I’m sure
@VladDrag has something to say about Frost’s WAR not being accurately representative/predictive of his microstats and in-season growth.
I'm going to start by saying I'm not 100% sure how J Fresh breaks down his WAR numbers. I know it's from Patrick Bacon, but that's about the extent of it. I don't know if these are total or if these are rates or what. I know nothing about them. I assume that the WAR number presented by J Fresh are full season as well. Which we know that Frost got better as the year went on.
Also, we need to understand what WAR/GAR numbers attempt to do. For the most part, WAR/GAR attempts to provide a total value of a player. It attempts to represent how much a player has contributed to their team. In essence, it's an attempt to identify a players impact, not potential or skill level. It's up to the person developing the calculations to weight what is most important. This is why you'll see variations in different WAR/GAR values.
What I can tell you is that the data from Patrick Bacon and Evolving Hockey (EH) don't quite match. From Evolving Hockey, Frost's ES Offensive GAR ranks him in the 50th percentile and his ES Offensive xGAR ranks him in the 66th percentile. J Fresh shows Frost in the 32nd percentile for ES Offense. EH also presents Frost as an above average defensive player as well (GAR - 50 percentile, xGAR - 57 percentile). So for some reason, Patrick Bacon's model shows Frost was less impactful than EH's shows. I don't know why, and I don't know which one is more correct.
However, what I can say is that Hockey viz presents Frost as a player who had an increase of xGF relative to the rest of his teammates and decreased xGA relative to the rest of his teammates(3% increase in xGF and a 3% decrease in xGA). Basically, this means that he was a bit better offensively and defensively than the rest of the team.
New to this year, Hockey Viz also has a new metric called synthetic goals (sG), and it's kind of like WAR/GAR stats, but also not quite. It attempts to do the same thing, but it does it in different ways. Hockey Viz does not give you values, but rather attempts to rank players by line (1st, 2nd, 3rd etc). Frost's sG from last season placed him at a very high level 3rd liner/low level 2nd liner. This makes sense given his poor first third, decent middle third and good final third.
Basically what I would say is that the numbers above from all three places are fairly similar. It shows Frost's 22-23 season as a middle-6/bottom-6 type player. That would make sense given his season's ark.
I could keep going, but it's not really worth it.