DOTA 2

Three Lane Highway: ways to think more usefully about your Dota 2 MMR

Chris Thursten at

Three Lane Highway is Chris' sometimes serious, sometimes silly column about Dota 2.

Last week I wrote off the concept of MMR as part of a not-entirely-serious list of 'meaningless' numbers in Dota. My thinking at the time was that discussing the problems raised by ranked matchmaking at all was going to attract a particular attitude in the comments, so I'd be better off treating it as a punchline. That was an error. I tried to use irony to mask something that I think and care about rather a lot, falling into the same trap that I'd accused certain competitive players of falling into only a week earlier. Sly winks don't carry well on the internet, and when you're discussing the relative worth of somebody's internet wizard skill rating it's fair to assume that most readers are going to take it pretty seriously.

This week I want to unpack that original point in a more detail without the protective cladding of irony. For the record, though, 'meaningless' is a self-consciously overblown descriptor to attach to any number or value. Of course a player's matchmaking rating has meaning. What's worth discussing is how that meaning influences the community and the types of discussions that take place; how it alters people's playstyles and their relationship with the game; how much value it really has.

My view was that Dota 2's MMR system has a disproportionate impact relative to its usefulness. That it is both important and a little hollow, which is tricky territory to navigate in any context. It's something that people invest a tremendous amount in when I'd argue that their efforts would be better directed elsewhere. It has the capacity to upset and divide people beyond the extent that Dota 2 is already capable of dividing and upsetting people. I believed then—and still believe now—that there's a subset of players that regard MMR with a dogged gravity that distracts from what's interesting about the game.

What I failed to take note of is how it is possible to process your MMR into something workable and useful. I still believe that, taken at face value, MMR has promoted a lot of bad behaviour. But I've come around to the idea that it's possible to grapple with the data in a way that is ultimately good for your experience of the game. Getting there, however, means accepting that Dota 2 isn't a single game with a single progression path: it's a hobby with dozens of legitimate approaches. Custom game modes will eventually make that fact obvious. For now it's something that players have to work towards themselves, even as a substantial portion of the community sets out to tell them that the way they play is wrong.

I don't believe in the 'trench' or 'forced fifty'. Enough studies have been done to prove that the MMR system is fairly good at placing players where they belong. In particular, kyuronite's experiment shows that a good player will eventually climb or fall to the place they need to be in. MMR becomes interesting—and problematic—at the point when a player decides that they're not happy with where the game has placed them. The default response of many is to blame the game or the people they're placed with. More level-headed players are likely to follow the advice given here by SirActionSlacks, assuming a support role that gives them maximal control over the outcome of the game. You might be able to achieve the same thing while playing independent heroes like Nature's Prophet and Tinker, but the principle is the same: that your random teammates must be assumed to be a liability, and that you have to treat Dota like a singleplayer strategy game where you have no direct control over four of your most important units.

My negative view of this system was grounded in the mistaken notion that I outlined above: that there is a single 'correct' kind of Dota, the type where two teams of five people fight to control the map by working together, and that chasing MMR by any means would ultimately distance you from the kinds of skills you should be learning. That this one-protect-four mentality is good for raising your MMR but only partially effective as practice for whatever comes next.

But that's okay, actually. I've come around to MMR a lot more by thinking of Dota as three different games divided up by the buttons on the menu. Unranked is where you go to experiment on the understanding that everybody else is likely to be experimenting too (or playing Pudge). Solo Ranked is where you go to play mother hen to four angry strangers, and Party Ranked is where you actually play Defence of the Ancients. Advanced players have access to a few more levels of play—competitive matches, scrims, in-houses, and so on—but those primary three are the ones that concern most of us.

Thinking about MMR usefully means being always aware of the context in which you're playing. Your solo MMR gives you a good sense of how well you're able to secure victory despite other people, while party MMR assesses your ability to play with them. Ultimately you'd want to be so good at both that the distinction fades, but chances are that's not where you are at right now.

Don't beat yourself up if your solo MMR isn't what you want: instead, figure out if the skills you'll need to 'fix' it are the ones that are the most important to you. If they are, lock Omniknight or Dazzle or Treant Protector or whoever and mother the hell out of your team. If they're not, feel free to ignore the next person that speaks down to you because of your rating. Listen to detailed advice when you can get it, ignore rage, and try to keep a clear head. Learning not to lose your shit when you drop 25 points is as big a victory as not losing those 25 points at all.

If a bad attitude and a lack of self awareness causes some players to wield MMR like a weapon, then the right attitude has the opposite effect. Your MMR is only as useful as you allow it to be.

To read more Three Lane Highway, click here. Chris will be covering this weekend's ESL One Frankfurt Dota 2 tournament from the event, so check back on Saturday and Sunday for a first-hand account of the action.