I feel like a lot of you are kind of taking the stance "well, I know what I'm doing now, how can I adjust the upper reqs so that I don't have to do any extra work?" which is probably not going to lead to much. The upper reqs exist partially so people don't have to do the work of writing paras, so why would it reward you if you don't want to do the work of lowering deviation, either?
I feel pretty safe saying that the odds of the upper reqs going up to 50 are pretty much zero. The deviation is already pretty generous - it's significantly lower than the normal requirements because it's adjusting the margin of error, but it's a lot lower than numbers like 35 or 40(which I think is the right number) we were throwing around, too. It serves as somewhat of a built in check against strength of schedule, and against streaks - 50 deviation is ultra low (to the point its really easy to do in an afternoon on a fresh account). 50 doesn't force you to play "a lot" of games, which makes the odds of not playing a lot of quality opponents higher. The whole "I played 80000000000000000000 games and my deviation didn't go down at all" thing is not how the system works so I'm not sure why people argue it - we should all know how the system functions at this point. Yes, if everyone you're playing against is terrible it's not going to go down much when you win because it didn't tell the system anything - you're "supposed to win" that match. Beating players ranked above you and losing to players ranked below you lower the deviation "more", but in general it does, in fact, go down as you play.
You're supposed to have to play a lot of games. The point is to reward players who played consistently throughout the period, and played consistently well. It's intended to make people play a lot of games, ideally to play a lot of games consistently throughout the period. That's why it qualifies you to vote basically on its own - it requires a great deal more games played, hopefully games played against quality opponents, and ensures that you have enough experience against the period's metagame fluctuations and different team types to "know what you're talking about." That's why it is an acceptable substitute for paras. If you can do it on a fresh account the last day it isn't doing its job.
I've actually played a lot, getting at least 5 alts up to the high 1600s rating wise, but then I get haxed or quit using it in fear of getting haxed, so yeah...I actually like the implementation of this new requirement because it actually gives me some insentive to want to ladder, so I don't have to write those blasted paragraphs.
You hit two things I want to point out here:
1. Yes, this does punish people who have alt-itis, which is a good thing in itself, since alts mess with the system somewhat (ie, if I started a fresh alt losing to me would hurt a lot more than losing to me if I'm playing with an account with 1800 rating or whatever, and beating me would similarly not be worth as much to you while not hurting me much). It has a good value too, in that it encourages people to play more of their matches on accounts with ratings that are built up a bit, so they're more likely to run into other players with accounts that are built up a bit.
2. That incentive to ladder is the main reason. The ladder feels more competitive this period than in previous ones(though still not as much as I'd like, I never seem to match up with leaderbaorders much outside of Thund and Flare... I need to start using an alt so people can't queue around me as easily). It's better testing if the average level of play is higher, and that is definitely the case when good players can't just sit on their easy to acquire lower reqs.
