Why Handicapping (MMR) is Wrong for Competitive Play

Competitive Discussion
Prev 1 25 26 27 50 Next
01/26/2018 02:43 PMPosted by Eclipse
So you agree in Handicaping in that a smurf can change the outcome of the match. What if the system knows who the high MMR player is and puts them on the enemy team while your team have low MMR DPS. The system can use this to calculate who is more likely to win or lose based on the DPS hidden skill (MMR).

If it thinks you belong higher then you get a streak of matches with the high MMR dps players. If it thinks you need to be lower then it can put those high MMR skilled DPS on the enemy team. Thats why the vast majority of players are stuck in what they call ELO hell.

I think a lot of these complaints come from Tank and Support mains who count on the DPS to get elims. The system can give you a good DPS player or a bad one and thus give you a greater chance at winning or losing which has nothing to do with how good a healer or tank you are.

I have seen matches where an entire team of high skilled MMR dps players won with no healers or tanks against an enemy team who did have healers and tanks.

I agree with you that a smurf can carry if he is a DPS smurf. I play Tanks and Healers. So if I am in a Diamond match and am healing good or tanking good then I can rely on the DPS on my team to be skilled enough to make it an even match.

In the last match there was a Widow in low Gold who got triple headshots in practically every kill cam. Clearly a smurf who was being pocketed by a shield tank and healer. If you went outside out of cover of shield or wall you got instantly killed and were unable to dodge it by random movements. Not to mention being shot at by the rest of the enemy team.

Those shield tanks and healers are winning because of that Widow carry. In Diamond though there would be an equally good Widow who can keep her in check. So now its a more balanced match. You won't see that in Gold. Either both Widows are terrible or one is carrying and the other is making us lose.

In Gold you can be the best healer or tank in the world but if your DPS are not getting elims and just dying constantly you will lose no matter how good a tank or healer you are.

Whenever I try to DPS I am forcing some other DPS to play healer or tank who does not know how to do those roles. I don't get heals and I don't get pocketed and the tanks are playing like DPS, out of position, feeding the enemy ULTs so they end up having one ULT after the next and we lose.

I think others know what I am talking about when a DPS main tries to play a healer or tank. Likewise, when a support main who never plays DPS tries to go McRee or Hanzo and gets killed too much thinking they will get healed but the DPS healer is just not as good as them in the healing department.


I don't know if the MMR system offsets smurfing, but it seems likely.

My mind reels every time I try to imagine how Blizzard is using the data they are gathering. But what blows my mind is that players accept the *explicit pretext* that they are gathering the data for the purpose of handicapping. In a "Competitive Game," as Principal Overwatch Designer Scott Mercer describes:
https://us.battle.net/forums/en/overwatch/topic/20749737390#post-3
01/26/2018 03:54 PMPosted by Cuthbert
01/26/2018 12:48 PMPosted by FriendlyFire
Stop lying about Scott Mercers post, it says nothing like that.


Does a single word of that post describe anything *other* than a handicapping system to you? What do you even think Scott is talking about there?


Yes, the whole thing. To be constructive though, I'll give a specific example.

The unfortunate truth is that there is not always a "perfect" match for you, especially at very high (and very low!) skill ratings where there's fewer players of similar skill.


It wouldn't be in issue to have "fewer players of similar skill" to match you with if they did what you claim, and threw high/low skill players on one team to balance each other out.

They don't do that, they try to put 12 people with the same skill in a game, so this becomes an issue if they can't find 11 other people with the same MMR. At that point, they find the best game they can, and appropriately adjust SR/MMR changes if the match was in fact unbalanced.

Most of the post simply talks about things separate from your conspiracy theory. Some of it contradicts your conspiracy theory. None of it supports it.
01/26/2018 04:06 PMPosted by FriendlyFire
They don't do that, they try to put 12 people with the same skill in a game


This post is the first I've read of any of this, so help me out here. If everyone in the game has similar MMR and similar SR... what's the point of the MMR, again?
01/26/2018 04:24 PMPosted by pootflute
This post is the first I've read of any of this, so help me out here. If everyone in the game has similar MMR and similar SR... what's the point of the MMR, again?


It let's them do stuff with players ranks (their SR) without messing with the rating used for match making (MMR). Like...

  • Decay in-active high ranked players
  • Adjust the distribution of ranks
  • Punish leavers (without making them smurf against worse players)
  • etc
  • SR is a lot like the tier icons. You wouldn't want to be matched based on tier right? Where if you go from 1990 to 2010, you go from all silver games to all gold games. That wouldn't make sense. The tiers you are in are a representation of your rank, not something used to make matches.

    SR is the exact same, just more precise.

    As always, to be clear, I'm not against visible MMR systems. I played Dota which had it, and it was fine. I also played CS:GO which had hidden MMR like Overwatch, and that was fine to. Personally, I don't care.

    If you want to say they should make MMR visible, you could make an argument. If you want to say MMR is hidden to rig games, you're delusional.
    For the record, I do not think it is rigged, or that there is forced 50%, or that there is a conspiracy. However I did not know this hidden MMR even existed and what OP describes makes a lot of sense. I've been suspicious of the matchmaking's effectiveness for a while and finding out that there is this totally obscured system really doesn't help.
    01/26/2018 04:47 PMPosted by pootflute
    I've been suspicious of the matchmaking's effectiveness for a while and finding out that there is this totally obscured system really doesn't help.


    Honestly, this is just ignorance. I'm not being mean saying your dumb, just the literal meaning of "not knowing". The system really isn't that complicated.

    You have your MMR. It goes up when you win, down when you lose, the exact amount influenced by a few factors. You are matched with 11 people who have the same MMR (or as close as the match maker can find). Your SR under almost all conditions simply follows your MMR. If you aren't leaving games or decaying (above 3000SR only), I don't think there is anything to make them separate.

    If you want all the details see this post (https://us.battle.net/forums/en/overwatch/topic/20761096463), but in summary it's pretty simple. You're put with people at your rating, and if you want to climb you gotta win more than you lose.

    A natural question is "why do so many people say they see the same thing?" The answer ...

    01/23/2018 12:51 PMPosted by alexfly
    People said the same things in league and cs go. People will always ignorantly claim that matchmaking systems are broken even though, ironically, every triple A title uses virtually the same elo based matchmaking system with minimal variance in implementation. You can search “forced 50/50” and all these other buzz words in league forums, people have been saying these things for a decade and will continue to do so.

    The reason stems from the fact that most people werent provided the necessary education to fully understand why matchmaking systems work the way they do. and while most peoples SR complaints can be swiftly nullified by an understanding of how statistics work, people tend to have a natural instinct to think they are better than they really are at tasks they do often. There is plenty of psychological research on this phenomenon.


    Also, the truth is it's a small minority that gets magnified on this type of forum. On reddit, these types posts are down voted into oblivion.

    Maybe you haven't played games with competitive ladders before. The fact that every game ever (despite many different ranking systems) has these same complaints is very telling. Even Dota, with visible MMR and flat gains/losses.

    If that doesn't convince you, consider why no one has posted a video of it happening. Why is it so easy to find videos of people climbing a bunch of rank, but impossible to get a video one person being help back by unfair teams?

    You sound like a reasonable person, so please don't get hood-winked by this nonsense. I guess you did say you don't think their is a conspiracy, but I don't know what being "suspicious" means specifically. As I've said before, anyone can feel free to add me in game (FriendlyFire#1121) we can chat about this and play some arcade games if you want. In 5 minutes I can easily tear apart everything "no-evidence-Cuthbert" has said.
    01/26/2018 04:47 PMPosted by pootflute
    For the record, I do not think it is rigged, or that there is forced 50%, or that there is a conspiracy. However I did not know this hidden MMR even existed and what OP describes makes a lot of sense. I've been suspicious of the matchmaking's effectiveness for a while and finding out that there is this totally obscured system really doesn't help.


    I feel you. I've always known a little about MMR, but I didn't concern myself with it when Overwatch was Quick Play only. It was only after the introduction of Competitive Play, and hearing that MMR was *included* in Competitive Play, that I started to give it some thought. It's quite a rabbit hole to go down.

    I agree that "conspiracy" isn't the right word for what Blizzard is doing, but there are some valid alternatives. "Deception" comes to mind. The fact that players like you and I can spend time in Competitive Play without even knowing that matches are handicapped shows that Blizzard wanted to keep us ignorant of that fact. That's why "Match Making Rating" is not visible or mentioned anywhere by name in Overwatch's user interface.
    01/26/2018 04:06 PMPosted by FriendlyFire
    They don't do that, they try to put 12 people with the same skill in a game, so this becomes an issue if they can't find 11 other people with the same MMR.


    You're ignoring the developer's *explicitly stated goal* of giving either team, in every match, a 50% chance to win. There's no way for them to do this if every player's MMR is just a single, equal number and teams are assigned randomly. It is *impossible* for Blizzard to be treating players equitably, based on what they have disclosed about handicapping/MMR.
    Yo, Cuthbert!

    I'm a three times master league Zenyatta main from the EU servers who is looking to improve as a coach/analysist by making vod reviews for anyone willing to improve!

    It's important to reflect on the games we play in a slowed down setting to truely figure out where we can improve/benefit the most to up that winpercentage bit by bit. Since 1 out of three games we will lose no matter what we do, the other 1 you will win no matter what you do thanks to teammates or enemy throwing, and then the last 1 out of the three are games where our own decision making actually can have a big impact on turning it into a win or loss. Anyway let's try to not solidify mistakes we have yet to spot in our plays by doing some reviewin 'n improvin!

    You can see previous made vodreviews on my youtube channel here;
    https://www.youtube.com/user/Gulzath/videos
    Or if you are interested in submitting a vod yourself feel free join my Discord here;
    https://discord.gg/rtRhSqH
    Hoping to help! Have a good one, and glgl hfhf my man!
    01/27/2018 07:20 AMPosted by Cuthbert
    01/26/2018 04:06 PMPosted by FriendlyFire
    They don't do that, they try to put 12 people with the same skill in a game, so this becomes an issue if they can't find 11 other people with the same MMR.


    You're ignoring the developer's *explicitly stated goal* of giving either team, in every match, a 50% chance to win. There's no way for them to do this if every player's MMR is just a single, equal number and teams are assigned randomly. It is *impossible* for Blizzard to be treating players equitably, based on what they have disclosed about handicapping/MMR.


    Scott Mercer's blue post that you keep referencing is a reply to someone who created a topic about matchmaking and premades, where the topic creator ended up having games where he was on a team of 6 solos vs 2 3-person groups, and also he and two other solos plus a 3-stack vs a team of 6 solos.

    What you keep trying to stretch Scott's reply into is some kind of idea that two players on the same team may have a visible SR of, say, 2300, but one of them could have an MMR of 2750 while his teammate could have an MMR of 1950 in order to make things "fair".

    MMR, according to Stylosa's conversations with Blizzard reps, is just a very large string of digits that isn't meant to be read by humans, and translates to your SR (not factoring in penalties/decay).
    https://youtu.be/1oXV5hsdYu8?t=185

    In a game with 6 solo players vs 6 solo players during busy hours, every single player will be approximately the same MMR, and thus, the same SR (assuming no decay that is). You'll only see MMR disparities happening at the absolute top or the absolute bottom of the rankings (something such as a 4650 SR Top 500 player that has a 3725 Masters teammate and a bunch of 3900-4000s vs 6 4100-players, and Blizzard's tweaks have tried to avoid that for a season now), or when there's a stack with the maximum difference in SR (1000 SR difference maximum Diamond and below).
    01/27/2018 07:20 AMPosted by Cuthbert
    You're ignoring the developer's *explicitly stated goal* of giving either team, in every match, a 50% chance to win.
    How is this so hard for you to understand?

    I mean, what do you think that means? Putting a GM player with a bronze player against another GM player and a bronze player?

    How do you not understand that this simply means putting all players of equal skill level into a game?

    I can almost guarantee that if you're on a 10 game win streak, there are players at that higher level. You're not all of a sudden the best player and being placed with terrible players because you won a few games. Use your brain man, ffs.
    I think both sides have their fair points, but recording and showing evidence of this theory would the best way to prove it. Do it.
    01/27/2018 08:30 AMPosted by Applesoup


    How do you not understand that this simply means putting all players of equal skill level into a game?


    It shouldn't DO that. that's the entire point. Short of placements and decayed masters and GMs the MM should just fin a group with similar *SR* and match them together.
    Instead it tries to be smart about it and creates "Tiers" of people of different MMR at the same SR.

    Here's an exsample, three people with the same SR:

    A: A player plays realy well over a fair few games but goes 50/50. His skill keeps him floating around and his MMR increases. Still the same SR. Game sees his MMR and matches him with people it thinks are his equal and on average they have higher SR than him.

    B: Some guy, average for his SR, goes 50/50. SR stays the same MMR stays the same.

    C: Some scrub lucks out and keeps his SR but his MMR decreases because the sytems thinks he sucks. SR is the same but he gets matched with people of his own MMR wich on average have LOWER SR than him.

    See? A is handicapped by being put into matches that are harder than they should be, C is getting easier matches than he should have.
    MMR is unfair. It has no place in a competitive system worthy of it's name. Let it stay in Quickplay where it belongs.

    Let competitive be competitive and just believe that people deserve their current SR and do matchmaking off that.
    01/27/2018 09:57 AMPosted by tawT
    01/27/2018 08:30 AMPosted by Applesoup


    How do you not understand that this simply means putting all players of equal skill level into a game?


    It shouldn't DO that. that's the entire point. Short of placements and decayed masters and GMs the MM should just fin a group with similar *SR* and match them together.
    Instead it tries to be smart about it and creates "Tiers" of people of different MMR at the same SR.

    Here's an exsample, three people with the same SR:

    A: A player plays realy well over a fair few games but goes 50/50. His skill keeps him floating around and his MMR increases. Still the same SR. Game sees his MMR and matches him with people it thinks are his equal and on average they have higher SR than him.

    B: Some guy, average for his SR, goes 50/50. SR stays the same MMR stays the same.

    C: Some scrub lucks out and keeps his SR but his MMR decreases because the sytems thinks he sucks. SR is the same but he gets matched with people of his own MMR wich on average have LOWER SR than him.

    See? A is handicapped by being put into matches that are harder than they should be, C is getting easier matches than he should have.
    MMR is unfair. It has no place in a competitive system worthy of it's name. Let it stay in Quickplay where it belongs.

    Let competitive be competitive and just believe that people deserve their current SR and do matchmaking off that.


    You're wrongly going with the assumption that MMRs can vary wildly from the same SR.
    This has been stated directly by the devs
    01/27/2018 07:26 AMPosted by Kozan
    Yo, Cuthbert!

    I'm a three times master league Zenyatta main from the EU servers who is looking to improve as a coach/analysist by making vod reviews for anyone willing to improve!

    It's important to reflect on the games we play in a slowed down setting to truely figure out where we can improve/benefit the most to up that winpercentage bit by bit. Since 1 out of three games we will lose no matter what we do, the other 1 you will win no matter what you do thanks to teammates or enemy throwing, and then the last 1 out of the three are games where our own decision making actually can have a big impact on turning it into a win or loss. Anyway let's try to not solidify mistakes we have yet to spot in our plays by doing some reviewin 'n improvin!

    You can see previous made vodreviews on my youtube channel here;
    https://www.youtube.com/user/Gulzath/videos
    Or if you are interested in submitting a vod yourself feel free join my Discord here;
    https://discord.gg/rtRhSqH
    Hoping to help! Have a good one, and glgl hfhf my man!


    I didn't ask for your help. That's not what this thread is about, so I rather suspect you're making that offer to imply your superiority and get my goat.
    01/27/2018 09:57 AMPosted by tawT
    01/27/2018 08:30 AMPosted by Applesoup


    How do you not understand that this simply means putting all players of equal skill level into a game?


    It shouldn't DO that. that's the entire point. Short of placements and decayed masters and GMs the MM should just fin a group with similar *SR* and match them together.
    Instead it tries to be smart about it and creates "Tiers" of people of different MMR at the same SR.

    Here's an exsample, three people with the same SR:

    A: A player plays realy well over a fair few games but goes 50/50. His skill keeps him floating around and his MMR increases. Still the same SR. Game sees his MMR and matches him with people it thinks are his equal and on average they have higher SR than him.

    B: Some guy, average for his SR, goes 50/50. SR stays the same MMR stays the same.

    C: Some scrub lucks out and keeps his SR but his MMR decreases because the sytems thinks he sucks. SR is the same but he gets matched with people of his own MMR wich on average have LOWER SR than him.

    See? A is handicapped by being put into matches that are harder than they should be, C is getting easier matches than he should have.
    MMR is unfair. It has no place in a competitive system worthy of it's name. Let it stay in Quickplay where it belongs.

    Let competitive be competitive and just believe that people deserve their current SR and do matchmaking off that.


    You're not realizing that in A and C the guy will rise in SR or fall in SR to match. If he is performing as expected at the MMR different than his SR then the system is designed such that his SR will catch up.
    01/27/2018 09:57 AMPosted by tawT
    01/27/2018 08:30 AMPosted by Applesoup


    How do you not understand that this simply means putting all players of equal skill level into a game?


    It shouldn't DO that.
    You're right. How could I have been so dumb. A matchmaker putting players of equal skill into games? Absurd.

    C: Some scrub lucks out and keeps his SR but his MMR decreases because the sytems thinks he sucks. SR is the same but he gets matched with people of his own MMR wich on average have LOWER SR than him.

    C is getting easier matches than he should have.
    I don't understand how you've come to this conclusion?

    If I understand correctly...player C has lowered his MMR and is now playing against other opponents with lower MMR. How are these games easier than they should be? I'm confused.
    01/05/2018 02:08 AMPosted by kaizoku222
    It doesn't matter what blue posts are alluding to, it doesn't matter what the discussion around how the system is *suppose* to work is. All that matters is that in practice the better you perform the more the system expects you to carry players below the average for your SR tier.

    So far there has not been a season where I get matched with better and better players as I go up in SR and/or winrate. Instead the matches get more and more stacked, the people that use voice, play well, pick around the situation/team disappear. They are replaces by people who, at high plat/low diamond, still pick a 3rd DPS as Hanzo and refuse to swap, or a Mei on offense that dives/solo ults, or a Torb that has no mic, doesn't communicate, and refuses to switch when asked by the whole team.

    I decided to play Moira whenever I could this season, so I could see how my SR would react to playing a new char and getting better over time with them. I know my play is getting better and better, but my SR doesn't at all reflect that. I had a 70%+ winrate about 7 games after placement, and my placement was 9w 1l with the loss being a crash, placed about 2850, got 2 games away from diamond (~2950), then I hit a loss streak that took me all the way down to 2600 while playing *better and better* with the character I'm using. The difference is the quality of players and the balance of matches I was getting. Several times during the loss streak I was getting matched against teams of 3 with level 600-1200 people with pubs on my team all sub level 500.

    This system doesn't match you with better and better players to test you, it matches you with worse and worse examples of your tier to try to make you carry to "prove" you can escape your tier. It's not about learning the game with good players and learning more advanced strategies while playing your character better. It's about learning to herd cats and negate toxicity as best you can.


    Sorry I missed this post. This is a great summary of what I'm trying to describe, glad it's on page 1 of the thread.
    Posted by Applesoup:

    How could I have been so dumb. A matchmaker putting players of equal skill into games? Absurd.

    (sarcasm)


    You're missing the point. The matchmaking isn't finding 12 players of "equal skill." It is hard to quantify skill. Every player in every match has a different level of skill, with different hero characters and classes. MMR monitors performance in every category.

    The handicapping/Match Making Rating system uses performance metrics, like levels of damage and healing, to figure out roughly how good people are. And then it arranges teams however it wants to give either team a 50% chance of winning.

    This arrangement typically singles out the best players, and puts them on teams with the worst players. If MMR knows you to be the best player in the game, welcome to the worst team. The second, third, and fourth best players in the game have all been placed to oppose you.

    You may still win the match, but you should recognize that the odds are *artificially* stacked against you. And those effects are only compounded when you group with other good players, so you get longer wait times and even worse teammates.

    Join the Conversation

    Return to Forum