@chubbyrain I really appreciate your effort to define a term here. Balanced Metagame = no deck maintains more than a 50% win rate over a sustained period.
Now let me say, I think that's not a good definition. It's way, way too low a threshold of "brokenness" (Here I'm using brokenness as just, the opposite of balance.) Just apply it to other games to see why -
In the past 10 years the Boston Red Sox have won more than 53% of its games. In that time it's logged 7 winning seasons and 3 losing seasons. Can we extend your definition, and look at that data and necessarily say that the AL east is "imbalanced"? (Taking that to mean, unfair or something like that?)
In 2016 Magnus Carlsen entered and won the Tata Steel (winning every game), Norway Chess Championship (4 wins, one draw), Bilbao Masters (losing only 1 game), Chess.com's Grandmaster Blitz Battle Championship, and the World Chess Title. His wins to losses during the year easily exceeded a 3:1 ratio. Can we extend your definition, and look at that data and necessarily say that grandmaster level chess is "imbalanced"? (Taking that to mean, unfair or something like that?)
Between 1940 and 1965, Sugar Ray Robinson won 173 or 200 boxing matches, 108 of those by knockout. Can we extend your definition, and look at that data and necessarily say that light/middleweight boxing was "imbalanced"? (I mean, I think it was rigged... but not because of these numbers. I also don't think these numbers are a function of the rigging.)
By your rule, we are trying to make a statistical argument infering "unfairness" from win loss outputs. (essentially unfairness is what we mean when we say a metagame is imbalanced, I think.) That's always going to be problematic, because as we can see, sometimes some entities within a metagame are just better, even though the game itself is still essentially fair. These kind of necessarily X given the statistics are always going to be trouble...
And even then, 50% is way to low a threshold. A good standard here has got to make some effort to tease out a difference between actual systemic "imbalance" and just a case of one entity being honestly better.
Of course, a way better method to do this is to simply playtest cards and see if one meets some other standard of creating imbalance... ironically, we (the largest possible scope of the word we here) are basically bending over backward not to just playtest at this point.