Hockey analysis has come a long way compared to just 5 years ago. The data is better, more research is done, people are understanding how to contextualize better, and we have more now to consume than ever. If you think about it, the on-ice ideas that are emphasized by those in the analytics sphere are essentially the same as what old heads have been preaching for decades.

Getting pucks on net = increasing your shot attempts
Get to the dangerous areas = improving your shot quality, which ties into expected goals
Net-front battles = prevent the opposition from taking those high xG shots and prevent your goalie from being screened
Creating/preventing odd-man rushes = rush chances are shown to be of higher quality than chances created on the cycle or forecheck
...and so on and so forth.

I guarantee you that 90-95% of the findings from analytics spheres align with the general consensus. The problem most people have are with the edge cases, which they then use to dismiss the entirety of analytics. The on-ice arguements are essentially the same, albeit the wordings differ. Most everyone will agree that today's game is far different than the game from even a decade ago. It's a possession-based game, physicality is emphasized less, etc.

A lot of the polarization comes from how off-ice/locker room effects come into play. It's pretty clear that guys like Ryan Reaves are not valuable to the modern game, in terms of on-ice impact; he plays very limited minutes and the team does not control play when he is on the ice. The arguments for Reaves state that he provides a morale boost to his teammates, he's a locker room guy, etc. The problem with that is how can we even know that's the case when we're not around the team and we don't know what's going on behind the scenes, other than the canned statements we hear in interviews? I'm sure that most, if not all, teams have guys that are good in the room, so as long as there aren't guys that are a detriment to team morale, I don't see the value in having a guy that also plays poorly on the ice.

“I don't see the value in having a guy that also plays poorly on the ice.”

NHL GMs sure do though and that’s been my point over and over again. What a guy like Reaves does better than most of the league is put in a short shift but throw a couple huge hits that get momentum back to his team. He’s made an entire career out of it.

Like someone mentioned earlier, just because an NHL GM does something doesn't mean it's a good move, as we see time and time again. How do we even know that his hits provided any kind of momentum? He's spent most of his career playing on teams that are good in spite of him, so one could argue that the teams could've succeeded and regained momentum without whatever his contributions were in the 8-9 minutes he plays.

Like someone mentioned earlier, just because an NHL GM does something doesn't mean it's a good move, as we see time and time again. How do we even know that his hits provided any kind of momentum? He's spent most of his career playing on teams that are good in spite of him, so one could argue that the teams could've succeeded and regained momentum without whatever his contributions were in the 8-9 minutes he plays.

There isn’t a stat out there that makes Reaves look like a strong contributor to New York’s success this season. The Rangers were outscored 30-16 at even strength with Reaves on the ice, and that minus 14 goal differential was the worst on the team. The NYR forward depth was their glaring weakness for the first three quarters of the season prior to the trade deadline, and getting outscored by nearly a 2:1 ratio whenever Reaves was on the ice was a major part of that issue.

One stat that highlights Reaves’ all-encompassing deficiencies well is the Goals Above Replacement model from Evolving Hockey. GAR (and its W(ins)AR and S(tandings)P(oints)AR conversions) are cumulative stats that incorporate even strength play, special teams play, and penalty impacts. Out of 427 forwards to skate at least 350 minutes this season, Reaves was one of only 11 to be far enough in the red in GAR to account for at least -1 WAR. That comes in spite of basically zero special teams time, which other bad players in his vicinity receive and give them more opportunities to be bad.

Taking out the special teams component of GAR and WAR, only five other players were such drags at even strength to account for -1 WAR. Being the team’s best chirper and releasing them onto the ice for games are nice things to get a chuckle out of once in awhile, but when having that requires icing a player in the conversation for the worst skater in the league, something has to give.

When the playoffs rolled around, Reaves spent even more time watching games from the bench than he usually did. No forward skated less than the 9:25 per game Reaves did in the playoffs. After a particularly rough four games against the Lightning in the Eastern Conference Finals. Reaves spent the final two games of New York’s season in the press box.

Bad


Good


Checks out, both eye test and stats.

If you’ve played the game you know that a guy like him can change the momentum with just one hit at the right time or one fight at the right time. And I’m going to trust the judgment of every NHL GM no matter how bad the team is over people on the internet with fancy charts

that's great. gm's are now finding guys who are able to do that as well as do other things too.

giving people concussions is no longer enough to earn a roster spot.

Too much analytics is bad
Not enough analytics is bad

Eye test doesn't tell the full story
Analytics don't tell the full story

Analytics are being used more in today's hockey with better availability to track those numbers and teams can get a better understanding of players without having to watch every game.

"Why would a smart team like the Rangers keep someone like Reaves if he’s so bad according to your analytics people on random corners of the internet?"
As for Reaves, the Rangers kept Reaves because he plays a role that not many NHL players play in today's game. In reality he isn't good enough to be a regular NHL player but his intangible impact on the ice keeps him in the league

Too much analytics is bad
Not enough analytics is bad

Eye test doesn't tell the full story
Analytics don't tell the full story

Analytics are being used more in today's hockey with better availability to track those numbers and teams can get a better understanding of players without having to watch every game.

"Why would a smart team like the Rangers keep someone like Reaves if he’s so bad according to your analytics people on random corners of the internet?"
As for Reaves, the Rangers kept Reaves because he plays a role that not many NHL players play in today's game. In reality he isn't good enough to be a regular NHL player but his intangible impact on the ice keeps him in the league

So you understand why a GM would keep him around… refreshing to see.

And he had a roster spot though so apparently it is. At any rate you can focus on Reaves but look at a guy like Wilson who can score goals and battle with the best of them, he’s worth his weight in gold because players like him are so rare. What do your analytics say about Tom Wilson?