Plenty of folks have written that bad match making / potatoes happens on every rank.
Wouldn’t that be a flag that this system is inherently flawed in some way? Shouldn’t the higher you go in elo the more accurate the match making be?
The only counter argument is “people play inconsistent”. But again the higher your rank should filter out the least consistent players, yet “there’s bad matches every rank”.
Can someone explain this phenomenon?
because MMR is only a number based on your win/lose ratio. It doesn’t filter anything; it’s prehistoric. And some people are good at one hero, bad at the others, which the game doesn’t care about, so a plat soldier might very well be a bronze Junkrat for sure 
You will never know.
All we can do is keep complaining and hope for a change.
Playing tank is strange. The tank has to kind of play to the team and it creates inconsistent play. Like, depending on how your team is playing you kinda have to switch up how you play. I play an aggressive tank, Im either forcing the enemy tank out of the space theyre trying to take, or if I see an opening ill go after supports/dps and force their tank to pay attention to me and waste resources to try and protect their team.
Some teams, simply dont follow up and play more of a poke style. The two just dont work together. So when im playing how I play, I tend to do much better. If I have to adapt to a slower poke/defend less aggressive style I simply wont play as well and teams will think im over extending because they dont share the aggression.
I see it on support too. You have the aggressive tanks that make space and push. then you have the more defensive brick wall tanks that ride payload or camp point. Its two different playstyles really. If you have some aggro players with some poke players the games just gonna feel bad.
Why is every rank prone to inconstancy…
I know you’re smarter than this. So this must be a troll thread looking bait people.
Humans play the game. Therefore game and ranks naturally feel inconsistent.
Because skill is relative. A gm player can struggle against some gm players sometimes, but not all of them.
This can be due just to hero matchups, or some people just have a different playstyle that catches people offguard around in same elo.
But that gm player will not struggle against plat players.
Or sometimes its just due to players in the same team not synergizing.
Then, people have good days and bad days. Sometimes people have slept badly, had too much coffee, eaten badly etc.
People are not robots, not even the pros.
Another way to look at it, its not rare to see, when you run into same players over a long time that:
Player A beats player B, player B beats player C and then player C beats player A.
However, they all have the roughly same rating. This same can then apply to teams.
tldr: mmr is relative to the playerbase. it is not an objective measure.
is it not? The biggest rebuttal out of your camp is people play inconsistent. Unlike chess elo, the higher the ranks the less “inconsistent” and stomps / bad matches there should be.
However… in the same breath people will say stomps and bad matches are at EVERY RANK?
So if MMR can filter out “winning” factor, are there more or less stomps in higher ranks? Mathingly it should be less, but it actually could be more…
Honestly on my ult account, I am seeing some dousey’s of games mostly on my hanzo account because its increasing my overall account win percentage. Oddly enough I constantly get terrible players right before I get to diamond. Like actual terrible.
After playing this last couple days I may take a step back and say EOMM is still in full force. Sure MMR is SBMM, sure the fundamentals are there. But the team balancer which we know nothing about nor is it talked about, maybe the EOMM link we’re missing.
Actual GARBAGE matches. Like so bad, it makes me want to buy a battle pass in MR for spite.
Because you are the only one who thinks there’s a separate entity called that in context of overwatch matchmaking.
Balancing teams is part of matchmaking process. And we do know how they do that, via the role delta system/matching nowadays. Where players are effectively ‘mirrored’ by their ratings, per role on teams.
You said we dont know anything about it, and thats simply because there is no such thing to describe separately.
It is however very easy to just conjure random stuff out of thin air and then say “well this thing XYZ i just imagined might be doing something weird”.
But thats not how things work.
People in high ranks are generally better at exploiting mistakes, etc. So, high rank games can easily become more ‘decisive’ sometimes, unlike low rank games where people just … do things.
You mean, anyone who’s thinking before they speak.
You can say that. But it would be incorrect as EOMM isn’t used in OW.
They’ll happen from time to time in any PvP game, at any rank.
Skill does not equal performance. Sometimes people just won’t click, or will have an absolute flapper of game.
Just look at pros in any sport. Some times matches are rubbish, boring and sometimes the best players just have an off day.
I’ve address this so many times before. I provided evidence that the dude who created this ow system mentioned before. I did not create the term / system of “team balancer”.
Gdc 2016 - Josh menke on match making systems. Goes into details of “mmr”, sr and happened to work at blizzard at the time.
He goes into how mmr is used but he gave existence to another system called a “team balancer“. Perhaps I’ll even look at it again to remind myself.
https://youtu.be/-pglxege-gU?si=AjzXbRDtGLGm2r7B
I’ll even add this dude did ANOTHER video I didn’t even watch - same conference few years later. I’ll be checking that out later.
So I’ll add a binary choice. Either Josh made a “team balancer” and they use it in ow. Or Josh referenced a “team balancer” and they don’t use it. Either way - if they don’t use one, maybe it’s time they use one, if they do use one, they need to figure it out more, it’s only been 7 years.
Yes the age old inconsistency argument:
Why is it I consistently get the bad day players around my sr peak, and consistently get the good day players a few games below. And this still doesn’t answer the question, if Mmr is working consistently - why are there stomps in every rank?
Think logically about what a “team balancer” might do.
Surely, logically, it would balance the teams. Thus making the MMR matching tighter.
It is probably the system they use to role match, and to ensure where there are stacks, they are match appropriately.
Not sure this is some “gotcha” for your conspiracy. If anything, it’s the complete opposite.
it would be more consistent if everyone was playing in the same stack every single time.
I play much better with a dive tank diving with me than with a full Team of scared poke players that don’t want to get close to the enemy.
Team A has a kiriko and a zen that kill more than the dps, so all I have to do to win is not charge in like a headless chicken. Next game I have a mercy lw duo and if I don’t pop off, I simply lose that match.
I got rolled twice in a row by this stack playing dive. Had no complaints, they were so well coordinated. It was impressive to watch.
And given that we know how matchmaking works already, including how it “balances teams”, what would you imagine this ‘team balancer’ would do beyond that?
What ow does now is more refined regarding balancing teams, than back in 2016. Because ow back then did not balance teams based on players mmr on roles.
Besides that, i am not sure if even Josh mentions team balancer on this video. Feel free to provide timestamp so we’ll get the context.
And that video is also seemingly about matchmaking systems in general, not specifically about how overwatchs one exactly worked back then.
You dont. You’re just anxious about ranking up and play worse, and blame your teammates immediately you see someone doing a mistake.
Because stomps can happen even with ‘perfectly matched’ teams with mmr.
I dont know why you think they couldnt happen.
Its not just about inconsistency, its also about player synergies and everything else as well.
The earlier mentioned flow: player A beats B beats C beats A , is one thing that also can cause things like this.
mmr is just a single number, and the predictions based on that are only about win chances. It does also not predict how a given match will play out.
A 50/50 predicted match can well be a ‘blow out’ for one team when played once. Put the same teams against each other 10 times, it could be blow out here and there and both teams still won 50% of games.
This part is wrong and shows a clear error in your thinking.
Matchmaking is more accurate the more players there are who are at your rank. This means that Gold/Plat will have much better matchmaking than GM/Champ. It also means that Bronze matchmaking will be much worse than Gold/Plat matchmaking.*
However, Gold/Plat players will play very inconsistently. For this reason, matchmaking can look screwy at all ranks. You could get matchmaking to look better. You’d need to somehow get major buy-in from the players (that is, everyone would need to affirm that they were going to try their best to win at all times, to do VOD analysis of their matches, etc and players who did not would no longer be allowed to play the game) and you’d need to reduce the complexity of the game.
If we want to allow random humans to play a complex game, however, we’ll just have to put up with matchmaking that can look screwy.
*If you want to get a sense of what matchmaking looks like to the players who are best at the game- the high GM and Champ players- imagine matches involving players of Bronze through Plat rankings. The range of skill in GM/Champ is simply much wider than the range of skill in Gold, for instance. And there’s not a damn thing anyone can do about that, it’s just the nature of bell curves.
Yea every single person who is just not playing well that match has 0 excuse and should either be banned or just never allowed to play or participate in anything. If you aren’t going to deliver 100,000,000% and carry every single match, then you shouldn’t be playing this VIDEO GAME that has 0 meaning outside of the few minute match. Great argument sgt let’s see how that works out.
I’m assuming that you are willing to be the first one to recieve whatever repercussions there will be for all of the matches you ever played inconsistently?
Not that deep dude. Some people have an off match. And it really doesn’t matter because it’s a video game lmao
I get decisive… I dont get stomps, which consistently happen, “can look screwy at all ranks”.
So “team balancer” was once made up by me, but now I have proof that I didnt make it up… LOL its not gotcha, its just that I am right - that it “exists”. Again if OW has it or not, it wasnt made up by me is all my point.
a tank that has 12-16 deaths while I have 5-8. is more than ONE mistake. My dps partner going 0-8 on defense is more than ONE mistake. This all happens when I am at my SR peak… its just very sus.
This is exactly the point. This is the gotcha. If stomps can happen outside of the perfect match, then the system needs work.
Tell that to the Match maker is fine camp. All I do is highlight that the system needs work and its blasphemy.
Two things:
So one, sounds like your agreeing, but not for the reasons I highlight. Stomps happen at all ranks. Higher ranks cause its less players, lower ranks its a player problem. But… if higher ranks had more people, wouldnt it still be a player problem? Why even highlight the low population in high ranks?
The devs admitted stomps happen even if “accurate” “on paper” matches occur, at what point do we actually stop blaming the players? Can we see if stomps occur more at night vs during the day or prime time?
Two, to tackle the actual rebuttal: How can we actually prove players play inconsistently. Isnt that an assumption surrounded in facts but not a fact itself? There has to be a “consistent” skill somewhere lurking right? Wouldnt it be ideal to find that for match making and then build around it?
The “inconsistency” argument is ironically inconsistent itself.
Let’s assume the shills are right about ranks being accurate. I have heard them say things like 500 elo is a massive difference in skill. So let’s assume that is true. Therefore is it fair to say that 500 elo is an acceptable range of skill fluctuation?
My current tank SR for example is diamond 5. Which it has been pretty consistently around there for many seasons. So is it fair to say that when I play my best, I play like a Diamond 3? And when I play my worst, I play like a Plat 3? Because that would be a 500 SR gap accounting for all of my possible inconsistency. Yet I often see other players in my games who play nowhere near that range.
Therefore either the inconsistency argument is invalid, or the ranks are not accurate. It can’t be both. The eye test would suggest the latter, as I often see players making bronze level decisions in my plat and diamond games.
I think inconsistency has more to do with things like aim, playstyle, map selection. But I can’t defend things like teammates not using their ults during last fight. Or somehow losing a 4v2 after I get 3 opening picks. Which both seem to happen far more often than I care to admit.
When things such as this occur it is difficult to take the game seriously. I get that mistakes happen in abundance at all levels. But sometimes you can just tell when a game is going to be unwinnable within the first minute.
All I do is complain about losing matches and blame anyone or anything other than myself and if someone disagrees they’re a shill or blaspheming me”
Fixed it for you (couldn’t figure out how to simulate the tears on your keyboard though)
Either way, boohoo game isn’t perfect. Show one that is and that the community doesn’t complain about the mm.
Still hilarious how you think the mm is responsible for how well someone will do from match to match. Especially with hero locks where you have a good chance of not playing your best heros.
I think the very first match you don’t play super well in at whatever rank you are at, your account should automatically be tanked down to the lowest rank possible. Because by your logic and argument players don’t have “off matches” and it is the mms fault for putting you together. So the first time you don’t play to whatever level you’re at, your account should be tanked in oreder to match. That’s what you’re advocating so I think you should be the first to volunteer testing it. Wouldn’t be too big of a drop for you anyway right
It’s not really an assumption. If it’s based on wins and losses, the rank is accurate for that moment in time on that account.
Theres very little to debate there. The issue is, people see rank = mechanical skill which is just nonsense in a game like OW. Whilst it’s a big part of you’ve got aim but no brain, you aren’t going to go far.
Remember. Your rank isn’t about how well you aim. It’s about how well you adapt to the challenge and win. Meaning, you can quite feasibly have quite significant differences in rank if you many accounts. Because the challenge to win is different in every match you play.
In order to do this, you’d need to do the two things I mentioned earlier: get extreme buy-in from all the players and eject players who don’t do what’s required to develop that degree of consistency and severely decrease the complexity of the game.
Without doing those two things, there’s no real way to ensure consistent play from the general populace. And even then, you’d just be decreasing the inconsistency in players’ skill.
What you are asking for would be much more possible if we reduced the game to 1v1s on a single map with a single hero- someone like Soldier would be ideal.
Ask yourself the following: do professional athletes being paid millions of dollars a year perform consistently, such that it is generally possible to determine the outcome of matches ahead of time?
If not, why not? Then you’ll have your answer for why a bunch of amateur randos who don’t train together, who may never have played together, in an asymmetric hero shooter might not perform consistently.
I highlighted the low population in high ranks, because it demonstrates that your intuition about more consistent matchmaking in high ranks is backwards. Your thinking about matchmaking had a fatal flaw in it, which becomes readily apparent once one considers the fact that high ladder ranks are sparsely populated. If you want to improve your understanding of matchmaking it’s an important aspect to recognize and consider.