Hence my point about the model flaw (or the model limits)
Yes, you're supposed to "modify" the 1% with whatever else info comes afterwards. But 1% as a raw number for Kaiden Guhle's offensive "results" (so to speak) absolutely does not align with what he is as a player.
I wouldn't say it's too much of a flaw when they have the context of QoC included. It only makes a major difference to those who are unable to understand what's going on and look beyond the overall war%.
Guhle's card imo is pretty indicative of what he is. Good defensively, hasn't quite figured it out on the PK yet (although MTLs PK is atrocious across the board), and currently isn't producing offense at all when on the ice.
However, this comes with the context that he's 22 playing a ridiculously tough role, so overall he's quite impressive.
The same thing happens with shooting%s.
Obviously if you shoot at 27%, you are providing a lot of value in doing that so your war will be very high for that year (as WAR is a results based stat).
However, if you can read the context and see a guys finishing%ile jumped up to 100% in 1 year specifically, you'd be able to see that it's likely not sustainable.
Important to understand what goes into the stats to paint the picture rather than just treating the end result like gospel.
Each of the stats is very good at highlighting what it is trying to say (especially the microstats which dig deeper into much smaller areas like puck battles, entries, exits, passing, puck retrievals, etc)