “Goals Saved Above Average.” These stats are relative to league averages for the season in question (see their glossary for more information). They give a quick sense of a goalie’s performance relative to the contemporaries. Here’s Tuukka Rask’s page:
League Wide Averages
Another useful tool hockey reference offers is a page with a variety of league averages for given years. Here you can get a sense of whether one season saw more scoring, power play opportunities, stronger goaltending, etc. over another season.
The best thing about hockey reference, however, is the interactive nature of its database. While it takes some time to master, this function is an invaluable resource.
Let’s say I want to find out who has the best combined save percentage in the post-lockout era. First, I select the “play index” button, than I select the “Player Season Finder” option.
It’s at this point that things get complicated and I imagine many shut the computer off and walk away. Don’t. It’s really just a matter of selecting the variables you want to learn about.
In this case, I’ve selected “combined seasons,” set the years I want to look at (post lockout), selected “goalies,” set a “games played” limit to ensure I’ve filtered infrequent players from the results and ordered the data by “save percentage.” The results look like this:
There’s a lot more one can do with Hockey Reference, but these are typically the things I use it for.
[Next Page… the #fancystats sites to visit when you are in extraskater withdrawal and how to use them]