I've been wanting to post something in this thread for quite some time now, but I've unfortunately been too busy to collect my thoughts into a quality post. Most of my recent 'Smogon time' has been spent working on my article for The Smog. However, the past couple days now I've taken the opportunity to get on the private server and test this thing out.
I'm sure it's not surprising that I'd be open to a concept like this, as I've always had issues with pokemon not being as competitive as I think it could be. I also feel like I'm one of the worst victims of game-changing luck out there. I think that, to an extent, my playstyle increases the potential for that to happen, but I'm sure that people who regularly watch me battle can attest to seeing plenty of more-than-ridiculous situations nonetheless. I think the arguments made in favor of this idea are compelling enough - the only real worry is that of 'changing game mechanics', which seems to have been adequately dealt with. We're only changing how the system rates each player, and I consider that a necessary and worthwhile change.
I mean, let's face it, a win in Pokemon means less than it ever has these days. There are even more alternate effects this generation, including moves for Serene Grace pokemon that they can actually abuse. Add that to the unavoidable fact that there is much more variety in each pokemon's usable movepool, and it becomes quite difficult to truly prepare. While we will always have to deal with the more dynamic factors (movepools and general unpredictability), I really think that if we can do something to curb the increasing 'luck-factor' of pokemon, we should.
Now, while I'm open to the idea, and I trust the ability of those involved to get the formula right, I'm actually a bit disappointed in the formula in it's current state on the test server. I really don't think that it goes far enough, at least not in the key aspects. I've yet to have a single battle overturned in my games, including a couple wherein I lost multiple key pokemon to hax. Right now, the formula doesn't seem to be able to take into consideration the potential 'lopsided' losses wherein you lose important pokemon to hax early on, and then end up losing 0-4 or so. From other observation, it does work well in closer games, but I can't help but feel like we're weighing certain types of 'hax' too heavily. For example, the situation regarding BK vs RBG is problematic. Regardless of the min-max on damage, pokemon still rewards you for making good switches. The crits and random burns/freezes/drops deserve a lot more focus than that. A fire move doing max damage instead of min to my Kingdra is almost never going to be an issue, but that same move burning Kingdra is a serious, serious problem.
However, as I said before, I have the utmost trust in you guys to be able to get this working correctly. I know how hard you've been working on the formula and it's implementation, especially this last week. I think that, with time, we can even find a solution for how the formula weighs close battles and hax which occurs early on. We might have to do something more along the lines of Obi's suggestion eventually, but for now I can see how a formula which overturns less battles would be preferable. It's better to go under than over for something like this, especially if there are actual plans to test this on the public server soon. Also, although it may seem like it doesn't do enough while battling, I know that in reality it's much closer to being workable than it seems. I'll try to follow this thread more than I have been, and, as always, I'll be around on IRC to discuss things.