We went over the relationship between Solo Matchmaking and Competitive Dota, and Valve has made a bold step forward to strengthening that relationship. Let’s take a deep dive into what this brand new update means for the Amateur scene and competitive Dota in general.
A polarizing topic in my previous article involved the idea of how MMR is perceived. An underlying theme throughout the article was that “MMR is just a number”, and that one’s rank doesn’t give a completely clear idea of how good a player is. I mentioned that a 7k Support spammer likely has 4k Midlane ability, and vice versa. This example is incredibly extreme, but many analysts in the scene seem to agree on this. Dota is incredibly complicated, and having a number be an indicator of your abilities in each and every aspect of Dota, is just not possible. That is where I expressed the need to pursue a competitive career, to build the other skills needed to become a complete player.
As I’ve mentioned earlier, Dota is an incredibly complex game. Some skills that a player could perfect include that of last hitting, efficient jungling, efficient movement, ganking, team fighting, warding, itemization, communication, and I haven’t even begun to scratch the surface. Having a single number be a judge of how well you perform all these things in a game, can be an incorrect perception of your skill. From Valves point of view, all that mattered is if you won or lost a game. Valve doesn’t care how many wards you placed, how many kills you got, or how many creeps you scored. After this update, they still don’t care, but at least they’ve taken a new approach to measure your skill.
Previously, a players MMR was split into Solo MMR, and Party MMR. The problem with Party MMR is that it is not useful after divine (about mid to high 4k), so there is absolutely no value in it to be used by professional teams. So, if by some chance a player had 9k party MMR, then no one would give a hoot. Party MMR was Valves attempt at bringing players together, but the decision just pushed us farther and farther away. That is until, this update, where Valve has split your MMR into Support and Core MMR.
Careful What You Wish For
In a perfect world, Valve could implement a new matchmaking system which involves a calculation of your performance in the use of last hitting, efficient jungling, efficient movement etc, rather than if you simply win or lose. Unfortunately, we do not live in a perfect world, so Valve has taken a much simpler route, to making MMR more accurate to a players skill. In the past, players just selected all pick, then pressed Find Match. Fast forward a few years, and Valve implemented a ranked roles feature which was exclusive to Dota +. Ranked roles gave you the ability to select your role before queuing, in the hopes of reducing conflict from the picking stage. You chose between Mid, Carry, Offlane, or Support.
Players have been pushing for Valve to make this a free feature for everyone since many of its issues stemmed from a lack of players. However, Valve continued to be patient, and just tested the waters with it being exclusive to Dota +. Now, armed with the knowledge of a year of ranked roles, Valve has made the service free, and better than ever before. Before you push that Find Match button, you will be prompted to selecting your role. Roles include Hard Support, Soft Support (brand new terminology for the position 4 role), Offlane, Mid, and Carry. The main tradeoff with this system, is that players (especially higher ranked players) may experience longer queue times, but in exchange, higher quality games.
Now what does this re-imagined system actually do for us, and how is it useful? I mentioned earlier that Dota 2 is too complex for one number. Valve has tried to remedy this issue by splitting the player base. On the one hand, you had Solo MMR players, and the other had Party MMR players. The biggest oversight was that Party MMR has received no incentives to grind for or focus on, so the community focused mainly on Solo MMR. Now Valve has decided to split the community once again, making a Core MMR and Support MMR.
At the start of a game, if you choose to select Hard or Soft support, then your Support MMR will be affected. If you choose Offlane, Mid, or Carry, then your Core MMR will change. This is a very elegant solution, and I mentioned that it kills 2 birds with one stone. The reality is, that it kills many more birds than we initially thought. Let’s quickly run down the pros that exclusively selecting your role does for matchmaking.
- reduces conflict at the picking stage (no more roll for mid, thank god)
- promotes conscious team building
- forces players to commit to a certain role (supports will fully support now)
- improves game quality
- reduces toxicity as a result of less reasons to argue
- Slightly reduces skill floor for newer players
Now for the cons
- Longer queue times
- Reduces flexibility in drafts (fixable in future updates where roles can be swapped)
- demotes experimentation to possible new roles
- Alienates high tier players
Valves main approach to tackling the solo queue issue is to be more specific to what your MMR represents. Instead of your “solo MMR” be a judge of your entire ability as a player, we now have more accurate measures, with a rank representing your support abilities and another representing your core abilities. I’d like to preface, that in a perfect world, we would have a Last Hitting rank, a Farming Rank, a Warding Rank etc, but that is simply too much work. Maybe with the advent of AI in the future, some sort of system could be placed, but as for now, the technology just isn’t there yet!
The first impression of many community members was that this update would be extremely exploitable. In a scenario where an individual plays an incorrect role, ruining the experience for the whole team. This most definitely would be an issue in the first few months. Valve, however, would easily be able to accurately identify these players, and punish them. For example, if a player who selects support, ends up getting 4 reports from his team, and has the highest net worth, then he most likely is not playing his role. The idea works both ways (for cores playing supports and supports playing cores), but the point is, the griefers are easier than ever to detect.
The main victims of this change would be the higher MMR players. I interviewed a few high immortal friends on their feelings on the patch, and a reoccurring complaint was imbalanced matches. Once you get higher up in the ranks, queue times get unbearably long. This is due to a much smaller concentration of players holding high ranks. The combination of a small player base, and being required pick a role means queue times must either go much longer, or matches become much more unbalanced.
- This graph shows us that the concentration of players in higher ranks is incredibly low. (The spike in divine 1 is due to a high number of boosted accounts in that mmr)
- Graph obtained from dota.rpg.io
But in my opinion, the biggest benefit that this update brings is the fusion of Solo and Party MMR. If a player had a Solo MMR of 2k and a Party MMR of 5k, then that player would be shunned as “party boosted”. If that player can continue to win games in a high bracket, then doesn’t that mean that his rank is accurate? Even if that player’s teammates are of much higher skill level than he is, at a certain point, his party would plateau and have a 50% win rate. The point is, the means of a player getting to a rank is not as important as his ability to remain on the rank. If a 5k Party MMR player is winning 5k games, then who cares how he got to 5k?
Not to mention that team-play will hopefully be more emphasized in the future. Stacks of 5 will be more willing to join together and grind together. This primarily helps the tier 2, 3, 4 scene and below. Let’s say you made a stack and wanted to participate for TI open qualifiers. Since you’re a weaker team, you want to test the waters but you don’t have any scrim partners. In the past, playing party queue with your stack would put you against much much weaker opponents, because the players would have much lower Party MMR than their Solo MMR. Now that your MMR is normalized, players can party queue against teams of similar rank, which is immensely valuable for any smaller up and coming teams in the scene.
As an avid supporter of the amateur scene, a change like this was a long time coming. Players have reached out to me showing their interests in experiencing team emphasized Dota, and this update gives team players a reason to play. It will take a couple of months to fine-tune the system, and many exploits will be found, but hopefully, Valve will swiftly address all the issues in a timely manner. I expect to see the level of amateur Dota rise up drastically in the next year since these teams will have a platform to grind on, and build ideas. Although this update is fresh out the gate, I’m incredibly excited to see the changes it will bring.
— DOTA 2 (@DOTA2) August 8, 2019