Mmr in overwatch

I’ve been experimenting this a bit. Your best bet is that the algorithm is Account/Person based i.e. sacrificing one role to losses will give other role some extra benefits. Probs they will make it role based at some point.

1 Like

My duo partner and I

Got any replay codes? Curious what that might actually look like.

1 Like

It’s funny because replay codes are probably the number one reason why this theory is thrown out. Of course if you watch us, you will see mistakes. GMs make mistakes. (Crazy right). The issue is, how does the rest of the team compare to these mistakes…

Like someone else previously said, just look at your own perception of your experience with the matchmaking and go from there

There’s also the elephant in the room that I haven’t mentioned yet. If this theory is true, would it not directly help struggling players. I mean, you do bad as a player, and the game rewards you with teammates that carry, therefore, you keep playing. Seems like that’s something that would benefit the company that owns the game but what do I know =p

2 Likes

Of course GMs make mistakes, but their mistakes are NOWHERE near anybody below them. Also you can see their intentions and approach in their games is very clear. They might not always be able to execute it correctly.

3 Likes

Found one of your problems right here.

1 Like

So, the thing to understand about “Competitive Overwatch” is that it is algorithmically handicapped with MMR. Sound like a contradiction of terms? It is. More info:

2 Likes

“proof that my theory is wrong is probably the number one reason why this theory is thrown out.”

3 Likes

There’s not enough players at the moment to make fair games so mmr has basically pooped the bed leading to drastically lopsided and terribly unfun experiences.

5 Likes

I do find it interesting how you immediately go on defense considering we’ve rarely (if ever) interacted with each other.

I think you fail to truly understand how minor/major mistakes compound over the entirety of a game (or multiple if its a consistent mistake you’re unaware of).

:face_with_monocle:

I mean, I could do that, but then I’d just be reverting to the “two party system” as the above user mentioned (in that exact comment btw).

Clearly you’re only looking for confirmation bias.

Good luck on your games

1 Like

This is true. I will say if you have a nuanced discussion, and don’t break into name calling you can have a discussion with a lot of people here. They won’t call you names, but they will try and rebut you. I personally think Rhain, and Abdullah are excellent in discussions overall. Taleswapper is good too. With that said… you won’t change anybody’s mind on your thoughts. Try as we might. So it is better if you want an actual discussion, with no hopes of getting anybody to see your way.

3 Likes

I apologize if I came off being defensive. Like I said, it was many games we were carrying. Going through one replay would just distract from the point of my post.

This is the reason why I said watching replays is why people dismiss this. You won’t be looking at the big picture and instead find all sorts of little mistakes the person did without realizing his team is making collosal mistakes.

All I can say is that as time went on, our team would get drastically worse and the enemy team would have seemingly pro players on theres. I would spend time observing and analyzing both teams. I did this for awhile before posting here.

So I’ve actually been reading a lot about this subject. There were more topics about it than I realized.

I’m going to test out my findings and see the difference. It’s actually pretty cool because I know now not to put too much pressure on myself to do well. It’s better to just go with the flow of the team I think (good or bad)

1 Like

All good I guess.

I mean… I’ve watched enough gameplay over the years, it wouldn’t matter. And I used to coach Silver/Gold/Platinum scrim teams.

Depends on the reason I’m choosing watching the replay.

Am I analyzing YOUR gameplay and looking for the turning points where your decisions failed to win the game?

  • There are always at least 2 points where an individual could have halted the enemy’s push and completely changed the games momentum.

Or am I watching the game to determine the overall contributing factors?

  • Which can always be accounted for/made up by an individual player (up until at least 3.7-4000 SR).

:face_with_monocle:

You’re desperately exaggerating.

Shouldn’t you be busy focusing on the game?

Make sure you’re investing equal time on either side of the argument. Because if all you do is seek out confirmation bias (exactly how your previous response was worded), I’ll just tell you exactly what the issue is and it’ll perpetuate the endless and unnecessary debate.

If your goal is to climb, its pretty clear you’re headed down the wrong path (in regards to mentality). If your goal is to enjoy the game as it stands, then you’re free to do as you see fit.

  • I just don’t appreciate people coming here with some kind of moral superiority and bashing players who genuinely just play the game for fun.

  • Its toxic and detrimental to the gaming community.

  • People who play the game for fun, generally make up Silver/Gold/Platinum ranks.

  • While you’re conducting your research, do look up Dunning Kruger effect.

2 Likes

Well, there are two schools of thought on this subject.

  1. Game the system like you are implying.
    or
  2. Put together a 6 stack and actually enjoy playing overwatch.

XD hahaha

Even if you lose you at least know everyone is trying and is willing to coordinate.


and here is the other problem…League of Legends:

  1. More Champions
  2. Better Champions
  3. Ability to customize your champion to your preferred play style.
  4. Better matchmaker
  5. 5 to 30 second queues. (Larger playerbase)
  6. Better gameplay
  7. Less pay to win (don’t need the highest fps monitor, lightest mouse, most alt accounts…etc)

So how does overwatch fix this?

  1. Stop reworking heroes, actually release more.

  2. Honestly, the easiest way to do this would be to buy paladins or hire their hero designers. A third and cheaper option would be to have community competitions on hero design. There is nothing a player base likes more than knowing that their feedback is being heard and implemented.

  3. This would be a fundamental change (Stop hero swapping during matches and add an item/card/hero/ruin/customization system like league) (Players won’t cry about hero swapping anymore which reduces toxicity (If players feel bad while playing your game they will eventually not want to play it)

  4. More players per skill bracket, Faster balancing win streak and loss streak SR Boosts 100-500 SR plus or minus after 3-5 wins or 3-5 losses (Move the smurfs out and the throwers out faster) Boosts would stop at the highest bracket or lowest bracket, add separate hidden queue for leavers and backfill system.

  5. QUEUE TIMES ARE THE LIFEBLOOD OF AN FPS GAME ((longer queue times = declining player base = dead game)Never forget this)) Fewer game modes, add backfill to comp.
    Game layout suggestion to optimize queue times:
    Competitive
    -Role Que 2x2x2 format (1-2 players per party)
    -Group Que 2x2x2 format (3(one of each role) or 6 players per
    party)
    Quickplay
    -Role Que 2x2x2 format (1-2 players per party)
    -Group Que 2x2x2 format (3(one of each role) or 6 players per
    party)
    Arcade
    -Deathmatch
    -Rotating game is chosen by the player base for the
    week
    PVE
    -Role Que 2x2x2 format (1-2 players per party)
    -Group Que 2x2x2 format (3(one of each role) or 6 players per
    party)
    Custom
    -Player designed game

  6. Adding more heroes, the ability to customize, remove hero swapping, and more defined roles. This will allow players to feel connected to their hero and focus on performing their role better. In other words, connected to the game, hero swapping not only encourages toxicity but also disconnects players from the game.

  7. This will make it less pay to win:
    -FPS lock the game at 120 Hz.

    Players won’t need a $6000 PC and $700 390 Hz Monitor to obtain higher ranks.
    (Lower cost of entry = more players).
    Allows players to play at higher resolutions in the pro league making your game
    look beautiful, new, and exciting (NOTHING DRAWS NEW PLAYERS LIKE HIGHER
    GRAPHICS GAMES).

I think this is a start. Other ideas and changes will need to be implemented along the way.

1 Like

The comment about the pro players was for sure an exaggeration lol. Was just making it clear there was a gross difference in skill level between teams.

I think you misunderstand me when I said going with the flow. On the contrary, I do want to climb. I just think making sure you have more consistent teammates is the way to do that. As soon as I feel either our team is steam rolling them or if we are getting steamrolled, I’m gonna real in my efforts in those games. In the case of our team steamrolling, I’ll still seek to secure the win of course. It’s just, having consistently high stats in this game seems to get you punished.

I saw one person mention it best. I’m paraphrasing but he said something like a GM player “will” get to GM rank but because of the matchmaking system, he may only have a 52% winrate where as he should be getting a 60%. He’ll get there, it’ll just take much longer. So in my opinion, why not game the system back. Ensure you get that higher winrate.

The idea of playing with a six stack is definitely appealing to me. And lots of respect for coaching teams! I would say doing this would also give you an edge on the matchmaking system for sure. I’d definitely check this out it I had more time.

I mean… I’m not going to say that you’re wrong here (because you’re not). But if you think it’s purely this simple, you’re missing some key bits of information.

  • One of which is remembering you’re playing against other human beings who also want to win the game.

I recommend the Elo Hell Discord Server. That’s one of the places you could start if you wish to find people to play with in your rank and interested in learning more about how to play competitively.

It’s funny that you think this. That’s not actually true.

  • You play good, the game tries to match you with players who also play good. Kinda like increasing the difficulty. That’s completely normal.

You’re free to play as you wish, but defeatist mentality isn’t competitive.

Have you ever stopped and asked yourself, why is my team getting steam rolled? What could “I” change about what I’m doing to prevent the steam roll?

  • To be fair, it is everyone’s responsibility to adjust their play style.

  • But since people will be people and they’re not going to even if you ask, you have to be the one to take control of the game.

This person probably isn’t a high level Overwatch player if they think this is a problem.

Disclaimer: I’m not a GM player, but I am a Masters player and I’ve spent a lot of time leveling my own accounts and I used to level accounts for other people.

Lets assume you asked one of my GM friends (Abdullah and Basil here on the forums are a good example of people I know outside of here) to run an unranked to GM account.

The initial 5 games in Gold/Platinum will be very easy for them because the relative skill of the environment is well below their ability and potential.

Once they get to Diamond, where teammates (in general) are more consistent, they’ll likely drop from an 80% winrate to something closer to a 70% winrate.

Once they escape Diamond and are climbing through Masters

  • Specifically 3500 to 3700/3750

Their winrate should drop to around 60%.

As they approach 3.8/3.9/4.0/4.1/4.2 (whatever their peek SR is), they should see a convergence between their main account’s winrate (around 50% +/- 5 depending on the day/week).

Why does this happen, you ask? Well think about it. In the game Of Overwatch, you’re on a team of 6 players. That means you’re responsible for 16% of the effort (minimum) assuming everyone on your team does their share of the work.

When you put someone in a rank they don’t belong in (for example a GM player in Platinum), they will be contributing much more than 16% of their share of the effort. Something closer to 35% or so (assuming they have a perfect game).

As you get into games (higher in rank) where other players skill is more in line with the specific player, their ability to have drastic impact on the game slowly drops until they’re no longer able to contribute more than their 16%

  • Which is 100% normal.

As you climb the ladder, you encounter other players who are just as good (or better) than you are. If you don’t correctly identify what the problem is (when you’re losing) you’re going to end up raging at the wrong teammates for the wrong reasons.

  • Keep that in mind the next time you flame your DPS in chat.

You can have a high win rate and still remain in your rank simply because the system identifies that you aren’t pulling your weight.

I think you’ll run into the same problems that you’re currently running into.

  • Why?

Because not everyone is capable of improving at the same rate as others. Giving you and your duo partner the benefit of doubt, whoever you select to play DPS for you or support for you, might not be as intelligent (with respect to game sense) as you are and they will anchor you.

Example: Overwatch was my first competitive FPS. At this point, I think I’ve put well over 500 hrs into Aim Training alone (between Overwatch custom games and Kovaaks). This has allowed me (on my worst days) to jump onto an alt account in 2.7/2.8 lobbies and still absolutely roll kids. Unless you find a DPS player who practices their heroes regularly and often, that player will anchor your performance.

You and your partner play tank duo. Ana is great for Rien or monkey!! But if you select a support player who is terrible at positioning themselves and constantly getting killed by the enemy Genji/Widow/Tracer/Soldier, this player is going to anchor your performance. And keep in mind, positioning is a dynamic and abstract concept that is not simple to learn and even more difficult to master.

Climbing in Overwatch isn’t a function of time. Its a function of understanding the game and correcting your mistakes so you can improve in performance overall.

There are plenty of these rigged match maker kids who have never made it out of gold and they’ve been playing since Overwatch released 6 years ago!!

  • Not really the kind of people “I” would choose to take gaming advice from.
  • There’s no “Get Rich Quick” guide to Overwatch, unfortunately.
2 Likes

What is mmr that everyone is talking about? Please show us an example dataset.

I personally believe that blizzard introduced this term to confuse players and lead us down the wrong path. No one benefits from using terms that are not properly defined, there is no way to discuss this in a proper way.

who is gonna tell this man that literally every competitive game uses mmr

2 Likes

MMR is not a number, MMR is more than just a function that estimates the future course of your SR.
I assume that it is a very large matrix that goes through many different algorithms to accomplish a lot of secret things and winning is only a result of all the preceding factors, with an accuracy up to the second decimal place.
But we don’t know the specific elements or the algorithms, so why would we want a reset or even talk about MMR, it is all speculation and not really helpful.

There are 100 other metrics that are not listed or visible.
At some point, let’s talk about these secret metrics and how they are weighted differently within the different ranks.
In Overwatch, players are compared to other players within their rank and adjacent ranks, and this is wrong.

It is not a universal term and can mean something completely different in different games.
In other games, I don’t think matchmaking is as shady or rigged as it is in OW.

Yes it is
It´s the internal value that dictates your “skill”
Your “true SR” if you want to call it that

SR is just the outside representation

Could you explain why this is wrong?

1 Like

No it’s not a single number, Jeff Kaplan said that in an interview years ago.

No, it is not right, and there is no way for you to prove it, that is the main reason why it is wrong. Blizzard could verify it, but they won’t.
Why would a company introduce a term that you can do absolutely nothing with and is completely unverifiable?
Please think about it.
They just threw you and many others off with this new term, that means nothing at all!

This only works well as long as you are sure about the authenticity of the player base.
With the introduction of lots of alternative accounts/rank boosters and very cheap cheats, a realistic representation of one’s rank is no longer possible.
There’s a reason most shooter games choose not to implement this.

Besides, it’s a team shooter, so the chances of someone playing dirty are not that small! and the odds are against you if you don’t cheat.
This way you don’t get punished just once, but constantly, because you are compared to players who play dirty.