last edited by Smmenen
These debates are messy because people are talking past each other. Even worse, some people are saying one thing when they mean another.
To disaggregate a bit:
Diversity - is the format diverse? Are there a sufficiently large enough number of viable decks?
Balance - is the format balanced? This is measured by overall win rates: Is there a deck with a much higher win percentage than others? Or is the format actually competitive in terms of win rate.
Polarization - Are the matchups close or polarized? This is measured by match-to-match win rates. You can have both diversity and balance, but still have a high degree of matchup polarization (The extrene version of this is Rock, Paper, Scissors, with 0% or 100% win rates, depending on the match).
Interactivity - do the matchups "feel" skill intensive? In other words, do decisions players make matter?
In some respects, all four questions can be assessed empirically. The first three are largely quantitative questions. Although the fourth is more qualitative, there are empirical tools that can be used, such as players surveys, likert scales, and also assessing whether skill correlates with outcomes.
This data set can help answer: http://www.themanadrain.com/topic/2419/mtgo-vintage-challenge-compiled-results
That database has both 1) overall match win rates, and 2) match to match win rates, so we can use that to assess the answers to each of the first three questions directly.
But we should be clear on what we mean. When @brianpk80 says it's "We know it's "balanced.", actually that might not be true. It's diverse (which is what I think he means), but that doesn't mean the format is balanced. Looking at the win rates, I see Shops win rates is creeping up in the last month. Or, even if it is balanced, that doesn't mean the matchups aren't polarized.
My concern is that matchups are increasingly polarizing, and that polarity is correllated with a feeling of non-interactivity. I fear that Lavinia may be polarizing match-to-match win rates, even if that isn't fully reflected in overall win rates.
Interesting to note, though, that in @k0dy's dataset that PO and Shops have 50% win rates against each other, with something like 700 matches in the dataset. That's as close a matchup as you can get!