The question is, between S2,4,6,8,10 and 12, is there anyone who is lv 37 to 94 who is NOT in the sage that you were in. And just how many 37 to 94 are there in total?

Fact of the matter is, there could be all of 3 people who is 30-40, another 5 for 40-50, 4 in 50-60 and 10 for 60-70 , each of them can't possibly have their individual sage, there is hardly enough people, in this example, for 30-70 to have a bracket (there is 22, which is barely above the minimal sage requirement of I think 21)

As it is, that cluster is a very old and small, in terms of # of server, server bracket, there probably aren't a whole lot of lowbies in them, so this is entirely possible. So as it is, the best matchmaking in the world won't be able to fix this "problem", because they are constricted by what they have to work with.

Just to be clear, I'm not claiming there isn't a problem. I'm saying that you can't prove one exist based on what you said. If you can find a single player who was in a different sage between lv 37 and 94, then you could prove that there is a problem. (hence the very first question I asked above)

Now, I would support more dynamic sage size, so when player are sparse the sage size shrinks to as small as possible to avoid giant gaps. However, hard bracketing won't work due to potentially below minimal requirement number of player within a particular bracket.

Furthermore, we need to think about the level freezers. They have a large enough advantage as it is, they stand a decent chance even if put into a bracket where they are the lowest level. If there were hard brackets, you can bet your hind quarters that they will camp at the highest level of a particular bracket, gaining even a larger advantage.