MR adding performance mmr

General Discussion
42

There is this mechanic called the “chronoshield” that boosts people out of gold by making one out of every 5 losses not count. It disappears in Platinum. And then there IS performance based SR in that game. They are just buffing it for next season.

Keep in mind that GM3 in that game is basically just diamond in Overwatch. So, boosting people out of gold in that game is similar to boosting people out of bronze in this game. It’s not a huge deal to anyone decent at the game whatsoever. People are just coping about inflation which honestly makes no sense to begin with because if they inflated everyone, there is technically no inflation.

So, it basically boils down to people getting mad that you get a little icon despite not being top 3% at the game. Lol

1 Like
43

Problem with performance based SR gain/loss is that unlike other games like Valorant/Counterstrike 2, your own stats can end up looking bad due to poor performance of your teammates.

For example as a support, it will be incredibly hard to get high healing per ten if your tank and DPS keeps dying over and over again, giving you little to no uptime in healing. Even maintaining low death rate can be almost impossible on some maps if your whole team dies. And with many denial abilities in the game, getting a lot of elims can be hard too if you don’t have a decent team.

Instead, I feel like the game should give you more SR gain upon win based on the stats of your hero compared to others in the same elo. On average, that is, over the course of dozen or hundreds of games. Not from individual matches.

For example, lets say that the average healing per 10 mins on Juno is 8K in Gold. If you consistently heal over 8K as Juno and you’re in Silver, perhaps you should be rewarded with additional SR gain upon a win.

For a loss, I think it should stay the same. Since you don’t want people to farm stats so that they don’t lose as much SR instead of trying to win the game.

But they do need to do something about the accuracy of Reversal and Uphill Battle modifier though. Because a lot of the times, the opposite happens where I get put in lobby where my whole team gets rolled only to receive a Reversal. And then I get put in a lobby where it’s essentially a free win and I get given Uphill Battle.

1 Like
44

Fixed.
Lets not pretend that you can be an amazing player with trash stats.

Stats are not the be all and end all of skill reflection, but they do offer a pretty accurate picture most of the time. Denying that is denying reality.

And talking about reality, this change means nothing for most players, but some players will have a rude awakening about their supposed “high” performance.

3 Likes
45

So you want Master widows in Plat lobbies because there matched off performance not wins

And you want Gold sombra in diamond games because they minmax performance but not winning.

Because that’s what performance mmr does and it makes the game unfair like you can say crap like this…

So you want to rewarded for a good play… Keep doing them and you’ll climb in the current system but if you want more remember if you make plays that win games but are see as bad by the mmr system you’ll get less sr. Distract both heals as tracer but never get the elims if that wins the game game you deserve your sr.

Lastly you rant about bad teammates holding you back… On average you’ll play against worse players if you better than your rank. Bad players don’t hold you bad they are why you climb.

2 Likes
46

They had it before. It boosts account inappropriately and promotes stat padding.

So no. Bad idea.

Also, heroes, roles, maps, modes… mean there is no statistical way to define “performance”. The best performance metric is simply, did you win.

2 Likes
47

No it didn’t. They claimed it did but they were lying. I can think of countless times where I hard carried the team, had best stats in the lobby, but still lost close to max SR

1 Like
48

I wouldn’t even give it that much credit. Legit gm in rivals looks like ow2 gold at best.

It wasnt about your performance on your team, it was sr gains based on that hero and being above or below the average performance for that hero in your region. so a one trick on a hero that wasn’t played much, they were able to climb faster and easier.

it used to be a sticking point during ow1 for a lot of us, but it made boosting accounts and selling them easy. There were a lot of changes to the system over the years though. pbsr was just one of the aspects of it for some of it.

2 Likes
49

PBSR existed in OW1.

initially for everyone, but for most of the life cycle it was just accounts under 3000sr that got it.

Above that, it functioned very much like it does now. With modifiers based on your rank relative to the lobby.

2 Likes
50

If you can kill it as Reed right now then I’d be pretty scared to run into you in the future. I have no idea whether this next set of buffs is going to improve his win rate, but it’s clear that they’re going to keep buffing him until he ends up in a good spot generally.

1 Like
51

He’s basically Reinhardt. Buddha was born to play him.

2 Likes
52

Huh, I hadn’t made that connection myself - but the most I’ve played Reed is in Clone Rumble, and I only played enough of that mode to get the purple coins before getting the heck out of there. I’ll take your word for it.

I figured that Doc was the closest to Rein, what with the nature of his barrier and the LMB → whip combo into Maelstrom shenanigans encouraging a bit of close-up combat. But I only have the most basic Doc you’ve ever seen while virtually never playing Rein so I’m not pretending to have any sort of insight here.

53

His arms are Rein’s hammer basically. If you “swing” them you hit multiple targets. Really outside of that he isn’t anything like Rein. :stuck_out_tongue_winking_eye:

2 Likes
54

Your missing the point entirely

It doesnt matter. With performance based ranking matchmaking does not need to be good to create a fun experience

No it doesnt lol

Winning games does not mean you are good

Performing well means you are good. Performing well does not mean you will win. You can perform well and still lose many of your games. Which is the problem

The game should be about rewarding player skill not rewarding who wins more. Whether you win or lose is entirely out of your control. How you perform is entirely in your control. Rank should be based on how you perform since that is the only thing you can control as an individual

Its not rocket science.

Its hilarious ppl saying it cant work WHEN OTHER GAMES DO IT AND IT WORKS AMAZING

Marvel rivals literally already has performance mmr and playing ranked feels so much more enjoyable than ow. The matchmaking is just as bad but even if i get some troll on the team as long as i play my part i barely drop from a loss. If you carry the game your rank shoots up.

That is fun. That is rewarding

NOT sweating your butt to win a game whilst your ball 1 trick refuses to swap and gets hard countered every fight. Leading to you losing 24% rank for something you could do nothing about. In marvel rivals if you are in that scenario you only lose 10% cus the game can blatantly see you werent the issue

Crazy how a 3 month old game is apparently more advanced than a 9yr old title

2 Likes
55

Not individual games, no.

Winning consistantly does.

What is “good” if not your ability to work towards the goal of the game?
Or in other words: win games.

Getting many kills is not “good” if you had to do something else to win.
Getting much healing is not “good” if it’s at the wrong time.

In which games does it work “amazingly”, which are as complex as Overwatch.

From what I’ve heard, ranks are all over the place in MR, with ranked integrity being non-existant.

2 Likes
56

Performance can equal towards skills, stats? No.

A otp reaper solo ulting repeatedly when his entire team gets wiped and kill 3 foes on the point. Will lead the team to lose. Yet he could, based on stats, lose less than its peers.

In general winning more than losing should do the trick towards what rank folks should deserve. Although due mmr discrepancies in matches that can lead to advantageous or disavantageous matches, that can lead to more or less mmr changes based on your mmr towards the match value.

If the game was deathmatch based, would make sense. Due kills actually matters. On ow matches, kills aren’t that important. Although can be, normally, the easiest way to control the objective.

You can win without killing anyone by simply pushing a payload or taking the point while the other team is not looking. Which is what pretty much C9 is.

I would understand solo hero queue on a deathmatch having performance metric. But on a teambased game that leans on objectives, not kills. Would be irrelevant.

I understand why marvels adopt it, tho. Because also OW1 adopted their current mechannism in the past, by providing more gains in wins than losing on losses.

Marvel rivals paired with, usually, more gains than loses. This system can offset that problem on higher brackets.

On OW, due it’s age, there are some really defined mmr points already. Paired with the flawed RQ system. Performance metrics only actually happens if were open due that could gauge ability to the player to perform in a match based on the matchup and map.

Due:
Specific maps favors specific heroes, specific parts of map favors specific heroes, specific metrics like (deaths prevented, players saves, assists, objective time, how many health packs picked or hacked and so on) can provide some info on performance paired with the adaptability and performance on each part of the match. But again, not raw numbers or stats, but instead cross reference favored heroes for specific map portions, kinds of plays and so on.

Performance metrics is way more nuanced in a game with objectives, which is why on marvel makes sense due how their objectives works.

Overtime limits, auto move payloads and so on. The game tries to make folks care less about objectives and more about deathmatchs. Which is not exactly what OW aims for. Push, hybrid and control points requires effort to be on the point to win. Marvel just requires “conquest it” briefly.

Fun aspect? Well that can be solved by splitting the playerbase towards what they should be playing, instead of the clown fiesta of casuals and try hards on pretty much all modes due it’s similarities.

1 Like
57

It’s dispiriting to see SwankyOrc refusing to engage or address the arguments some have presented here. I guess acknowledging them would mean admitting PBSR in Overwatch is flawed and it’s why it was removed in favour of the superior system we use now.

58

Overwatch originally had this and it kind of didn’t work. It often led to people padding their stats each game rather than doing the objective. So killing people became a higher priority than trying to win. Because if you have big stats you either get big gains or minimal losses depending if you win or lose.

59

No it doesnt lol

You can queue up and win 10 games just from luck alone. Many ppl do it every day.

Winning games does not equal skill. Good performance is what reflects player skill. Wins and losses are just the outcome of the match, not how you got there

Not the fault of the reaper at all lol. If the reaper killed 3, what were the other 4 ppl doing every fight? Reaper is literally the only one doing anything while the rest of the team chew tarmac

Also, how do the 4 people respawning lose the ensuing 4v2 when they get back. The 3 enemies are respawning with reaper so how are the reapers 4 teammates failing to capitalise on the reapers ult.

That literally is an example of why performance based sr matters. If im that reaper I am beyond steamed that every fight im killing 3 but my 4 teammates cannot handle the remaining 2 and we lose cus of it. Why should I as that reaper lose that game and full sr when I am getting 3:1 K:D every fight while my team are literally just existing all game

A perfect demonstration of why we NEED performance based rank. Congrats on proving my point completely

2 Likes
60

If you win 10 games on luck alone you will likely lose a lot in the next 10 matches until you get back to where your ability can carry you again.

10 matches is too inconsequential overall.

1 Like
61

10 matches can be an entire week of games lol. Saying someones entire week is inconsequential is exactly why the ranked system is a failure in ow lol

Every match should be consequential otherwise whats the point in playing…

No one wants to play ranked because nothing you do as an individual matters. People are sick of being stomped or having to deal with bad teammates that lose you games from spawn

Anyone who argues against performance based mmr i ask you, how are you intending to address the issues ruining comp rn? The poor matchmaking. The stomps. The mismatched teams?

Its easy to shoot down ppls ideas when you dont provide any solutions yourselves.

Heres some facts for you.

Matchmaking can never be fixed.
You can never make equal matches.
Stomps will always happen

So your solution is to do nothing and accept the game is a dumpster fire??? Really productive. At least im providing solutions to the bigger issues of the game

1 Like