The current state of ranked: the major problems, with some proposed solutions
Let's talk about some of the major issues that currently affect ranked games.
A lot of people are complaining about Dynamic Queue vs Solo Queue, and I'm not going to rehash that argument.
Instead, I'm going to list the main problems that are currently affecting ranked games, along with some possible solutions to each problem. Each of these problems lead to frustration and/or unfair matches in various ways (which, in turn, lead to more frustration, and maybe eventually to players no longer wanting to play ranked, or even quitting League altogether).
A fair number of these problems were introduced in Dynamic Queue, but instead of just saying "bring back Solo Queue", I will list the specific problems, and offer suggestions on how the situation might be improved upon.
Feel free to bring up more points and/or more possible solutions to the problems I've listed.
Premade groups and communication
This is probably the main reason that people dislike dynamic queue.
A group of friends playing together and using voice chat can coordinate things much more efficiently than solo players who need to stop and type in order to communicate with the complete strangers on their team. It is an indisputable fact that people with voice communications have an unfair advantage to those who do not, far more so than e.g. the jungle timers previously provided by Curse Voice This leads to scenarios such as the jungler on one team going bot for a gank, and the enemy top-laner instantly starting a teleport, their jungler and mid running to the lane, etc. As a solo player, that kind of thing feels really unfair because you cannot possibly coordinate with your team that well.
Possible solutions:
The problem is that the ability to communicate is uneven between the two teams. To solve this problem, the two need to be more equal.
- The communication ability of the premade could be reduced. This simply isn't viable. Even if third-party voice communication was forbidden, it would be unenforceable. Players could still run the software on other computers, on their phones, etc.
- The communication ability of the solo players could be increased.
- The game could provide built-in voice chat. This has the advantage of giving all players access to the same tools. It also has risk of increasing toxicity and making the game less pleasant. In addition, it is difficult to censor voice chat in real time, which could lead to other problems such as making the game unsuitable for younger players because it is filled with people shouting profanity.
- The game could provide better non-verbal tools for communication. The current ping system is fairly primitive. An expanded system with more options could help to decrease the communication gap, especially if combined with more auditory cues (such as being able to make your character play useful voice quips to their team, similar to Dota2's "mid missing."). Another suggestion might be a shot-caller's ping wheel, where someone could indicate that another player should move to a given position (e.g., a "Shaco go here" ping.)
The see-saw effect
Another part of the above problem is that the matchmaker has to adjust the MMR of teams to compensate for their higher chance of winning due to their communication tools, and also that players who win through communication don't have to develop their game skills to the same extent as those who compete solo. This leads to games where the premade team dominate in the early game (due to coordinated ganks, towerdives, etc.), but fall off heavily in the mid-late game due to their lack of game skill. Games often end up as a kind of see-saw, where first the game is unfair and frustrating for the solo players because they are being ganged up on by the team, but once they survive that early coordination the game instead becomes unfair and frustrating for the premade group, because they are being matched against players who are significantly better than they are. Finally, on a somewhat related note, applying an MMR handicap to premade teams is also extremely frustrating for the solo players on the same team as the premade, since they are being intentionally matched with players who have less MMR / skill than them, which can often lead to a huge skill disparity between the different members of the same team.
Possible solution:
I'm not sure if there is one. At the moment, ranked teams already win more games than solo players, and Riot said they are going to increase the "MMR handicap" even more to compensate for it - but that is only going to exacerbate the above problem and make the skill gap between the premade and solo teams even larger, amplifying the see-saw effect. Solving some of the other problems listed here, especially the ones about solo player communication, might help to even the playing field between the premade and solo players, which would help to alleviate this problem.
Teams vs groups of solo players
There are two types of premade teams - ones made up of people who know each other and use third-party voice comms, and ones made up of people who met in solo queue and decided to play together again for a game or two. The first group have the major advantage of knowing each other well and having voice communications - causing a lot of the issues with dynamic queue, requiring major MMR penalties and so on, while the second group are more-or-less solo queue players. There's a huge different between the two groups, yet the matchmaker has no way to know which type of group is which and applying appropriate adjustments accordingly. That means that games between a voice-comm team and a group of solo players will invariably be unbalanced.
Possible solutions:
There's no real way that the matchmaker could tell which of the two groups is which, so the only real solution would be to try to narrow the communication gap between the two, as described in the first issue.
4+1
The other side of the premade communication problem. When a solo player is grouped with a 4-player premade, then those players tend to communicate almost exclusively via voice chat, which means that they don't really cooperate with the single player at all. At best, the solo player feels like an outsider and has an even harder time coordinating with their team than with 5 solo players; at worst, the four-player premade heap all of the blame on the solo player and flame them constantly. In general, this is the type of team that solo players most dislike being in.
Possible solution:
Remove the ability to queue as four players. Only allow, 1-, 2-, 3- and 5-player teams.
Getting stuck in the queue
One of the most frustrating things about dynamic queue is that you can often get "stuck" in the queue, where you will fail to find a game for a very long length of time. Aborting the search and rejoining the queue will often find you a game almost instantly. As I understand it from various hints Riot have dropped, the reason that this happens is a bug / flaw in the matchmaking algorithm. The matchmaker creates "protogames" (games in the process of being created) and then tries to fill those with players. When a player joins the queue, they are either placed into an existing protogame, or a new one is created and they are placed into that (if there are no protogames looking for their roles.) The problem comes in when a solo player gets put into a protogame with a premade group, especially a large one. The matchmaker has to try to find another group of similar size and skill to place on the other team in order to make the game fair - that means that the large group will logically have a longer queue time. However, because the solo player has been assigned to the same protogame, they also get the same long queue times as the premde group. Because of this, there can often be many protogames at a given skill level, each with enough players to start games, but because each of them are waiting in their own protogame, no games can be started.
Possible solution:
The only real solution to this problem is to fix the algorithm. A matchmaking system shouldn't prematurely assign players to protogames in the first place. Instead, it should look through all of the players in the queue each second (for example) and try to build valid games with them, with a fair MMR distribution. Players who are in the queue longer can be allowed a larger margin of error when matching MMR. This would mean that as soon as the correct number of players of appropriate skill were in the queue, a game would be created immediately. No one would end up stuck in protogames.
Incentives for teamplay punish solo players
This one is simple. Providing incentives for one group of players punishes the other group of players. Playing as a group gives bonuses, depending on the size of the group, in terms of key drop chance, chest acquisition chance, IP (team IP bonuses, etc.). I don't think many players mind about the IP, but the key drop rate feels especially unfair, since a lot of Riot's long-time supporters are solo ranked players. Giving them less chance at keys than "more casual" players who play in large teams sends the idea that Riot care more for people playing casually with their friends than with the hardcore playerbase. There's also the box drop mechanics. If that is supposed to reward you for doing well on a champion (and also encourage you to try many different champions over a season), then being able to get boxes by queuing with someone else who does well seems very odd. However, since the number of boxes which can be found per month is limited anyway, this is much less of an issue than the key drop rate. Solo players may get there slower, but they'll still get the same amount in the end.
Possible solution:
Normalise the key drop rate. Have everyone get the same rate, regardless of how many players they are queuing with. This doesn't necessarily have to mean putting the drop rate for 5-player teams down to the drop rate of solo players; it could also mean increasing the rate for solo players to match 5-player teams, but realistically it should probably be somewhere in the middle (e.g. everyone gets the drop rate that teams of 3 have now.) (Low priority) Either have any S rank on your team reward all players on the team with a box, or have none do it. Alternatively, add some kind of compromise, such as all players on the team get a box if someone gets an S+, or have a chance of each player on the team getting a box if any player scores S- or higher.
Role selection and queue times
One of the reasons that queue times have gone up is because the players are now sorted into buckets by which roles they are willing to play. Each player can fill at most 2 positions (unless they pick Fill). On the flip side of the coin, people have to choose to play at least 2 roles, so someone who only wants to play one role exclusively can get a role they are much worse at in roughly half of their games. Role selection is a sliding scale, from a guarantee of getting your role on one side, and shorter queue times on the other.
Possible solution:
Allow the players to decide which of these two things is more important to them. Allow them to queue as 1-5 roles, with a warning indicating that queuing as a single role will lead to significantly higher queue times. Show the players the estimated queue times right there on the role selection screen, changing dynamically as the number of roles and the selections are changed. This will allow players who want shorter queue times to get them by offering to play more roles, while also allowing other players to opt for longer queue times for themselves but a guarantee of their chosen role. Also, this would be a far more reasonable place to give incentives to players - maybe offer a small reward to players based on how many roles they are queuing for or how highly in demand their roles are.
Competitive Integrity
In previous seasons, ranked was an indication of a single player's skill and ability to carry and win games. With dynamic queue, players can queue up in teams, putting less emphasis on the individual players' skills. Moreover, a weak player can queue with several stronger players and use those players' skill to boost themselves to a higher ranking. Ranking is now as much an indicator of being able to find and cooperate with the right premade as it is of individual skill. Also, since the rankings are shared, played who normally play in teams may find themselves matched in games that are out of their league when queueing as solo players (though the MMR penalty helps alleviate this problem somewhat.) In contrast, however, I feel compelled to point out that solo queue has never been only about individual skill either - while skill was a larger factor, wins and losses were still always affected by which players were randomly grouped together (with different skill levels, playstyles and toxicity levels being mixed randomly and greatly affecting game outcome.)
Possible solution:
I can't really think of anything. Narrowing the bands within which you are allowed to dynamic queue together might be a start - for example, only being able to queue with someone ~4 tiers away from you (so Gold III can queue with Silver I and Plat V, but Gold I can't queue with Silver V any more.)
Premade threat
Playing against a premade group as a solo player feels unfair, because they can have much better coordination than you and your allies, and you often have no way to know if they have this advantage or not. Whenever someone does something well-coordinated, there's a tendency that they must be a premade group abusing voice communications, even when they are not, leading to increased frustration. Also, with no way of knowing which players are premade, you can't make informed decisions such as expecting the enemy to camp your lane if they are premade with your laner.
Possible solution:
Show an indicator linking the players of a premade group, visible to all players. http://boards.na.leagueoflegends.com/en/c/gameplay-balance/E5Trdqkj-riot-at-least-copy-the-simple-line-connecting-premades-in-dota-2
Toxicity
This isn't really related to the queue types, but it's still a major factor in ranked games (and in other game types.) The level of toxicity in the game has certainly gone down this season, but it's still there. Most games still have people asking for false reports on anyone who does badly, other types of flaming are still fairly common, and other negative behaviours (such as going AFK) still crop up from time to time. These kind of things can still affect the outcome of games, and each game which is ruined by toxicity makes the matchmaker slightly less accurate (e.g. if a player ragequits and his team who would otherwise have won ends up losing, then 9 players end up with their MMR moving in the wrong direction, leading to less balanced games.)
Possible solutions:
A lot of toxicity - certainly the more minor stuff such as report-calling - seems to be caused by people just not understanding that what they are doing is toxic. A quick glance at the number of "I don't get why I was banned" threads certainly points in that direction. Warnings and bans shortly after their games is one thing, but it still doesn't really let them know what is and what is not acceptable behaviour. A system of real-time notifications could help to teach the playerbase how to behave. For example, if a player types "report Garen", the system could block that message and instead display an onscreen prompt indicating that asking for reports serves no purpose and is considered negative behaviour. Similar prompts could be added for other (more serious) types of negative behaviour, including gameplay behaviour, such as detecting when someone says "I'm afk" and then stands at the fountain and immediately informing them that they are breaking the rules. (I'm using "report X" as an example here because, while it's a very low level of toxicity, it's also one of the easiest things to systematically detect.)
Champion design
Another thing that has nothing to do with queues, but is a major factor in deciding ranked games, and also leads to a lot of frustration and unfairness, just like the queue problems. Certain champions are oppressive. No one wants to play against them. It might be because they are overpowered, but mostly it's because their kits are poorly designed and either offer little counterplay or (worse) allow the champion to negate too many of their opponents' plays. These types of champions decide games, both by being oppressive to the champions that they hard-counter, and by putting players on tilt because there's almost nothing they can do against them if they are playing one of the champions that they counter, which in turn leads to even more frustration and negativity. In this season, the outcome of ranked games has moved even farther towards rewarding those who are willing to just play the most broken champions at a passable level instead of those actually willing to invest in learning their favourite champion inside-out. Meanwhile, Riot almost never make meaningful changes to these champions. They might make a few minor number nerfs here and there, but never want to address their actual gameplay problems.
Possible solution:
Prioritise reworks of problematic champions that everyone hates to play against, instead of focusing on champions who maybe don't have enough unique about them or are a little undertuned. A weak/undertuned champion with no true niche is sad for the players who love that champion, but generally doesn't affect the rest of the playerbase that much. A badly-designed champion who is oppressive and turns every game they are in into a nightmare for the enemy team, regardless of win or loss, is a much bigger problem that needs to be addressed far more urgently. Riot have a very, very bad history of not prioritising this enough, and not listening to the community on such issues.
