I could, would just be a death sentence for me :)

P.S. even with two layers of fingerless gloves while typing, there's still some shuddering and obvious typos.
Also forgot to tag @sierkovitz.bsky.social as it's likely be up his alley.

For now, this will do but it's exploratory. Just processing data (badly) and coding (worse) to get an impression of availablve avenues. Maybe there's a route to a tool that helps players identify the cards that shouldn't appear in a deck for a colour pair or highlight cards that should. 6/eNd

Prevalnce in trophy decks (which is just what % of decks in a colour pair combo had that card) multiplied by the average number of that card in decks, multiplied by the ratio of the overall GIH WR to that of the GIH WR for the specific colour pair. I'm sure there's a more elegant link but...5/N

This is all feeling around to see what looks interesting kind of analysis, and just a pet projecject, but I'm getting convinced there's something here (not least from the @lordsoflimited.bsky.social discord feedback)
FWIW the "AdjScore" is just a combo of... 5/N

And with the prevalnce of how often a particular card occurs in trophy decks of a given colour pair. So in BG undead Butler doesn't perform much better or worse than in any trophy deck it appears in, but BG decks tend to have more of them on average than even premium removal like Ob-Nix Cruelty 4/N

...colour pair. Such as like in the blue cards here:
Looks like Cloudfin Raptor is kinda pants outside of UW for instance. And as far as the "Adj(usted)Score is concerned, I think there's some link between GIH WR (game-in-hand win rate) for a card when it's in a given colour-pair, and overall. 3/N

I've put some output from about 600-ish pioneer master decks from the recent @magic.wizards.com set.
tinyurl.com/AJMpioTDD
Exploring at the moment but maybe some others might be interest. Main aim is distinguishing between card performance overall vs that in a trophy-winning deck of a given 2/N