Great work as always. Your comment about the eye test is fantastic, and really illustrates why people seem to think this is the only way to understand hockey: it gives you all the context you need - or rather, the context you think you need.
Also, if you're looking for quality YouTube analysis, I don't think anyone does it better than Along The Ice Hockey (if you haven't already heard of him).
Agree it's totally worthless for goaltenders, there's nothing in these "advanced" stats that hasn't been available for almost twenty years now. For skaters, I think it's more trivia than useful, and that's after spending a lot of time trying to dig into what's worthwhile. On occasion, I've found it useful for explaining things not captured in regular advanced stats (Jason Robertson somehow even slower at the start of last season while recovering from offseason surgery), it mostly doesn't add anything though.
I think the biggest problem is the lack of transparency. They show only season totals, they don't break things down by game state, except for zone time (also, puck zone time would be more interesting than player zone time) and distance skated, the two things with arguably the least utility, and the biggest thing is there's no table view except for the top 10 for each category. You have to go to each player's individual page to see what their numbers are. Something this clunky has to be intentionally designed to be difficult to use.
The interface is slightly better this season, and on the backend they've added API endpoints, so it's a bit easier to scrape the data, but still, it's very clearly not Baseball Savant.
The decision to not specify exactly how bad any player/team is when they're in the bottom half of the league remains symbolic of many NHL EDGE problems.
As you note, they present it as though it's hockey's answer to Baseball Savant. But Baseball Savant, thanks in large part to Tom Tango, has an ethos of data-sharing and facilitating collaborative research, because in the big picture that's good for the league and sport: You get amateurs doing free research that is either valuable on its own or gives teams candidates to hire for analyst roles in their front offices.
But NHL EDGE has the opposite of transparency. Don't quote me on this, but I think I read somewhere that it was because, to get the NHLPA's authorization to publish this data, they agreed to not show a full ranking or who was the worst in any given category. Obviously this is pathetic, to hide low performers in the data to shield people's feelings or keep them from getting flak from fans and media. But it also prevents us from doing almost anything useful with the data! Imagine a Baseball Savant where we couldn't know who the bottom 50% performers were in any category. It would be useless.
It’s actually so funny you mention not showing who’s worst in a given category because I tried to see today who has the lowest max skating speed, and I found out that I could not invert the best-to-worst sorting.
That also would explain why there’s no leaderboard function in any statistic.
Not that you’ve pointed this out, I can’t unsee it.
That would make sense re: NHLPA approval. That's what I suspected the reason was, but it was just a guess. This is the league that didn't like having a player picked last at the All-Star draft!
Great work as always. Your comment about the eye test is fantastic, and really illustrates why people seem to think this is the only way to understand hockey: it gives you all the context you need - or rather, the context you think you need.
Also, if you're looking for quality YouTube analysis, I don't think anyone does it better than Along The Ice Hockey (if you haven't already heard of him).
https://www.youtube.com/@AlongTheIceHockey
Agree it's totally worthless for goaltenders, there's nothing in these "advanced" stats that hasn't been available for almost twenty years now. For skaters, I think it's more trivia than useful, and that's after spending a lot of time trying to dig into what's worthwhile. On occasion, I've found it useful for explaining things not captured in regular advanced stats (Jason Robertson somehow even slower at the start of last season while recovering from offseason surgery), it mostly doesn't add anything though.
I think the biggest problem is the lack of transparency. They show only season totals, they don't break things down by game state, except for zone time (also, puck zone time would be more interesting than player zone time) and distance skated, the two things with arguably the least utility, and the biggest thing is there's no table view except for the top 10 for each category. You have to go to each player's individual page to see what their numbers are. Something this clunky has to be intentionally designed to be difficult to use.
The interface is slightly better this season, and on the backend they've added API endpoints, so it's a bit easier to scrape the data, but still, it's very clearly not Baseball Savant.
The decision to not specify exactly how bad any player/team is when they're in the bottom half of the league remains symbolic of many NHL EDGE problems.
As you note, they present it as though it's hockey's answer to Baseball Savant. But Baseball Savant, thanks in large part to Tom Tango, has an ethos of data-sharing and facilitating collaborative research, because in the big picture that's good for the league and sport: You get amateurs doing free research that is either valuable on its own or gives teams candidates to hire for analyst roles in their front offices.
But NHL EDGE has the opposite of transparency. Don't quote me on this, but I think I read somewhere that it was because, to get the NHLPA's authorization to publish this data, they agreed to not show a full ranking or who was the worst in any given category. Obviously this is pathetic, to hide low performers in the data to shield people's feelings or keep them from getting flak from fans and media. But it also prevents us from doing almost anything useful with the data! Imagine a Baseball Savant where we couldn't know who the bottom 50% performers were in any category. It would be useless.
It’s actually so funny you mention not showing who’s worst in a given category because I tried to see today who has the lowest max skating speed, and I found out that I could not invert the best-to-worst sorting.
That also would explain why there’s no leaderboard function in any statistic.
Not that you’ve pointed this out, I can’t unsee it.
That would make sense re: NHLPA approval. That's what I suspected the reason was, but it was just a guess. This is the league that didn't like having a player picked last at the All-Star draft!