... that is is Monday once again. Okay, technically it is Tuesday after a Monday holiday, but I ended up doing quite a bit yesterday and somehow writing up a post did not make it into the schedule. Instead, Guntir and I spent the majority of the day playing arenas with a Hunter we knew. Junglecleave, as it is called, is a silly team composition. It provides a lot of offensive pressure, good spacing for all teammates, and a ton of instant-cast crowd control. We ended up doing very well against all the teams we faced except for RLS and ThugCleave (although, one ~2k ThugCleave was farming us and in desperation we tried the strat "throw everything we have at the Priest" and accidentally killed him before he used Pain Suppression). Interestingly, what I want to discuss has little to do with the match-ups and composition strengths and weaknesses, though that is interesting; Guntir said it best when we were discussing the StealthCleave we were running with Tent as "StealthCleave is a solid comp, but it gets countered by too many of the FotM comps [RLS, VIP, ThugCleave]."

No, composition and matching is interesting, but I want to talk about the Elo system for match-making. Long-time readers will, just now, have said "again!?" rather loudly. It is true, I have discussed this system ad nauseum back when it was first explained to the player base, but with recent MMR changes I feel it is worth another look. For those readers who are unfamiliar, the Elo system of "fair match-making" was originally introduced as a means for rating and matching professional Chess players. The idea is fairly straight-forward and easily understood by anyone who has stepped into the arena: a weighted system that suggests, rather accurately, whether a given player should or should not defeat another player. Specifically, if you are 1800-rated and you face a 1600-rated, you should defeat your opponent if both ratings are solidified.

Interestingly, what becomes contended is when and how ratings become solidified. As an example, I had something like 111 games played in 3s on my current team up to yesterday and we had something around an 1800 rating. In fact, we very quickly got to around 1800 and then sat there for a large number of those games; something like 50-60 games bubbling right around 1800. Yesterday, we decided to give JungleCleave a try and ended up playing 65 games; we went 39:26:0, which is exactly a 60% win-ratio. Again, for long-time readers, bells should be going off. At a 60% win-ratio, our team should be showing signs of improvement in the Elo system. Essentially, we were at ~1800 and getting matched against other 1800 teams (having been firmly solidified at 1800ish), and we were winning more than half of these matches. The Elo system corrects rating by giving out proportional rating amounts to winners and taking (what should be equal but is not any longer) some from the losers.

Okay, so this rehash is sort of boring and old-hat at this point; you go above 50% win-ratio and you move up; I ended up finishing around 1900 yesterday. My concern is the distance traveled, however, in that we ended up playing 65 games over the course of many hours and ended up only gaining ~100 rating with a 60% win-ratio. Obviously, other things factor into the mix; we did end up losing to a fair few 1700-1800-rated teams to whom we should not have lost, but we balanced that out by also beating a few 2k-rated teams. Essentially, our low-losses were about 50-50 with our high-wins (we farmed a 2k Kittycleave down to 1900; they had to be pissed), and we ended up carrying a 60% win-ratio against evenly matched teams.

What I am mainly annoyed by is the rate of change. I ended up having, and I use this term lightly because it was fun doing so, to play 65 games over the course of probably 4 hours to gain 100 rating. I feel a 60% win-ratio is pretty strong, though definitely not gladiator material. In fact, I would have to imagine that a 60% win-ratio, if sustained, would eventually get a team into gladiator range all things being equal. In fact, the Elo system subtly points us towards this fact by showing that once a team sits around 50%, their rating stops moving. Okay, so I guess that if we continue playing this team for weeks and weeks and maintain our 60% win-ratio, we will eventually hit 2.2k, but that seems like an awfully hard grind, when compared to the PvE content. Maybe it should be hard; maybe the MMR-reset that has made 2.2k almost Gladiator-range on my battlegroup was a good thing overall. We are seeing fewer 2400+ teams and basically no 3k teams, and maybe that means taht the MMR system is finally evening out.

I do not know... I just know that it is frustrating playing 4-5 hours of successful arena matches and only gaining 100 rating when I am still 300 rating away from the best weapon I can get. I keep joking with Guntir that if I ever get the shoulders in an LFR, I am going to respec to Blood in the Water and find a Resto Shaman to run double-healer Feral again. Recall, I made the double-healer Feral team work with a Holy Pally who did not speak much English and it was before the days of Vengeance being hilarious overpowered; largely, nothing has changed for that comp from season nine to season eleven except that the PvE gear available is much more easily attained and much less of a detriment for me to wear. I have trouble giving up two pieces of iLvL 403 for two pieces of iLvL 384; twenty item levels is quite a bit of stats to give up, but the increase in damage and pressure from a 60%-range Blood in the Water is extremely appealing.

On a different swing, the fine folks over at MMO-Champion are of the opinion that the internal alpha has started and is likely pretty far along over at Blizzard. The reasons for this belief are pretty lacking except that the community managers have been answering questions on the forums, which is unusual, and posting in-game screen shots of tooltips with the spellbook opened in the background with texts like "NYI [not yet implemented".

I tend to agree that this evidence looks strong enough to support a claim that the internal alpha has begun. Usually, this means that while things are being nailed down, information will start to trickle out more and more; however, since Blizzard is planning a press event, rather than BlizzCon, in mid-March, I have to believe that the Friends and Family Alpha will not begin until then. The FaF-Alpha is always the time when information, that is supposed to be under a strict NDA, starts to leak out and Blizzard always has a hard time keeping such things under wraps. In fact, the NDA-breaking was so rampant and rough with Cataclysm that Blizz eventually lifted the NDA early and pushed the Beta a month earlier than planned (according to my source who got me into the Beta last time).

I would not go so far as to say that Blizzard likely will give us no more information until the press release, but I certainly would not be surprised by that should it come to pass. I am, of course, hoping for additional information like everyone else, but I suspect that Blizzard has given us our large information "hay bale" and expects us to live through winter until the press release.