Forced 50% Matchmaker Theory

My theory is that one-tricks that are masters of only one hero, like me, are far more susceptible to getting hit by the forced 50% matchmaker than someone who is good at playing a lot of heroes since it’s hard to get a winstreak with a hero you are only “good” at. All the time, I get winstreaks followed by loss streaks and vice versa.

Win win win win win win win win win win win win win win loss loss loss loss loss loss loss loss loss loss loss loss (yes, this actually happened to me, a 14 winstreak followed immediately by a 12 loss streak) and that is my longest winstreak and loss streak I’ve ever had on my main account.

And that wasn’t just a one-time thing, it happens all the time. I pretty much only play Sylvanas because frankly, I think every other character in the game is boring and irrelevant. I don’t play HotS because I like the game, but because I love playing Sylvanas. (victory for the Forsaken, we will slaughter anyone who stands in our way, and all that jazz)

For someone who just enjoys HotS, you probably will never feel the sting of the forced matchmaker so of course you will deny it’s existence, even though I’m pretty sure when HotS was still in alpha, one of the features they bragged about was a system that tried to force 50% winrates.

EDIT: Yes, if you want to avoid the 50% matchmaker entirely, just play in a 5man of people you know are good all the time. The people who truly exploit the game will do that.

3 Likes

Uh…OK. I can’t really speak to the one-versus-many heroes part. Interesting idea.

But I really wish people would just admit that a 50% win rate is the goal of every well-functioning match maker by definition, and that the implementation depends on the size of the player base and resulting rules. That’s all there is to it.

Instead, it’s a conspiracy to “force” losses.

Vanity.

5 Likes

if the odds of winning and losing are 50%, then the odds of having 14 wins in a row naturally are 0.0061%, meaning that if you play ten thousand matches, you will get a 61% chance of that happening. that is abso-freaking-lutely absurd because it would take you YEARS. if you play a single match per day then that is like 27 years.

followed by ANOTHER 12 losing streak, it has a 0.024% chance of occurring.

put them BESIDE each other and well thats a lifetime.
( 100,000 matches for 14% chance, and 1 match per day, makes u need to live approximately 300 years )

not making any assumptions just pointing that out

ps: the propabillity of every single match is ALWAYS 50/50. but the chance of this chain to occurr naturally is abnormally low ( if my calculations are correct u know im a moron :stuck_out_tongue: )

1 Like

People don’t care that we are getting a 50% win rate. People care how they are getting a 50% win rate.

9 Likes

Yes. That is why I stipulated that the implementation depends on the size of the player base. MMR averaging is probably the best they can do with the size of the player base. There simply are not enough people at progressively higher echelons to match with.

In any event, the problem is “discussed” ad nauseum and you’d think people would be tired of complaining about it by now. Guess not!

4 Likes

If the matchamker was really trying to make you lose it would match you against people smart enough to ban your only hero and win more than 50%.

4 Likes

People are just going to leave the game before they stop “crying” about something that is slowly killing off the player base.

2 Likes

Well the buddy system sucks no way to defend that.
But saying people would accept a system that does not care if they have 20% wr on a regular basis is a bit…because well only the very top would have something close to 60%wr

FTFY. Going by queue time QM at least has way more than enough people not to use averaging.

This presumes independence of these events; humans though can tilt over series of losses, get high on wins, and so on.

And this is not even taking other factors like having a bad day into account.

5 Likes

I agree with this.
The buddy system trying to pair low wr people with people with higher wr is a thing.
They could pair people with similar wr in QM.

3 Likes

I’m not so sure. ‘Expanded’ search kicks in for me after just 30 seconds on EU – which, so I understand, is one of the healthier servers? I’ve heard that elsewhere it kicks in even faster (someone on this forum mentioned 10 seconds the other day). And usually in those games where I can tell that I am the strongest on my team, I can spot my counterpart in the opposing team.

Can you upload one of these losses where you were match fixed?

:wink:

1 Like

Werent you plat? Or do you mean only that hero

I have had expanded search within a few seconds, that means almost instantly. After 1 AM.

30 seconds would be a luxury. I’d always cancel at 25-29 then requeue until I get a normal match where I’m not supposed to mentor anybody new to the game or at 35% win rate.

2 Likes

Doesn’t the expanded search also kick in because not enough tanks and healers to get from the pool?
Not sure numbers of player can do anything for this one.

2 Likes

As i understand it, while QM does not enforce those comps anymore it still tries to create such teams first: due to healers and tanks being chronically rare, this then leads to expanded search fast.

Wasn´t there a patch note a while back reducing time to expanded search exactly for this reason?

I do think i´ll check by queueing with healers with assassins and see if i can tell the difference.

Well, yeah; this is another factor: time of day.

As someone who plays only Murky or Qhira, I agree with this and have experienced the same thing. It feels like a forced loss because after you win so much, the game tries to bring your win rate down to 50%. It does this by matching you against heroes that perform well against the hero you picked, so you’ll find yourself against 4-5 heroes who do extremely well statistically against Sylvannas. You are also expected to carry, since your MMR is likely the highest on your team at that point, so you’ll be given 4 teammates with lower MMR and are likely on heroes that don’t complement Sylvannas. This is an example of a forced loss.

For example, I go on long win streaks with Qhira all the time. Suddenly I start getting matched against extremely mobile team comps that feel like they were designed to solely counter me like Tracer, Genji, Uther, Li Ming and ETC, while our team comp has absolutely no hard CC.

1 Like

easier explanation is just that any random system with a binary outcome will eventually yield a 50% result. And their matchmaker isn’t much better than a random system - maybe it can tell the difference between a grand master and a bronze, but pretty much everything between that has a lot of uncertainty…

4 Likes

I absolutely love the broken, tangled logic surrounding every “forced loss” conspiracy. Does the matchmaker force losses through heroes or players? Guy A says it gives him useless teammates, Guy B says it gives the enemies unbeatable comps. Depends on what they lost to today I guess.

Imagine maining one or two heroes and playing tonnes of games as just them (particularly ones that aren’t very strong and have lots of hard counters to begin with) and sometimes running into bad matchups. Statistically you’d think you’d just run into bad matchups by playing more, thus increasing the odds of a bad game, but CLEARLY it must be intelligent design.

When you win it’s because your skill carried you above the mark, but when you lose it’s because eeeeeeeevil Activsion needs to feed. That’s why they do that obviously, just to feed on you. You, specifically, and not the other nine people. They’re just numbers used to very specifically manipulate just you.

It’s almost as if… you play weak heroes in QM… and you lose games… because of random chance :exploding_head:

5 Likes