Back in my Dota Matchmaking FAQ series, I said that in my opinion a lot of the perceived failings of Dota’s matchmaking are social issues rather than technical failings. Today I want to talk about one of those social issues that keeps popping up: the problem of highly rated stacks, stacks being a group of players who queue up together and play in a highly coordinated way.
Dota 2’s matchmaking allows players to play in a group. It needs to allow this. Some people cannot enjoy the solo queue game, and Valve is obligated to offer them a form of casual matchmaking so they can play with their friends. In order to expedite the process, matchmaking is willing to allow these pre-made groups to face solo players because it selects solo players with a higher overall skill level to compensate for the pre-made group’s greater level of coordination. It’s not a perfect system, matchmaking never is, but for the most part it gets the job done, and I expect it’s still undergoing constant refinement.
The problem this system runs into is that when a group of sufficiently highly rated players queue together, the matchmaking system doesn’t have a sufficient population of even higher skilled solo players to pick from. The matchmaker does what it can to make decently competitive games while not allowing queue times to skyrocket, but the results are not pretty. The picture here is a little clouded. We can only see who these players have played with in the past month, inactive players are not filtered out, and Dotabuff cannot show us players who have decided the remain anonymous. Nevertheless, if you click on most of the players on that list you will find many who regularly play in groups that achieve +70% win rates. It’s true that Dotabuff only shows us 50 people who have a greater than 70% win rate, but it is very likely that win rates are distributed in such a way that the 60-70% group has a much higher population than the +70% group.
What this means is that there must exist a group of players who are absorbing those 30-40% win rates on the other end. Solo queue players who are constantly matched into games that we know are not statistically competitive but that are created for expediency. What this does is create a burnout point, where the primary way these players interact with the game is so frustrating that they lose interest in the game as a whole. Some of these players will quietly quit, or at least scale back their playtime, and we will then just throw the next tier of players into the grinder so that the cycle can continue.
Some say that these players should be grateful that they have the opportunity to play against the pros and learn from the experience, but I find this argument rather unconvincing. If you were to create a game with 8 talented amateurs and a profession player on either team, that could easily be an educational experience for everyone involved. You’d have a source of direction and insight that you might not normally have while also being tested by a similarly directed team. Being stomped by a pro+friends stack in under 15 minutes? Yes, there are things that you can learn for it, but if it is the 5th time that night you’ve been stomped in such a manner then you’re probably at the point where you’re no longer learning much. And let’s remember that if you can recognize a pro player’s stack by the names, it’s pretty likely the other 4 people on your team can as well. Some of those players are going to give up internally before the creeps even spawn, and being essentially down 4v5 against a far more coordinated team is not an environment well-suited for learning anything.
A alternative solo queue would address this issue, but the danger is that the solo queue could become so popular that finding a casual team match would become difficult or impossible. As I mentioned earlier, casual team matching is a necessary feature of the system and is as important as addressing this burnout point. So, how can we create a solo queue that won’t completely cannibalize the more general queue? My answer starts with this:
![[Feb]Modes](https://dotametrics.wordpress.com/wp-content/uploads/2013/03/febmodes.png)
It’s not a perfect sample by any means, but a 75% representation for All Pick jives with everything else that I’ve seen. The most straightforward solution is to make a solo queue that does not include All Pick. The most appropriate default mode for solo queue would be some variant of Random Draft, and if it’s proves popular enough it could eventually add something like All Random or Single Draft. Meanwhile, solo queue is treated as the opt-in mode and doesn’t include the most popular mode in Dota 2. These two facts should allow the general matchmaking queue to maintain at least a 30-40% All Pick representation, which should be more than enough for efficient group matchmaking. General Random Draft might suffer, but in exchange we would have a solo queue Random Draft with a significantly higher population than we have now, which should improve the match quality for dedicated RD players.
Most importantly, this would give highly rated solo players the means to avoid stack burnout. The question then remains, what does matchmaking do with the stacks who now have even less competition than before. My suggestion is to just let the matchmaker search for however long it needs to to find an actually even matchup, warn the stack that the queue might take a significant amount of time to process, point out that they would have a much shorter queue time in Team Matchmaking, and then automate team creation and switch them over to that queue. This would serve as a much needed boost the TMM population, and frankly, if your stack is at a +60% win rate in general, TMM is probably where you belong.
This is not to say that TMM doesn’t have issues right now, but the best way to address those issues is to get more players playing that mode and giving feedback on the problems they have with the mode and how they’d like to see it develop in the future.
DotaMetrics 
Wow once again, your article pointed a big problem in the game! Good solutions as well! I hope valve change the matchmaking like how you adviced.
I’ve actually thought about how to improve my winrate as it is only 52% .( I solo queue most of the time) and I actually thought about playing all random to prevent stacked teams.
Last time I posted and asked you for advice you told me to try Natures Prophet and get medallion, watch the map to improve map awareness. I tried it and NP turned from never being played by me one of my fav heros. I play him everytime! People even compliment my map awareness in games. Because np is a difficult hero to master, due to having to many role possibilities and builds, my winrate with him is just a little under 40% . I’m becoming really good at him, just can’t pull the low winrate up.
However as a dota player, carry player to be specific, my abilities to play a carry has gone down, as I try to tp to fights and worry too much about my team when I farm now. I don’t know, maybe I should take a break from the game and pick it up later. As I improve I’m actually making friends with the top skilled players in the game, the ones on the front page of live games. I do we’ll in those games tbh.
Overall I think I have improved, I just gotta put very thing together :)
Hope to read more articles from you!
I’m glad that you’re having fun forcing yourself out of your comfort zone. The best heroes to grind away at are the ones that you simultaneously love and struggle with.
As for knowing when not to join fights as a carry, it’s tricky. Sometimes the right thing is to simply continue freefarming if you have the most favorable gold -> power ratio in the game. The timing on when to stop freefarming also depends heavily on the hero. Faceless Void and Spectre might do it best in small bursts in conjunction with the cooldowns on their ultimates. Lifestealer will definitely want to do it earlier than Antimage does because of how different their strength curves are despite both being relatively hard carries. Also keep in mind that having good map awareness sometimes means being able to tell which situations TPing into are sheer suicide. Many times the best choice is to just farm faster and try to create a tower trade.
It’s a bit second hand, but I recall someone saying that according to Dendi if you leave mid before level 7 and don’t get a double kill, you’ve lost the lane. Part of playing a hero with a heavy carry or semicarry aspect is just learning how to make the call on when the opportunity cost of roaming and ganking is no longer prohibitive. If you’re mid Zeus and one of your sidelanes is struggling, it’s very tempting to go help them. But if you go and fail to accomplish much, you’re delaying your ult a minute or two , and that one lost ult could be enough to cost you the game.
dblu —
The danger of that system is that if you demand rigid team constructions you could get to the point where you have too many groups of 2, 3, or 4 people without complementary groups of a similar rating to match them. A group of 4 people is actually a construction of 4 different MMRs, and it could be a nightmare to find another group of 4 people with a relatively equivalent MMR distribution. It might be the case that a 4+1 grouping is a better competitive match to some 3+2 grouping than it is for any other potential 4+1 grouping in the next 30 minutes.
Effectively though, what I am saying is similar in that you have an extremely successful 5-man group, we should be seeking to match you against other extremely successful 5-man groups. I don’t have enough statistics available to say whether 4+1 vs 2+1+1+1 gives the matchmaker conniptions, but I do have reason to believe that top end 5-man groups are not being sufficiently challenged.
Like I mentioned in a reply to another post, you’d just have to match groups with two people with the equal amount of groups of two on the other side and so forth. For example, q’ing with three people you’d face a group of three and either two solo q’ers or a group of two, depending on what the system decides to put on your side. No matter what, the “group composition” would always remain in balance.
Even though HoN has way smaller population than DotA 2, the queue times there using this system aren’t too bad at all. When the change to this system was made in HoN, I didn’t notice any difference in how the games were matched in terms of MMR either.
As far as I know this is what happens now: whenever my low rated friends queued in a stack, the enemy team would also get a stack. (at least, we could check this easily before when DOTABUFF was fully operational). The problem that Phantasmal is referring to is in the top 1% of matchmaking, where it’s difficult to get an opposing stack at that skill level because the population is so low. Which by definition makes it a “top 1% problem” and probably not something that will bother the 99% of the players… I queue nearly every game into Very High and I don’t honestly find stacks a problem. I remember playing HON in the beta from before there was matchmaking, and you had to manually search for games. I loved that aspect, honestly. It didn’t take as long as finding matchmaking goes now, even, you filtered by game type and ratings and picked a game from the browser. You could choose to play 5 solo queue vs 5 solo queue, or you could play stack vs stack, or the final option, which people rarely took, was play 5 solo against a 5 stack. I found that win rates for all 3 modes were roughly equal… provided the ratings matched up. Sure, stacks have an advantage, but a lot of that advantage is already built into their ratings, so it’s “fair”. Also, I felt the mentality of the players a lot better in that scenario, where you KNOW before going in that it’s going to be 5 solos against a 5 stack, and you’ve made a conscious choice to seek out that game and play it rather than playing a 5 solo vs 5 solo game. Compared to DOTA, where you see a stack and go “Hell no, I didn’t sign up for this…”
Can’t remember if I’ve said it here (I probably have) but my ideal solution for matchmaking this game would be to allow matchmaking only for full 5 stacks. Everyone else gets a manual server browser, so you can pick from true all solo queues, to stack versus solo, to partial stack vs partial stack. Giving players the feeling of control is often underestimated: a lot of the rage in DOTA stems from feeling helpless when the matchmaking “system” gives you a bad game.
As much as this sounds like a decent idea, it’s one that would likely fall flat on its face. In dota1, pub games were almost entirely comprised of the lower skill brackets, with inhouse leagues, external clients, and clan inhouses often snatching up a lot of the better players. Even then games were often stomps due to the lack of matchmaking, and something as simple as a single/duo player with a decent gimmick could absolutely run away with a game. Shit, I used to run Alchemist/Bloodseeker as a “pubstomp lane” with my best friend, and we tended to carry the majority of our games with that setup simply because of a little character knowledge and a game plan.
It really wasn’t until the advent of public inhouse leagues (Throneit being the one that springs to mind) that games started to feel more consistently even, just because they used a matchmaking rating to place better players in separate games. As much as I’d like to hop on and have a few relaxing pubs where I stack Mjolnirs on supports, this isn’t the kind of thing that’s going to lead to fun games for the majority of players.
Deathbot: What I’m proposing is not a “free for all” public game system with stomps galore – it would be what HON implemented in the Beta test – there were ratings restrictions and game balancing (for example, a room could be rated for a certain band of players, restricting who could join – when you searched the server list, you would only be shown games you were eligigble to join). Additionally, the game had “win predictor” that took into account eeveryone’s rating and if it displayed the win probability for each side,also it allowed the host to auto-balance the game by shuffling people around to achieve a close to 50/50 match. Most of time given 10 players within a narrow band (20% percentile skill band) it could shuffle players around to get to a 50% win prediction.
It worked well, I thought, and it allowed me the freedom to deliberately seek out stacks and play them.
That sounds very interesting, I guess I misunderstood you. An interesting marriage between the openness of WC3 pubs and the matchmaking we’re looking at right now. The only problem I really see with it is Valve apparently preferring to not have public winrates/mmrs, but that could likely be solved with more vague indicators for whether a game is going to be above your head.
Where is the proof that solo queue players get matched against better players? i almost always solo queue and im sure that i play against people of my level, stack or not.
Just, where is the proof? people with good winrate playing together? well, they are just good players and they are in high skill, so who cares?
I have played 2000 hs and i never had this problem, you can always have an occasional stomp.
Solo queue players generally get matched against worse players when they’re playing against premades. This is part of the balance that exists to allow premades to get quick games. When you have a premade that is, as you put it, just good players with a good winrate playing together, the system can’t find the necessary players to create that handicap.
Yes, you will always have an occasional stomp. We’re never going to eliminate occasional stomps. The problem is that if you’re a solo player and you’re repeatedly running into these teams that can maintain +65% win rates, you’re getting more than just an occasional stomp. You would be regularly running into games where you know ahead of time that you have poor odds of winning. And if you’re running into this problem today, you’re likely also running into it tomorrow. And the next day, and every day after that until you tank your rating far enough that some other group of players is now getting thrown into the grinder.
It also doesn’t matter how many hours you’ve played. We have no evidence that you’re in the range of players that would run into this. You could be running into it and just not care about it, but this doesn’t mean other people are wrong to care about it. Or it could be a somewhat overblown problem, but providing a way to avoid it might make people feel better about matchmaking in general because it removes the very perception of the problem, with the added benefit of creating a second populated queue option that isn’t AP.
Tonight played against same 4stack (may have been 5) in solo queue twice. Lost resoundingly twice (they had good co-ordination).
MMR tried to match me against them an additional 4times.
*_*
I may be experiencing this “burn-out” tier.
Very interesting read. I like the solution you proposed as well except for one thing; people don’t like to wait. Long queing times would cause a decline in popularity of SMM and thus increased interest in TMM which would lead to further stomps.
The system is very good in my opinion, but it is just ahead of its time. When the population increases wait times would automatically be reduced while aloowing balanced matches.
All I can say is we should just keep waiting for the official release and the problem should rectify itself.