You might be onto something here. If the queues get long enough, players might revert to custom games, like the good ol’ days when there were dedicated servers with human administrators to ban trolls and balance the teams. Imagine?
Tbh, I would rather have all roles within the same metric. While I understand that each role today has somewhat different metric. At same time the mmr of the player should be the combination of those 3 mmr (sup, tank, damage) due most of the knowledge being transferable.
Which is why, even after most of it’s lifetime being RQ. My Open/classic matches are way more balanced and actually “pretty equal” on Open/Classic, than ever is on RQ.
Over specialize a metric in a role that folks mostly plays with 2 heroes if something negates both at some point or that player would change it’s role for some matches, the performance of the matchmaking would be way off. Like one tricks and main something often doesn’t stay long on the same spot, similarly to folks that almost never play other roles and decide to play those roles.
The balance around this kind of thing goes way off. While if that said player were actually that rank material, that player would know at least the basics for how that rank performs.
The RQ conditioning is what throws several stuff off, due RQ itself only actually makes sense if folks couldn’t swap or/and folks queue for their hero, meaning the best metric in that regard would be hero metric. Which goes against most of the core and purpose of the game.
That and the fact of individual blame in a team based game instead of a team failure is what I think are major flaws on the system.
Open shares the burden and gauges everybody by the same metric, meaning you will matched against folks with similar behavior and performance. Which is a good thing if you play as intented but pretty bad if you not. But wouldn’t differ much on the negative part if you don’t play as intended, aside you being more likely to play what you want but not meaning that your experience would be better if you actually aren’t playing as intented. In theory you would experience more toxicity and more negative impact from it. But in both cases the experience would really painful.
One has potential to give more balanced matches on the long run, based on both time spent and player behavior. While the other the team will always be all over the place due queue time being painful otherwise.
RQ was a mistake trying to solve a problem on the most backwards way possible. Now folks got used and conditioned to it, to the point of 5v5 highlighting the problem they actually created to themselves.
The problem with this game is that never was actually planned and intended to have RQ, if they change something at core level, they could reach some sort of equilibrium between queue time, match quality, in game experience and player’s performance.
But to do that, would be way more overcomplicated than folks actually realize. Meaning that Occam’s Razor by simply going Open and create systems to avoid certain situations would be the best solution on the long run.
Their goal with placing folks at similar mmr mirrored but not in the entire match, was to reduce queue time which impacts on match quality. By praying that the match quality would be better than actually spending time to have a somewhat unbalanced match.
That is backwards, due how ranks works. Due that they opted to Narrow/Wide methodologies that themselves aren’t rigid enough due having failsafes to go wild if means saving queue time.
Ping and queue time are the highest priority in the stack. Which is why they’re coming with gimmicks. Sadly the impact of queue time is not that broad on MMR than their change, due if was their metric wouldn’t be required. But on the other hand queue time impact of the player perception and experience of queueing.
So, folks wait less to have way worse matches instead of waiting way more to maybe have a worse match. The more time spent the odds of worsened match increases but the degradation on the match quality is more conservative and narrowed than the alternative. Otherwise the alternative wouldn’t be able to find the match that faster.
I understand that folks want more quality even at expense of queue time. But is not okay to reach like 1 hour queue time for play a game. Which happened in the past for some reasons like avoid and certain ranks.
Queues became a problem because RQ, they could had solved the Open/Classic problem by other ways, but they didn’t and now any change would be unpopular and disruptive enough to have backlash. Hero bans were an excelent system if the game was open to balance the game and shift metas, but they introduced after RQ being implemented and conditioned folks to play 1-2 heroes. The result? Backlash from folks not being able to play their mains. Which Shouldn’t even be a thing, while I understand that some folks plays more with a particular hero, only playing a single hero or two is not how the game was intended to be played and devs never actually changed the game to work that way.
That’s the cronical problem about the queue time debate and their tricks trying to solve. There’s no real “good” solution to the traveling salesman problem which suffers from similar outcome to this. Their system (SBMM) doesn’t handle well groups with wide difference and can’t find fast matches if the population isn’t big enough due splitting the population in three(rq) pools instead of one(open)