Well, when one creates a binary tree, you don't do it about how the data feels, more about a way to split the data.
The amount of US teams VS teams outside the US is a crappy way to look at it, yes. But the amount of US teams is very close to the amount of non-us teams. This creates two "bins" that are roughly equal.
If we did it by continent, Canada would be lost in a list of almost 100% US teams.
If you have another method to create equal sized "bins" for the teams, I think it would be a good move. Otherwise I don't see the point.
Well, when one creates a binary tree, you don't do it about how the data feels, more about a way to split the data.
The amount of US teams VS teams outside the US is a crappy way to look at it, yes. But the amount of US teams is very close to the amount of non-us teams. This creates two "bins" that are roughly equal.
If we did it by continent, Canada would be lost in a list of almost 100% US teams.
If you have another method to create equal sized "bins" for the teams, I think it would be a good move. Otherwise I don't see the point.