I wrote about lacrosse stats for IL Indoor for almost ten years. Now that I’m not doing that anymore, someone creates a web site with a zillion NLL stats. Figures.
laxmetrics.com is a new site created by San Diego PxP guy Cooper Perkins, and it contains an insane amount of data, way more than I have had access to over the last ten years. I’ve generally been dealing with directly-measurable stats (goals, loose balls, penalties, etc.), and then doing math to combine them, aggregate them, average them, and so on. Some of the data available here is the same – taking the data we get from game sheets and such and “manipulating” it to try and get something meaningful. For example, the “plus” stats (goals+, assists+, and so on) basically compare a player’s production against the league average, and gWAR (goalie wins above replacement) uses goals for and against to attempt to “quantify how many wins a goaltender is directly responsible for creating”.
However most of the stats require more work and you just can’t get them from the boxscore. For example, there are several types of assists listed here:
- “First order assists” are different from regular assists in that the intention of the passer is taken into account. For example, if a transition player casually tosses the ball to a forward before heading off the floor and the forward scores (with no intervening passes), that transition player gets an assist which is arguably not as “deserved” as other assists. First order assists only counts passes that are directly intended to lead to a shot.
- A “second order assist” is the equivalent of a first order assist but for second assists. Sometimes second assists are meaningful and necessary for the goal, while others are not.
- An “unrealized assist” is a pass that results in a scoring opportunity but no goal is actually scored. We’ve all seen outstanding passes that result in a shot that misses the net or that the goalie saves, and of course no assist is credited.
- A “pick assist” occurs when a player without the ball sets a pick or does something else off-ball that directly contributes to a goal. Because the player never touched the ball, he won’t be given an assist.
Of course, teammates and coaches notice these kinds of plays and sometimes broadcasters will mention them as well, but normally they get no other credit. Now they do.
Dhane Smith, league leader in Facilitator Score and Weighted Assists
The problem with those sorts of stats is that the league doesn’t keep track of them, so someone (Cooper, presumably) has to sit and watch every second of every game, looking for these things and recording them. He has to hope the feed stays up, the cameraman catches everything, players names or numbers are visible so you can tell who did what, and so on. NLL games are generally around 2h15m long, and there’s probably a lot of going back and forth, watching a single play a dozen times to make sure you got everything. You can skip timeouts and commercials and such, but I imagine it still takes several hours per game to gather all of this information. (Update: I heard Cooper on the Off the Crossebar podcast the other day and he says it takes him about 35-40 minutes per game, so perhaps this isn’t the time commitment I thought it would be but it’s still significant work.)
In addition, most of these stats are very subjective. Was that pass really essential to the goal? There was a great pass followed by a shot from a bad angle that didn’t go in – was that enough of a quality scoring chance to warrant an unrealized assist? But even loose balls and face-off wins can be somewhat subjective, and we rely on someone else to make those decisions, so this is really no different.
Honestly, I don’t love how the data is presented on the site. Most pages look like an Excel spreadsheet embedded in the middle of a blog post. Given that the site is created with WordPress, that’s probably exactly what it is. In some cases, this is just as good as showing an HTML table. But for example, the leaderboard page makes you scroll left-to-right to see the data. Given the amount of unused space on each side of the chart, this is ugly. But who cares, really, it’s the data and the interpretation of the data that really matters. The page has only been up for a week or two so perhaps “make it pretty” is still on the TODO list. I’ve had this blog for ten years and have put pretty close to (read: exactly) zero time into making it pretty, so I really shouldn’t complain.
I appreciate the amount of work all of this is, which is why I don’t do it. But the fact that someone is doing it and publishing the results of the analysis is crazy awesome.