Nobody said Brendan Smith was better than Roman Josi? It's obviously a sample size thing and results for a specific 4-7 game span is not reflective of overall ability nor even predictive of future results, it just shows performance for that series for that specific skillset and nothing more. Also for the record Josi did have a poor series defensively vs Colorado, doesn't take away from the year he had overall and his teammates owed him a bailout for carrying them into playoffs to begin with, but lets not pretend he didn't get caved in defensively by the likes of Makar and MacKinnon.
It doesn't show anything tbh. JFresh came out hot to the masses who don't understand analytics as a whole but love to pick low hanging fruit. They are truly the TMZ/Enquirer of stats, and I would not be suprised if Eklund was the head editor.
As fast and popular as they came in, they went out credibility wise when during FA and expansion draft they were pumping out some ridiculous stats trying to quantify some f'ed up reality of player values with asinine comparisons. Embarrassing really
You lose mega points simply by trying to cite them as a barometer in talks. They target and appeal to ppl like you who want soooo bad to bend a narrative
Hence, Ilya Lubushkin was better at defense than Roman Josi. Get real ffs
Plus/minus was always a misused stat, or I remember when giveaway stats became very popular. I admit I got sucked into that honey trap for a minute also till you realize all the elite players have the most giveaways. Context matters. JFresh make their $ citing outta context
Context, even in those horrible games Keith had, zone entries were never a problem. His giveaways on the exits killed us. But he played well in all but 2 games, so those zone entries you mention never even remotely cost the team any strife