Professional gambler explains why Competitive matches are rigged and how to negate the forced win/loss

sooo…do you have any tips

You typed alot of words to essentially agree with him under the pretext you were arguing.

The thing is, people also do not realize that this matchmaking system takes into consideration your games outside comp to “tweak” your MMR.

That is what leveling up your account in QP heavily influences your outcome in comp. I even tested this by buying a smurf and playing with friends who are gold, plat, diamond, etc. I went from 3.8k to 3.2k and been around 3.3k ish since then.

Whereas on my main account, I am able to easily win GM games, of course, I lose a few too but there is no way to justify a 700+ points gap UNLESS a system such as “hidden MMR” works in the background.

1 Like

I have noticed this for about three years now.

1 Like

Should have been 5 - 5 by OP’s core evidence…

2 Likes

Some here would attest that you don’t belong in GM, and that you were ‘boosted’ to get that rank, because the MMR system is infallible.

1 Like

Professional gambler

If there is anyone to trust regarding issues of scientific accuracy and reasoned thinking, it’s definitely a compulsive gambler.

Yes yes.

6 Likes

Weighted bayesian skill systems aren’t rigged.

Without getting too specific, Blizzard isn’t guilty of implementing this kind of matchmaking any more than any other matchmaking system in the world, including Chess dot com.

The matchmaker tunes itself repeatedly.

  • Over the course of many games, it discovers what factors in your gameplay are prominent, and assigns those characteristics a number value.

  • It compares your number values with other players with similar number values.

  • The matchmaker uses a secondary filter, your SR, to filter you into an initial skill “bracket”/pool (people in your rank) + sift out players that are not around your rank to make them ineligible to be in a match with you.

  • SR is then used as tertiary filter to compare your numerical values to the average numerical values of of people in that pool.

Using those numbers, the matchmaker sorts a game. There’s nothing special about this, nor is it “rigged”. It’s weighted, there’s a difference. If the averages are accurate representations of players, then the natural chance of winning SHOULD hover around 50%…

The alternative of pure anarchy of sorting the fastest possible matches ends up being an unsavory experience for EVERYONE involved.

That’s not what a forced 50/50 is.

A forced 50/50 would be saddling you with actual literal bots (CPU players), beginners, or players whose average numerical values are too far below what they should be immediately after a win, where it’s a statistical impossibility to win. That doesn’t actually happen.

What actually happens is that the pool of available players to sort into a match has to widen beyond the initial sorting because no one is around your skill level. It has to do this to sort into a match.

Widening the pool means that the players that get sorted into the match with you have varying disparities in their numbers. And that’s exacerbated by role queue where if their numbers don’t match the average, they underperform.

4 Likes

if this is infact true there would be actual numbers to back it up, there isnt and until theres multiple datasets to show it is true then its not true

1 Like

Bayesian skill systems drive most online matchmaking games.

Every eSport forks one of the more common versions of those skill rating systems like Chess, gives it flavor with differently named ranks and titles and prerequisites.

Datasets of players are actually some of the most closely guarded company secrets, but that doesn’t stop developers from talking about them. That’s how we know they exist.

Just because the public doesn’t have the datasets doesn’t mean they don’t exist.

edit - I’m just realizing, are you talking about datasets regarding OP’s forced 50/50 or my Bayesian skill systems? It’s hard to tell.

2 Likes

There is a difference between it not being rigged and it being infallible.

People are arguing that it is most likely not rigged. Not that it is infallible. People who actually understand the science know that it is fallible. Your MMR can never be determined with 100% certainty. ELO systems always are, as everything is, affected by tolerances. These tolerances have no significance to the average player, though. And of course they have nothing to do with the system being allegedly “rigged”.

2 Likes

I started at Bronze and am now Diamond.

Never, you say?

1 Like

So when pros do unranked to gms they turn off the algorithm am I right.

2 Likes

5050, youd be pretty easily able to see if you win a game the enemies sr should get higher compared to your team and same visa versa.

With all due respect, unless you have first hand experience developing the Overwatch Matchmaking Rating algorithm, your claims are as baseless as the people claiming that it is rigged.

Neither you, nor I, nor anyone who isn’t an agent of Blizzard knows how the MMR functions, or even what it looks like. We have not been privy to such information. It could be as ‘complex’ as Kaplan has claimed it is through multiple lines of code/data, or it could simply be an apple that adds or subtracts an apple depending on if you win or lose. Nobody knows, only the developers do.

Thus we know nothing.
You don’t know I don’t have an invisible dragon in my backyard. Your claim that I don’t is as baseless as mine.

Edit: I will repeat my claim just to make sure:

People are arguing that it is most likely not rigged.

Most likely not rigged. Most likely. Yes, knowledge is probabilistic. Yes, there might be a huge conspiracy by your government trying to make you feel comfortable in Overwatch, so they rig the system so you don’t realize that aliens are taking over the world. Or… or maybe you’re reading too much into win- and losestreaks.

Just maybe.

6 Likes

There’s a difference between a compulsive gambler and a professional gambler.

Professional gamblers work through statistics, not impulse.

2 Likes

I know, but then my joke wouldn’t have worked as well :penguin:

Also, considering what that guy wrote, he probably doesn’t know much about statistics.

1 Like

I don’t mean anything toward or against OP, just stating the difference.

Sorry. Asperger’s. Definitions are important to me :stuck_out_tongue:

1 Like

I didn’t make any claims purporting that which I don’t know to be truthful, or factual. You might have a dragon in your backyard, who am I to say you don’t?

As I said, neither you, nor I has any information on how the MMR functions, because the devs haven’t disclosed any such information.

And I say this respectfully, not to insult you.