Yeah! I decided to maybe go a less traditional route with answering the rankings going purely off the data I have instead of my personal opinions. Will be controversial im sure, but I think it might also highlight some cards that fly under the radar and fill the categories of being an awesome cube card based on the criteria you outlined.
Using the criteria of:
- How critical this card is to winning. (I used my data of how much decks with this card win)
- How frequently you would maindeck this card (I used my data of maindeck percentages)
- How important is this card within your cube. (Here I reserve some consideration for cards that might be narrow affecting its maindeck percentage but help support archetypes and the overall cube environment)
I weighted maindeck percentage and win percentage evenly.
As far as paladin and vanguard being ranked higher, here is how my data resulted in the ranking:
Accorder Paladin:
Main %: 81.82%
Win %: 57.89%
Overall: 69.86%
Adanto Vanguard:
Main %: 72.73%
Win %: 63.83%
Overall: 68.28%
Balance:
Main %: 71.43%
Win %: 58.00%
Overall: 64.71%
Just a little bit of a different take. I was surprised at some of the results and it was hard to not just change my rankings based on opinion vs actual cube performance in my list. Never realized how much cast out gets played and wins.
I like objective, hard data.
But the reality is, even with a pretty large sample size there are going to be lots of flukes and oddities - it's simply unavoidable.
I think everybody here should take the realistic view that no matter what approach is settled on, the results may be interesting and informative but they will have their flaws and imperfections.
I view Accorder Paladin being ranked over Balance as one such fluke. Not only is that result in strong disagreement with my prior (subjective, personal) belief, but it's also in strong contradiction to other metrics with reasonably large sample size.
Once again though - nothing against your approach - more power to you.
I like objective, hard data.
But the reality is, even with a pretty large sample size there are going to be lots of flukes and oddities - it's simply unavoidable.
I think everybody here should take the realistic view that no matter what approach is settled on, the results may be interesting and informative but they will have their flaws and imperfections.
I view Accorder Paladin being ranked over Balance as one such fluke. Not only is that result in strong disagreement with my prior (subjective, personal) belief, but it's also in strong contradiction to other metrics with reasonably large sample size.
Once again though - nothing against your approach - more power to you.
360 card powered Chicago cube:
https://cubecobra.com/cube/overview/e7r
2020 Numerical Power Rankings:
https://www.mtgsalvation.com/forums/the-game/the-cube-forum/cube-card-and-archetype/817969-2020-numerical-cube-power-rankings
2018 CubeTutor Power Rankings:
https://www.mtgsalvation.com/forums/the-game/the-cube-forum/cube-card-and-archetype/803301-cubetutor-power-rankings-2018-by-color-and-cmc