Algorithmic Handicapping (MMR) is Wrong for Overwatch

well i was typing while trying to play :slight_smile:

also i edited the first sentence

I have done many tests myself. Play bad, win or lose, your nextgame will be HARDa

People who benefit from a system will rarely find fault with it. It makes sense that “low rank, hard stuck” players will find fault with the system because those players are the ones victimized by it.

Just because people are low rank, doesn’t mean their complaints are without merit.

2 Likes

How many games would a player need to play a season to make teammates irrelevant?

If the total amount of games to make teammates irrelevant in a season is (x) then can the majority of the player base play (x) amount of game in a season?

1 Like

If comp would work like this it would stop being a skill based system and would just be a system that ranks people by the amount of time they spent playing.

As soon as you are able to maintain a win rate above 50% ( most if not all top 500 players have >50% win) your place on the ladder becomes a function of how many games you can grind.

It would rank a 55 % winrate player above a player with 90%+ winrate if the former player just plays enough games.
That is not a measurement of skill, it is a measure of how much time you can spend on grinding.

1 Like

Are u trying to say matches should be based on sr alone?

  • also ranks/titles(bronze to t500) are only given to the top performers in each tier at the end of a season?
1 Like

Nope MMR can stay. They just need to remove all the factors effecting it, like other team rolled us, we lost by a narrow margin etc. Let MMR go up/down lets say 10-20 points faster (cumulative) than your SR.

The more games you play, the more accurate the rank. This corresponds to reducing the influence of random factors (including teammates) on your rank. However, accounts bounce around their true rank, and lower accounts bounce around more than higher accounts.

My analysis suggests that observing a history of several hundred games is necessary to get an accuracy of less than 100 SR.

I’ve also done some matlab simulations of what happens after a full reset (it’s bad), but I’ve never posted it.

Have you ever considered listening and testing things the way I’ve proposed? Many people are playing Overwatch by choosing hero’s that counter the enemy hero choices for the good of the team, sometimes changing 3-4 times during a match. For instance, if a person is using their best hero doomfist, and the other team has a Pharah, they can change to Soldier to discourage the Pharah which helps their team immensely, even if they are a somewhat mediocre Soldier. They may not kill Pharah often, or at all, but as long as they prevent her from going into the sky and spamming they are providing a huge benefit to the team by making one of the enemies more powerful hero’s obsolete.

By the end of the game this persons stats might not look very good, they didn’t get too many kills or inflict that much damage, but they helped their team win. The matchmaker will see these poor stats as poor performance and will punish them in upcoming matches.

A good Doomfist changing to an avg soldier is just one example, I could give a bunch. This is one of the main things that’s happening with the matchmaker that is causing huge issues. These huge loss streaks the matchmaker are putting people on due to the fact that the matchmaker has identified them as under performers when they were actually helping their team.

What you should do in order to test it, is to try playing flex instead of one tricking with zenyatta as your profile shows every season. Then the problems that people talk about will become blatantly obvious. Play for long hours and try to rank up and see that it’s almost impossible. Your 10 hour a season play times are not nearly enough to understand what the community is conveying.

1 Like

I will however say that you testing anything in this game isn’t going to be a proper test since you know how to avoid the matchmaking nuances like getting placed in the throwers queue etc. The general public do not know how to do this. I’m sure this is one of the reasons you ranked up from gold to plat in the first place - you attribute it to your skill when it’s actually because you know the underworkings of the matchmaker.

1 Like

No no no no we dont award top 500
or the top players for each tier.

I would determine rank reward based on SR.

Meaning

For GM’s only those with the most SR out of all GM’s will get the top 500 mark.
Of course to qualify u need at least 4000 sr

The same for every other rank. Master,diamond,plat, gold,silver and bronze.

Most likely tho blizz will have to add more ranks in between such as bronze 1-4 and so on.

1 Like

Last time we suggested this he claimed it would be throwing. Speaks volumes.

2 Likes

Yes. An example where using a hero you aren’t good at isn’t throwing: Reinhardt isn’t a strong hero of mine because I don’t spend time on him, so if I start the game instalocking Reinhardt before someone else can then someone could maybe say I’m throwing…but I have won countless games because I went reinhardt when nobody else could/would and we needed it.

1 Like

Like Blizzard corporation itself, you are afraid of real competition. Real competition is not engineered to give all competitors an equal chance of winning, according to nebulous proprietary metrics and algorithms. Real competition is ugly, unpredictable, and unfriendly to marketing.

Match Making Rating is Blizzard’s way of making their game as appealing and addictive as possible. It guarantees the near wins and losses that make matches feel dramatic, compelling us to repeat the experience over and over again. Like a slot machine, it is an evil manipulation of players’ human nature. And you, Kaawumba, are a tool for defending it.

Yes cuthbert people are so scared of facing your silver might in their competitive games after mmr is reset, yep that’s totally the reason some people don’t want it to happen, not that it would ruin ladder for a month or more…

3 Likes

So playing against people at your skill is not “real competition”? So what is? Is stomping players you are better than real competition? Is getting crushed by better players real competition? Because it has to be one of the three. Those are the only options. You either can play against worst players, equal players, or better players.

Personally I want to play against equal players. I’m not particularly interested in getting crushed by pro’s or stomping newbs. Having a balance match provides the most fun. I guess stomping people you’re better than can be fun for a little bit, but it gets boring real fast.

3 Likes

Would love to see your team sport player selection system, where you’re balancing the teams to ensure every team is equal.

And then call it “real competition.”

1 Like

But you still post on the forum? Sure, your life is soooo awesome that you don’t need to play Overwatch, but you still need to complain about it.

(Cuthbert still spams other threads with this crap).

1 Like

That is a real competition imo.