Actually in a game, and 8 silver, 4 bronze, opposing team has bronze tank, bronze dps.
Our team, bronze dps, bronze support.
If SR is an extension of MMR is it fair to say that the bronze are in “different” roles and thus fundamentally shift the balance of the game?
Would it be so terrible to match up the “lower” ranked folks in each role? And also mind you if “everyone” was in silver is it just some are about to be gold vs the “low” silver?
You know statistics would REALLYLLLLYLYLYLLYLY help right now.
Ideally, the roles themselves would be better matched. SR is not so much an extension of MMR, as it is a representation boiled down to a number. Think of MMR as a set of variables, through some mixture of science and witchcraft that is then reduced down to “SR” as a representation.
The issue I have with using SR (and its stated somewhere its a 1000 spread, but some claim its actually MMR on this range), is that say you have a silver DPS, but they are a Plat or even Diamond Support and have very good positioning and game sense. Even with weak mechanics, they are still way better than a person who is silver in all three roles. This is why the games in silver I think in particular often feel harder than Gold or even Plat. Some is random play, but think this aspect contributes more than Blizz counts it as. I think your strongest role should count a bit more by the matchmaker than it does because those skill seem to really help more than it gives credit for.
In theory those Plat to silver players should excel faster, but what seems to occur to me is some kind of relative “blurring” effect in a tier, where players of radically different skills seem to get lumped in when they ought not to.
Why would they be better than a person who’s silver in all three roles? Obviously their gamesense is way above silver, but their aim must be way below silver - otherwise they would be ranked higher already.
IIRC ranks/SR are basically percentiles. So if I’m a silver dps, that means I’m a better dps than 15% (made up number) of all dps. Different combinations of skills can get me to that SR, but my average skill is the same.
Obviously the ranking system isn’t perfect, but I don’t think that being better at one role means the system breaks for you.
Ahhh… and that is my exact point. They should be, but I have been playing enough to spot the trend looking at open profiles. A player who often is Plat or higher on a role does substantially better than a full-Silver player despite the relative rank for the role. They won’t stay silver long (but oddly I see them often across several seasons in that tier), but in the meantime they are pub-stomping the true Silver player to death.
The compound problem is if their mechanics are so-so, then the system just says “welp, despite being Plat on say Support, this DPS is pure Silver” and actually keeps them in that tier longer than they should be because instead of just win/loss, its using PBSR.
My entire point is they are “basically percentiles” but how that percentile is derived is not taking into account enough their other role rankings. Literally, that is the entire point. There is a vast difference lately in Silver and even Bronze unless Smurfing is far more rampant than anyone cares to admit.
The game uses too many metrics to determine a players rank. That and the fact they won’t allow outside programmers to inspect the code to help them find flaws shows they don’t really care about it being a competitive game.
I honestly think the reason they don’t is they actually do force a distribution. The median for PBSR tiers took years to move from Gold to just into Plat.
Many have asked why a win/loss system is simply not good enough, but I think I understand why. They also I am sure fear people “gaming the system” which could only be possible if there are certain player-controlled variables not obvious (like say elims/damage) that can be manipulated.
i thought the whole pont of role q was so they could match mmr role for role, but they still just go on team mmr where its comparing apples and oranges. makes role q pointless and infuriating since you cant switch to properly counter your true mirror.
I might agree with you on most things, but playing with a friend 1000 SR lower in comp is more like boosting, sure nobody really tries in QP, but even just myself, if for some reason I was playing even 500 SR lower than my current roles, it’s like playing DuckHunt level 1, things hardly move, but to others (new players or people that arent at my level) the game is moving very quickly, on this topic, I 100% agree with Eyevou, if you want to play with friends, do it in a casual mode.
Edit:
Just stop, the conspiracies are getting out of hand, stop twisting the devs words to fit your needs, you are wrong, give up. MMR wouldn’t be an issue at all if people only had one account and didn’t screw with the system, and you know it.
I find this all funny because people anchor that MMR / SR is really accurate. All gaming has been figured out - just like personality tests…
Can we get off the high horse and even admit that perhaps MMR / SR may not be 100%, so if we even attempt to make even teams lets at least ball park the PARTS of a team to be “as even” as possible with roles matched up.
Exactly this. There are several esports integrity advocacy groups on the rise, that are trying to force auditability and transparency in order to receive esports certification from a trusted authority.
No more “haha play our esport” with hidden rigging that secretely just tries to bin and rig around what hitbox sizes you can track, ignoring other factors and calibrating off of smurf corrupted rank data.
The industry just isn’t going to stand for it in the 2020s.
IT helps no question, but that doesn’t mean it can’t use extra tuning. Let’s not do a throw out the baby with the bathwater approach by raising the issue. RQ still provided a much needed structure to the game, and the majority support it even given Blizz own statistics on what people select.
The team average under RQ is still superior to one simple SR for a player because the level of play obviously now that we can see it is one or more tiers apart frequently between their mains and off characters.