Ultimately, Stu Hackel wasn't asking for very much – for the most part, he wanted the NHL to post stats that it already tracks on its website, or possibly cut a finer slice of the stats that it already tracks. We think a little bigger here – there's significantly more data that the league could track that would help us understand the contributions various players make during the game.
1. The NHL needs to standardize its recording of real-time stats.
Giveaways, takeaways, blocked shots and hits could all be useful statistics, but we sometimes see 10:1 variation from rink-to-rink in how these stats are recorded. There are other issues, as we've seen: posts hit, shot locations (particularly at MSG) and even what counts as a shot on goal or a missed shot. To my knowledge, there is no standardization, and each team does things differently.
2. Scoring Chances.
Lots of people record chances – teams, TV crews, random people on the internet. But a lot of the data is kept secret or handed out haphazardly. The random people on the internet do a good job, but this shouldn't be a volunteer gig.
3. Zone Entries and Exits
The crews over at Broad Street Hockey and Copper & Blue have done great work in this department. We don't yet know what the effort will yield, but it gives us a great quantitative feel for what players need to do to be successful.
4. Passes
We recorded passing data for one game last year, and the results were pretty instructive. The collection time is prohibitive, so it's exactly the kind of project a large organization with tons of stringers could get involved with.
5. Coordinate-Tracking of the Puck and Players
This is the holy grail, clearly. We'd like to see electronic tracking of every player and the puck at all times. Yes, it would create a mountain of data, but that's exactly the kind of thing we'd like to hammer away at with a computer. Major League Baseball has benefited greatly from Pitch-F/X, its real-time pitch tracking system, and there's no reason to think the impact would be any less in hockey. (Added benefit: if this came to pass, it's not clear that we'd need stats #1-#4.)
