Alright I'm going to try to explain why I believe most rating systems collapse in Toribash, because I've talked about this at great lengths and changed my view point many times.


ELOs are designed specifically for One vs. One matchmaking systems. The perfect example of a pretty much perfect ELO/MMR system is Chess. So now having said that, we can establish two things Toribash would need to be to have an actual ELO system:

1) One vs. One Fighting

2) A matchmaking system

We've got #1, we don't have #2. Their is absolutely no form of match making in Toribash. The closest we have are rooms that are separated by belts, but as we all know, Games Played is far from 1 to 1 with skill. Much like their are people with 200 games played in Challenger rank in League of Legends and people with 1,000 games played in Bronze 5, there are good Toribash players with only 1,000 games played and some crappy ones with 6,000 games played.

Because you can play against anyone at any time, and you get to choose who you play, the system is entirely broken. If Bobby Fischer wanted to get a higher rating in chess, he couldn't just target 1000 MMR players, and farm them like nothing. He exclusively played the absolute best players in chess, as that were literally the only games that affected his rating.

Any "pick your opponent, every game counts" system is going to get exploited no matter what your variables are. It's extremely obvious that this is the chief problem in rank both in Clans and on the individual level. I guarantee you "Frost" (who?) didn't get rank 1 globally by playing a bunch of boxshu_mushu vs. high level players like Shmevin. And I hardly think "Litmuz" (who?) should get rank 3 globally because he played a crapton of Shaolin Soccer. If they had to actually play good Toribash players, they would be the ones getting farmed. And on the Clan level, (make no mistake I mean no malice or ill intent by saying this. I respect the shit out of these clans for what they bring to the game): WAPOW keeping ranked 1 in Clans for it's wins against "Shaolin"(who?), "MHE Dead Clan"(who?), "Sprite"(who?), and "TeamAsu"(who?) is the same equivalence. Every year in clan league a lot of these clans try to compete, which is awesome I'm glad they do. But every year, they show why the top 10 clans in the Clan List Ranking System are all not actually anywhere near the top clans in strength. Small exception being Obey, who haven't come close to winning any major event, but at least usually make the top 16 and do bring a lot to the table to mid-tier to low high-tier play. Nitro last year in Clan League shocked the scene by destroying Inq to pieces. Their performance in Clan Siege; however, was a bit on the weak side. I believe their were roster changes.



So we would have to drastically change how we rate players and clans in order to make them reflect strength. The problem with that is, there isn't incentive for developers to do that. If newer unofficial clans of younger people, who are extremely active, see themselves rising the ranks to #1, they get super excited and play a bunch more. They play a bunch more as a team too. They get hooked to Toribash, and they spend money on this game. Toribash devs are probably going to value the money brought in by keeping the "Ranking System" as a "Pat You on the Back System" over actual assessment of skill among clans and individuals. Anything that encourages activity in Toribash, especially to this degree, is a homerun. Luckily, they still host the biggest event in the year, Clan League, where we can see how strong clans actually are in a fair setting using actual competitive mods.