dota 2 is 50% forced win rate

[OnlyFarm]

its just too real. 1 win 1 lose repeat until you realize you already old and this loop will never end. it is said this game is dead. but i hope it is.

[Ragnoraok]

Първоначално публикувано от hyker233:

no thats just your average matchmaker going by completely random numbers and also taking into consideration stats and variables but mainly going by random. Thats why if it is operating at random and you stack a team up of 5 friends it becomes not random at all. You want to know why because it cant be random you have 5 people on a team there problem solved . It only seems 50 percent win rate cuz thats how random works actually its probably less than 50 percent win rate cuz if it was 50 percent win rate then no on would lose mmr. As some very smart people have stated so i believe it is actually not rigged at all and any average player has around a 25 to 35 percent win rate. So theirs your proof right there if your losing mmr then you've proved the 50 percent win rate theory to be absolutely false.

LOL, you have no idea how matchmaking or programming works; that much is obvious from your inane posts.

Matchmaking does not work through randomization. It matches people based off fixed parameters, with the only variables being player avaiblity. The output of a mm algorithm sohuld consistently create games where everyone in the lobby is of equal skill - that is to say that if there are only 100 people playing, then you should ALWAYS see the same 10 people matched together.

The only way that you can have such a narrow spread in winrate, to the point where even the top 1% of player can barely scratch 55%, is if there is some additional variable(s) that forces people towards some predetermined threshold. This is incredibly simple to implement in an algorithm. The difficult part is trying to make the manipulation seamless or not apparent, but Valve fails at this since its easy as hell to show that the game will strive to keep you at a 50% winrate. I have tried this myself where I intentionally tried to throw games in a 30-game window and STILL had around a 50% winrate despite my games being a 4v6; that can't happen with a fair matchmaker.

[:D]

ye imagine all us having 80 % winrate

[Ulthwé]

Първоначално публикувано от huez:

Първоначално публикувано от uɐɐılʎʇs:

These are always a fun read. People who throw conspiracies to explain how these things just naturally work out.

Apart from those at the very top and very bottom you will pretty much get close to a 50% winrate if you play enough. You hit your skill level and are facing those of similar skill and win roughly as much as you lose. After a certain point youve now played that many games that moving up even a percent will take a massive win streak, which will raise you to a new skill level and if you cant retain that, you get a loss streak. If you can you now stagnate at a new skill level and start losing as much as you win and back to 50ish %.

They dont need to add some elaborate matchmaker to skrew you over, you will just hit that point naturally.

Those at the top get to play people below their skill level for majority of games and can buck the trend, no one/less people above their level to get matched with. And the reverse is true for those at the bottom, higher chance of loss.

But no, obviously its the matchmaker thats keeping you down and you should be winning far more then you are losing for your entire career, which would eventually put you at the top end of the leaderboard. Coz everyones a pro.

You are kind of right but honestly, sometimes people win even if they want to lose.
I have tried that by just being as useless as possible, without feed etc. Just being zero in my team. Still kinda won nearly 50% of the games.

Problem really is not that people reach their limits but problem is that teams are often very unbalanced. This why people rage mostly. Like if u really trying to win and u almost win alone but your team is ding really nothing useful but rather pushing towards losing the game. This is where people rage.

When i tried to lose games i did not feed or push the game to the extreme losing state but i was more like, minding my own stuff. Acting like doing something useful or straight afk jungling while my team just stomps the other team. 10/35 score or something. Then winning like 7 games in a row like this. Doing pretty much nothing at all. Thats my skill boys!!! Im the winner here! LOL

It really made me feel that winning or losing has nothing to do with my skill tbh.

Next games were like opposite, where my whole team got stomped. I did not mind it but it was funny. Sometimes creeps seem stonger than heroes, games get to that point.

I believe that this is where people get angry. When you try your best but your team members are like 0/13/1 like every one of them. What u gon do?

In this case yes. Someone whos legendary can beat those games in picking stage. Picikng one of the most unbalanced heroes depending on enemy picks and just wreck them.

Yea, system works, but the way it works is just frustrating for people. There is no good balance in teams. While its not easy to predict player picks, laning, items and everything possible in dota games. Sometimes it rather seems that games are created unequally.
It feels like u have 2 teams with a 3000 MMR difference. And this is multiplied by 5.

this is why people come to cry in forums, not because they want to be a legendary player.
Because their games are uneven.

If you win more when you do nothing or almost nothing than when you try to actually play, then I have some bad news. You probably play really badly... Although an insight within your open dota profile could give us directions to give you proper advice.

[Ulthwé]

Първоначално публикувано от Ragnoraok:

Първоначално публикувано от Technomancer:

So, if there was no matchmaking 'rigging', what would the average player's winrate be?

Each player's winrate would be reflective of their contribution to matches. So someone who routinely is a major asset in teams would have a winrate near 60%, whereas someone who is routinely the worst-performing would have abysmal winrates.

In DOTA, nearly everyone with more than 1K games is within the 45-55% winrate, even people who play majorly in premades, who should easily have winrates in the 70%.

So you say it is forced 50, then you say it is forced 45-55%, how does this make sense?

Spoiler, it doesn't. Because if there is a 10% winrate range variation, you can't claim it is forced anymore.

[Hadji*]

No, You're just bad

[Ragnoraok]

Първоначално публикувано от Ulthwé:

Първоначално публикувано от Ragnoraok:

Each player's winrate would be reflective of their contribution to matches. So someone who routinely is a major asset in teams would have a winrate near 60%, whereas someone who is routinely the worst-performing would have abysmal winrates.

In DOTA, nearly everyone with more than 1K games is within the 45-55% winrate, even people who play majorly in premades, who should easily have winrates in the 70%.

So you say it is forced 50, then you say it is forced 45-55%, how does this make sense?

Spoiler, it doesn't. Because if there is a 10% winrate range variation, you can't claim it is forced anymore.

Matchmaking is a deterministic output meaning that if all variables are the same, then the program will always give the same results. However, there are confounding variables that cannot be accounted for such as smurfs, disconnects, etc., and as such you are going to have natural variance in winrates. Nonetheless, there are expected deviation that can be measured and predicted. A +/-5% deviance from 50% is VERY indicative that either individual performance does not matter or the matchmaker is creating so many 1-sided matches that it causes winrates to "average" out to 50%.

[Ulthwé]

Първоначално публикувано от Ragnoraok:

Първоначално публикувано от Ulthwé:

So you say it is forced 50, then you say it is forced 45-55%, how does this make sense?

Spoiler, it doesn't. Because if there is a 10% winrate range variation, you can't claim it is forced anymore.

Matchmaking is a deterministic output meaning that if all variables are the same, then the program will always give the same results. However, there are confounding variables that cannot be accounted for such as smurfs, disconnects, etc., and as such you are going to have natural variance in winrates. Nonetheless, there are expected deviation that can be measured and predicted. A +/-5% deviance from 50% is VERY indicative that either individual performance does not matter or the matchmaker is creating so many 1-sided matches that it causes winrates to "average" out to 50%.

You must have a poor understanding in maths or at least in logic to think that.

[uɐɐılʎʇs]

Първоначално публикувано от Ragnoraok:

Each player's winrate would be reflective of their contribution to matches. So someone who routinely is a major asset in teams would have a winrate near 60%, whereas someone who is routinely the worst-performing would have abysmal winrates.

They will have those sort of winrates in the short term until they reach the rank that suits their skill level. And those good or bad players would then face people equal to them and start losing as much as they win. And that 60% win rate will slowly trend closer to 50% the more they play without climbing.

Lets say 60% winrate in first 200 games. So 120 wins, 80 losses. They reach the suitable rank for their skill, face opponent teams of equal skill and lose as much as they win. They play 1000 more games without getting better or worse. thats roughly 1120 to 1080 w/l. close to 50%.
Now if you continued to win 60%+ permanently you would never stop ranking up and end up in the top maybe 10% or so.
Its the top end and bottom end that buck the roughly 50% trend, as the top has far few people above them to be matched.

Първоначално публикувано от Ragnoraok:

In DOTA, nearly everyone with more than 1K games is within the 45-55% winrate, even people who play majorly in premades, who should easily have winrates in the 70%.

Coz the more time you spend playing those of equal skill means more time winning roughly 50%. Doesnt matter if you win your first 200 games, after 1000 more games played against equal players and getting 50% will put your average down to about 60-70% another 1000 games and closer to 50%.

Premades can get beaten by solos, they also face a lot of other premades. being in a group isnt a sure fire way to win. And if 2 premades face eachother 1 has to lose.

Its no conspiracy. its just the natural order of things.

[Ragnoraok]

Първоначално публикувано от Ulthwé:

Първоначално публикувано от Ragnoraok:

Matchmaking is a deterministic output meaning that if all variables are the same, then the program will always give the same results. However, there are confounding variables that cannot be accounted for such as smurfs, disconnects, etc., and as such you are going to have natural variance in winrates. Nonetheless, there are expected deviation that can be measured and predicted. A +/-5% deviance from 50% is VERY indicative that either individual performance does not matter or the matchmaker is creating so many 1-sided matches that it causes winrates to "average" out to 50%.

You must have a poor understanding in maths or at least in logic to think that.

Right....as if someone who thinks that matchmaking uses randomized components and doesn't understand how variance works is in any position to talk.

[Ragnoraok]

Първоначално публикувано от uɐɐılʎʇs:

Първоначално публикувано от Ragnoraok:

Each player's winrate would be reflective of their contribution to matches. So someone who routinely is a major asset in teams would have a winrate near 60%, whereas someone who is routinely the worst-performing would have abysmal winrates.

They will have those sort of winrates in the short term until they reach the rank that suits their skill level. And those good or bad players would then face people equal to them and start losing as much as they win. And that 60% win rate will slowly trend closer to 50% the more they play without climbing.

Lets say 60% winrate in first 200 games. So 120 wins, 80 losses. They reach the suitable rank for their skill, face opponent teams of equal skill and lose as much as they win. They play 1000 more games without getting better or worse. thats roughly 1120 to 1080 w/l. close to 50%.
Now if you continued to win 60%+ permanently you would never stop ranking up and end up in the top maybe 10% or so.
Its the top end and bottom end that buck the roughly 50% trend, as the top has far few people above them to be matched.

Първоначално публикувано от Ragnoraok:

In DOTA, nearly everyone with more than 1K games is within the 45-55% winrate, even people who play majorly in premades, who should easily have winrates in the 70%.

Coz the more time you spend playing those of equal skill means more time winning roughly 50%. Doesnt matter if you win your first 200 games, after 1000 more games played against equal players and getting 50% will put your average down to about 60-70% another 1000 games and closer to 50%.

Premades can get beaten by solos, they also face a lot of other premades. being in a group isnt a sure fire way to win. And if 2 premades face eachother 1 has to lose.

Its no conspiracy. its just the natural order of things.

In order to be in higher skill brackets, you have to have a high global winrate because you need to win substantially more to progress up the rankings. The higher the rankings you go, the higher the global winrate you should have. In order to have a 50% winrate at high levels, players would have to spend the same amount of time in their current bracket as they did in all their lower brackets.

Immortal players in particular should have winrates at least in the 60% range since they have demonstrated competency that puts them higher than the top 1%. Yet few immortal players have 60% winrates.

Premades are provided with such an advantage that even Valve acknowledges this hence why there are separate premade and solo ques. Premades in particular are one of the major driving forces behind boosted accounts making it to ranks that they have no business being in.

You don't know what a conspiracy means. There is clear evidence that shows the matchmaker creates far more 1-sided matches than "fair" matches, and that the end result is a universal threshold that is met by all players regardless of skill. You can watch any high level match and see that teams are grossly imbalanced in terms of skill. You may not appreciate or respect evidence that is not in the form of Valve flat out admitting how their algorithm works, but that does not mean that the evidence is nonexistent. Also, there is no natural order of things, because even if the matchmaker made fair games, it is STILL an artificial algorithm giving a specific output. The only way for things to be natural is if matchmaking was completely random.

[Ulthwé]

Първоначално публикувано от Ragnoraok:

Първоначално публикувано от Ulthwé:

You must have a poor understanding in maths or at least in logic to think that.

Right....as if someone who thinks that matchmaking uses randomized components and doesn't understand how variance works is in any position to talk.

Ask those who complain if they would have complained with a 55% winrate. Also, seeing them co.plain about a so called forced 50 is laughable when they have lower than 50% winrate. I saw only one person on this forum being disadvantaged by matchmaking. It was a guy who had a significant gap between dire and radiant, and as most know, there is a gap in winrate between both sides.

[Ragnoraok]

Първоначално публикувано от Ulthwé:

Първоначално публикувано от Ragnoraok:

Right....as if someone who thinks that matchmaking uses randomized components and doesn't understand how variance works is in any position to talk.

Ask those who complain if they would have complained with a 55% winrate. Also, seeing them co.plain about a so called forced 50 is laughable when they have lower than 50% winrate.

Again, that is how variance works.

[BigDaddy RiRoS]

If this was true, then there wouldn't be any boosters.
Case closed.

[uɐɐılʎʇs]

Първоначално публикувано от Ragnoraok:

In order to be in higher skill brackets, you have to have a high global winrate because you need to win substantially more to progress up the rankings. The higher the rankings you go, the higher the global winrate you should have

Not necessarily, i could lose my first 1000 games and be herald 1. if i win my next 100 ill go up ranks but still have a WR of about 10%. So there is a huge variance in what your winrate could be based on how long it took you to improve, how much of an improvement and how many games spent at each level of skill.

Първоначално публикувано от Ragnoraok:

. In order to have a 50% winrate at high levels, players would have to spend the same amount of time in their current bracket as they did in all their lower brackets.

Nope, if im herald and have 50% winrate with 500 games and move up to guardian in the next 100games(75%wr) then that would only move me to around 54%.[/quote]

Its just as much about how you got to where you are as to where you are. Many players have thousands of game the harder it is to have a higher then average winrate. you give the best player in the world an account with 5000 games and 50% winrate at herald and theyll climb extremely fast back up the ranks, but those few games it took to get to a high ranking will have little influence over the 5k games earlier.

Първоначално публикувано от Ragnoraok:

Immortal players in particular should have winrates at least in the 60% range since they have demonstrated competency that puts them higher than the top 1%. Yet few immortal players have 60% winrates.

Because no matter what your rank(apart from very bottom and top) you will eventually end up at the rank that matches your skill. You will then face opponents of equal measure and begin to lose nearly as much as you win and your rank/mmr will show minor variance. The more games you play like this the further and further towards 50% you will go.

To retain a 60% winrate you obviously need to be winning more then losing consistently without periods of stagnation that would drop it, the longer you stagnate the harder it will be to push your WR up, as each win will be worth less and less towards your winrate.
How many immortals just installed the game and were good enough to never get stuck at crusader or archon or legend for any extended time.
If you stagnate at legend for 3000 games before things click and you start rising to immortal then each percent up or down is worth just over 30 games, assuming thats 30 in a row up or down.

All you need to reach high ranks is to win more then you lose and have a consistent net gain MMR. A 52% winrate from day 1 and held consistently will eventually get you a high rank, albeit over a vast amount of games.
You could potentially get immortal with less then a 10% winrate if an immortal player got a hold of an account with 100% losses for many games. divine is what 6k mmr. so 300 wins in a row as a solo. Now thats extremely unlikely obviously since its smurfing/account sharing or whatever, and 300 wins as a solo is unlikely, but even if they were hitting 75% for those small number of games it wouldnt affect overall win rates much at all.

Just looking at the end result and ignoring the journey makes little sense.

Първоначално публикувано от Ragnoraok:

In DOTA, nearly everyone with more than 1K games is within the 45-55% winrate, even people who play majorly in premades, who should easily have winrates in the 70%.

Premades are provided with such an advantage that even Valve acknowledges this hence why there are separate premade and solo ques. Premades in particular are one of the major driving forces behind boosted accounts making it to ranks that they have no business being in.

(Added both your quotes together from 2 posts)

So premades should easily have 70%,
Im sort of lost on your arguement since you countered it yourself in your next post by mentioning that premades face other premades, therefore nullifying the group advantage.

Those premades boosting are also most likely below their real rank so less likely to face their equals, also with many hours playing together and using meta heroes etc. And to boost would also be bucking the 50% forced when they play seriously, then purposely losing to lower rank again to avoid facing those of equal skill. Basically smurfing to a degree to be able to sell services.

Първоначално публикувано от Ragnoraok:

You don't know what a conspiracy means. There is clear evidence that shows the matchmaker creates far more 1-sided matches than "fair" matches, and that the end result is a universal threshold that is met by all players regardless of skill. You can watch any high level match and see that teams are grossly imbalanced in terms of skill. You may not appreciate or respect evidence that is not in the form of Valve flat out admitting how their algorithm works, but that does not mean that the evidence is nonexistent.

If there was overwhelming proof this wouldnt even be a thread. It would be a post of links to irrefutable evidence. Not i feel this, or it just seems to happen this way etc.
what was said was

Valve senior software engineer Jeff Hill
“The Dota matchmaker will optimize for each individual game made being well-balanced, defined as games where the matchmaker predicts each side has an equal chance to win. As a consequence of this goal, over the long term all players will tend towards a 50% personal win rate… A 50% lifetime win rate isn’t an explicit goal or constraint of the matchmaker, rather it’s a consequence of trying to make the teams for each individual game fairly…” said Hill.

Which immediately got misinterpreted as proof.
The game tries to make a balanced match based on similarly skilled level, one where it thinks both sides have an equal chance of winning. and people immediately thing since you have a 50% chance to win its proved 50 winrate.
Facing a balanced team of your equals should at face value give you a 50% chance to win. By getting better you start consistently beating those in your pool to progress to a slightly harder pool, till you get better again to consistently beat them, stagnate and win as much as you lose, or dont meet that standard and fall.

[Zagryzaec]

Първоначално публикувано от Arimanius:

Първоначално публикувано от Zagryzaec:

Ever imagined how it works?

Imagine if a system actually create a ladder from players according to their skill. From wrakest to strongest.

And as you win vs similar leveled opponents you moves up on this ledder and down if you lose.

At some point when you reach your skill level and once you dont improve, you will meet opponents that are just better than you. You cant beat them unless you improve your skill. Like they can lift 120 kg, and you can only lift 100 kg. If you wont train to lift 130 kg you will never beat them in powerlifting.

So reaching this level you will always be losing.

From the other side below you will be people way weaker than you. Even at their best they cant beat you. So whenever you drop far enough you cant really lose. Only intentionally.

So within this reach you will be swaying back and forth, winning raising to top of your abilities, snd then struggling, untill you are exhausted and fall down along with your skill, untill you reach your bottom. From which you easily get back. Overall since you stuck in same bracket your winrate will be moving toward 50%+-.

There is no conspiracy, its natural, and the component you are missing is improvement over yourself. You will never raise by grinding without changing. You may catch a winstrwak if you are lucky but then will be put back down.

This is the most retarded BS I have heard. You don't understand how the 50% winrate policy works; It doesn't put you against good opponents, but rather trash team mates... you don't win because the enemy is better than you; you lose because part of your team is actively griefing the game.

For real, I don't know what is worst: if the so called "conspiracy" about the 50% winrate policy or Valve Fanboys like you defending a flawed system.

Those are real people. Why do you imagine system is revolving around you? No one cares about you specifically. System gives you just random teammates of the same rank. I remember bieng put with and against the same teammates in consequtive games. Was im the one who supposed to grief him or was he the one who was supposed to grief me?

My expirience shows that usually both teams has chance of victory, even when it may seem that you dont have it. Its just some players like you just used to give up.

Also every time i was playing cavern crawl in dota i was systematically falling 2-8 stars (up to three ranks) and afterwards playing my main characters (all of which are supports) ive been easily climbing back.

Which by itself proves two things.

1)There are no 50/50 system.
2) you and you exactly is the defining figure in the outcome of most matches. No one but you.


I would explain how support affect game of carry, but its too long and you dont want to listen anyways.