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.

[Zagryzaec]

Also immortal player in herald is pretty much guaranteed 99% winrate.

[76561199091735933]

Its forced POTENTIAL 50% winrate. Of cause if actually like 49.5-50.5 etc. Dotabuff are full or 40-ish winrate accounts - ruiners. Even system can't help them. Same for true smurfs/boosters - they can have up to 100.

So if you play ALL games without any ruining activities then yes you will at least stay at same MMR. If you improve you can slowly climb.

[76561198874287325]

İlk olarak Ragnoraok tarafından gönderildi:

İlk olarak hyker233 tarafından gönderildi:

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.

lol ok noob all i can really tell you since your caught up in your delusions is a your a noob and you hate that. And you need to get better cuz crying here isn't gonna get you what you want. No matter how many tears you shed, and if your pretending to have a noble cause that still isn't gonna change things FOR ANYONE LOL. So you have NO CHOICE BUT TO GET BETTER OR GET A 5 MAN TEAM NO MATER WHAT YOU BELIEVE ABOUT THE MATCHMAKING. And if your talking for your friends then tell them they have no choice and no amount of tears or excuses is gonna change this 100 percent correct and true matchmaking process lol you can cry now im not gonna respond anymore. I also think you noobs are forgetting how much people have improved in this game. Ever since guides and dota 2 gold and just the spread of dota 2 knowledge and tactics have become very very common. Thats why i think you noobs used to be ABle to win but after 7.25 and up the matchmaker became way more consistent and behavior score became a big factor as well. So dota 2 is even more strict now when you get reported so taking all those things into consideration. Anyone who isn't willing to play like a pro is gonna get stomped repeatedly.. Its almost as if dota 2 has a divided player base one that plays with complete dedication to the game and the other half are griefers that get carried always and believe that they actually contributed. Adding to your delusions that you should be a higher rank and that the matchmaker is rigged. By the way only thing i need to know is are oyu an american player cuz if that is the case i can already tell you why you cant win. Hence the phrase said by all foreigners that play dota 2
Stupid spoiled Americans easy. Now Im not saying i say that but any match with a Peruvian team against american team will tell you whats going on here. And yes im an american player but not your average one. I will admit though when i started dota 2 i was the worst player in the world literally the worst had like 1000 s of reports. Defintly not the case anymore but ya that will tell you alot about american players.

[Ragnoraok]

@uɐɐılʎʇs

1) No, you have to have higher than a 50% winrate in order to move up a rank. In your example, if it takes 1000 games to rank up, then you needed to have won at least 500 games to make it to guardian. When you make it out of guardian in 100 games, you need to win at least 50/100 games. Your global winrate in fact has to ALWAYS be higher than 50% to rank up; under no circumstances can you have a 40% winrate and rank up. The only time that your winrate stagnates at 50% is if you are incapable of ranking up for some reason. If your winrate is below 50% then you rank down. If anything, the very small spread in winrates is more than enough evidence that there is something in the code coercing some sort of threshold, because otherwise you would see people's winrates reflect their skill and thus winrates would be all over the place. The variance in skill in a given rank is monumental (especially at lower ranks), yet everyone has nearly the same winrate. Its a paradox that only makes sense if the matchmaker is routinely matching people of wildly different skill against each other, which of course means its not working as intended.

2) I already explained why that is wrong. Immortal players would have had high winrates to even make it to immortal to begin with. Once you reach immortal, some players are still more skillful than others, and winrates SHOULD reflect that but they don't. People in the top 10 should have higher winrates than everyone else, yet even Dendi has a 53% winrate. You are essentially arguing that everyone in immortal is equal to each other that that is objectively wrong....yet the matchmaker creates an output where everyone in immortal has similar winrates. The only people with really high ranks near 70% are pros who only play in premades with each other, which is the most purest merit-based game style out there.

3) If you are a 2,3, or 4 man premade then you will have solo que players on your team. I have the option to only play solo que matchmaking, and from to time I still get put with/against a premade. Also, premades themselves are not equal, and if anything have the geatest variance in skill. Premade matchmaking is the most pubstompy matchmaking that you will ever experience, and groups of peopl intentionally boost acounts so they can rank up easily by exploiting the terrible premade mm system.

4) There IS overwhelming proof, but you don't want to acknowledge it mainly because you think that that its normal for everyone's winrate to be around 50%. Why even quote a Valve employee? The first sentence alone refutes the entire point since the matchmaker fails terribly at creating even matches. Like in most of my games the 3 worst people in the lobby are on 1 team, when any functioning matchmaker would at least put 1 of them on the opposing team to make the game more competitive.

[uɐɐılʎʇs]

İlk olarak Ragnoraok tarafından gönderildi:

1) No, you have to have higher than a 50% winrate in order to move up a rank.

You dont need a higher global winrate to move up, just a higher winrate in a certain period of time. Guardian is about 900 mmr. thats 30 wins in a row as a solo.

lets say you start at 1 mmr. (hard cap on lowest possible)
30 wins(up to 900mmr) then 270 losses after puts you back at 1. 10 final result, 10% global winrate 1mmr.
270 losses first(cant go lower then 1) no change in mmr, then win 30 in a row. Final result 900mmr, 10% global winrate.

Not likely ofcourse but its an example that proves you dont need a global positive winrate to increase rank. Just a positive winrate in more recent games.
The many possible ways to get that rank reflect your winrate, the end result just shows the rank.

Now also factor in that gaining in a group is less then solo. Around 30 solo and 20 in party. so you can lose 3 times in a party to get -60 mmr, and win twice solo for +60 mmr. 40% winrate over 5 games with equal gains and losses.

So there is more complexity to it then just simple sayings/rules thrown around.

İlk olarak Ragnoraok tarafından gönderildi:

2) I already explained why that is wrong. Immortal players would have had high winrates to even make it to immortal to begin with. Once you reach immortal, some players are still more skillful than others, and winrates SHOULD reflect that but they don't. People in the top 10 should have higher winrates than everyone else, yet even Dendi has a 53% winrate. You are essentially arguing that everyone in immortal is equal to each other that that is objectively wrong....yet the matchmaker creates an output where everyone in immortal has similar winrates. The only people with really high ranks near 70% are pros who only play in premades with each other, which is the most purest merit-based game style out there.

I never said everyone at immortal is equal.
And i already explained how they ended up at that final rank will reflect on their winrate.

Pros never just installed the game and on day 1 were pros. They worked their way up. You are merely looking at the end results and totally ignoring what it took for them to reach that result. Anytime they got stuck at a skill level for any period of time their global wr wouldve started trending closer to 50%. Have you got a full history we can analyse to or are you just throwing around a stat that has 0 context. The context(how you got that stat) is far more important that just the final stat in determining how it came about.

İlk olarak Ragnoraok tarafından gönderildi:

3) If you are a 2,3, or 4 man premade then you will have solo que players on your team. I have the option to only play solo que matchmaking, and from to time I still get put with/against a premade. Also, premades themselves are not equal, and if anything have the geatest variance in skill. Premade matchmaking is the most pubstompy matchmaking that you will ever experience, and groups of peopl intentionally boost acounts so they can rank up easily by exploiting the terrible premade mm system.

Your in a premade queue which limits any pubstomp advantage, so effectively your facing what you are. If you just faced solos, yes it would be a huge advantage. But you dont.

İlk olarak Ragnoraok tarafından gönderildi:

4) There IS overwhelming proof, but you don't want to acknowledge it mainly because you think that that its normal for everyone's winrate to be around 50%. Why even quote a Valve employee? The first sentence alone refutes the entire point since the matchmaker fails terribly at creating even matches. Like in most of my games the 3 worst people in the lobby are on 1 team, when any functioning matchmaker would at least put 1 of them on the opposing team to make the game more competitive.

Overwhelming proof, but no proof listed, just opinion, feelings and what people notice and arguments to that effect.
Post the proof, if theres so much overwhelming and available proof go for it.

Me, im just basically arguing maths and how mmr systems work in general in many games. hell i even posted from someone responsible for it but like always theres an immediate way to discredit that person with opinion.

Simple fact is, there is no proof either way,
You obviously arent willing to change your opinion, and thats all it is. Unless someone breaks open the codes and many people can confirm that it is legitimate dota code and can understand it, there wont ever bee proof. Just opinion.

And what i say isnt proof, just how these ranking systems in general work out, and im yet to see anything in dota that bucks that trend.

Ive posted ways people can get ranks with lower global winrate, which is possible but not likely, but the fact it is possible contradicts a lot of what you say,. maybe i didnt word it right, maybe you just didnt understand those points.(which again isnt proof of the topic, just proof some of your arguements are laughable.

either way, luck and later. at a certain point you just have to give up arguing. Have fun believing whatever you want.

[Grimmz]

1 win 1 lose loop? oh man i wish. i got usually 5 losses until i get 1 win and my win rate is still at 56% which is kinda not accurate. i think i'm in forced 25%

[Katzenspeck]

be glad it's 50% otherwise scrubs like you would never win

[Zockolade]

You know if Jeff Hill from Valve is saying:

https://www.reddit.com/r/DotA2/comments/ppkz46/icymi_jeff_hill_explains_the_dota_2_matchmaking/

it must be true .. but what is more true than true is that Dota 2 is not fun anymore

[Dark Sun Gwyndolin]

İlk olarak Zockolade tarafından gönderildi:

You know if Jeff Hill from Valve is saying:

https://www.reddit.com/r/DotA2/comments/ppkz46/icymi_jeff_hill_explains_the_dota_2_matchmaking/

it must be true .. but what is more true than true is that Dota 2 is not fun anymore

That's why I mainly play Unranked, Turbo, and Ability Draft. I've admitted that I'll never be Top 100 Immortal.

Meanwhile, you make statements like "Dota 2 is not fun anymore" while continuing to play the game and 'pay' Valve by providing a noob they can use to 'rig' matches.

[Metal Arms]

İlk olarak Ragnoraok tarafından gönderildi:

İlk olarak uɐɐılʎʇs tarafından gönderildi:

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.


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.

Wouldn't immortal players have a 60% or at the least 51% wr against the bracket lower than their current rank? If most had a 60% in immortal they would force the skill level way up or increase the amount of immortals total. Thats where you get those below 1k, then those below 500 and then those below 100 who have vastly different skill levels in immortal alone.

[Ragnoraok]

@uɐɐılʎʇs
1) We are discussing the average person who is playing fair.. What you descirbed occurs when someone is boosting or using some other mm abuse. Otherwise, noone is mracuolously going from struggling in a low rank to a master against higher ranked players. Still, you do show that it is possible, so I acknowledge that I was wrong when I sted it was an impossiblity.

With that being said, global winrate is indeed an indicator of how well you perform throughout the history of the game. Someone who routinely makes it to high levels of rank should a substantially higher winrate than someone who does not, and this should scale based off the difference in player skill.

2) Except that in order to even make it to immortal you have consistently play better than the majority of the community. You correctly state that the pros didn't all of a sudden make it there, but you ignoring that they had to demonstrate a level of competency way higher than everyone else. To even make it to higher ranks you need to on average win more than you lose across the entirety of your game history.

The pros mainly play against each other, and in those closed systems we actually see global winrates above 60% because they mainly play against each other in coordinated games. so when the mm is barely a factor, we actually see a wide range in winrates which actually reflect one's performance. When the matchmaker is a factor, we see that the spread of winrate is small as though your ability to win is similar no matter how skilled you are

3) People in premades indeed face solo players a lot, even if the solo players choose to only play against other solo players. But in any event, premades inherently have the highest variance in skill since you typically have someone who is more experienced queing with a less experienced player. Thus, 2 premades playing against each other would almost always have more of a gap in skill than a lobby of solo que players. People exploit premade mm by smurfing and playing with lower ranked friends so that they can play against other premades where everyone is at their natural ranks.

4) You can still make logical conclusions based off observations. You don't necessarily have to have the code in front of you to know that something is broken. I personally believe that the matchmaker does not strive to create fair matches, which creates the observed effect of having a 50% winrate or some other threshold. The matchmaker creating more 1-sided matches than anything else is obvious, and is NOT something that other multiplayer games do on the regular. I am talking about the vast majority of games, like 90%, being 1-sided stomps. I have never experienced such a thing in any other MOBA, even Paragon which was in beta. DOTa's matchmaker

5) I am basing everything on my experiences. There was a point where I wanted to tank my behavior score by intentionally trolling and griefing in unranked. I was astonished that my winrate was still around 50% despite me putting so much effort into sabotaging my teammates. In those 30 games, the mm was effective at making sure that I won games by consistently putting the best people in the lobby on my team and the worst on the opposing team. Of course this isn't anything new, but it was disturbing to experience this as the "worst" player in the lobby.

[Ragnoraok]

İlk olarak Metal Arms tarafından gönderildi:

İlk olarak Ragnoraok tarafından gönderildi:

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.

Wouldn't immortal players have a 60% or at the least 51% wr against the bracket lower than their current rank? If most had a 60% in immortal they would force the skill level way up or increase the amount of immortals total. Thats where you get those below 1k, then those below 500 and then those below 100 who have vastly different skill levels in immortal alone.

Lets say that in order to get out of herald you won 6/10 games. Then to make it out of guardian you won 12/20 games. Then to get out of cruader you made it out of 22/40. So overall you won 40/70 games which is around 57%. The exact numbers don't matter, but if you are consistently ranking up, then your winrate has to be higher than the threshold needed to bypass a rank; that is the baseline that everyone has to meet. But when we are talking about people who are more skilled than the average player, then their winrate should be higher than the winrates of everyone else doing the bare minimum to level up.

As for immortal players, you should see that the more skilled immortal players have noticeably higher winrates than the average immortal player, but in reality we see that the winrates are more or less the same. The only immortal players that I see with absurdly high winrates are pros who only play against other pros so the system is enclosed.

[Metal Arms]

İlk olarak Ragnoraok tarafından gönderildi:

İlk olarak Metal Arms tarafından gönderildi:

Wouldn't immortal players have a 60% or at the least 51% wr against the bracket lower than their current rank? If most had a 60% in immortal they would force the skill level way up or increase the amount of immortals total. Thats where you get those below 1k, then those below 500 and then those below 100 who have vastly different skill levels in immortal alone.

Lets say that in order to get out of herald you won 6/10 games. Then to make it out of guardian you won 12/20 games. Then to get out of cruader you made it out of 22/40. So overall you won 40/70 games which is around 57%. The exact numbers don't matter, but if you are consistently ranking up, then your winrate has to be higher than the threshold needed to bypass a rank; that is the baseline that everyone has to meet. But when we are talking about people who are more skilled than the average player, then their winrate should be higher than the winrates of everyone else doing the bare minimum to level up.

As for immortal players, you should see that the more skilled immortal players have noticeably higher winrates than the average immortal player, but in reality we see that the winrates are more or less the same. The only immortal players that I see with absurdly high winrates are pros who only play against other pros so the system is enclosed.

Thats basically everyone though, the average at every skill level should be outpaced by the better of every skill level. thats why we have ranks and mmr to put people in different skill brackets. Pro have insanely high winrates because they play against mostly immortals who are below their own skill level. They don't actually play pros that often and when they do its mostly 1-2, not the whole team. There just isn't enough for great majority of games to have everyone below 100. Everyone below 100 is well above the skills of those above 500 and even more so than 1k+.

the variation of skill level among people on the teams, can greatly influence winrates.

[Dark Sun Gwyndolin]

Let's say we accept as true that Valve is rigging the matchmaking. The question is why? Is it to make people think they're better than they are so they don't quit? If so, it isn't working; 'forced fifty' conspiracy theories are endemic in the lower ranks. Is it to hold certain people back from greatness? Doubt it, Valve benefits from the truly great players climbing to the top, as pro players earn them massive income from Battlepass sales. Is to hide the effects of smurfs and boosters? Doesn't make sense; 'bad actors' take money out of the game (by making people quit), while spending nothing themselves (since they don't care about their account).

So, lets hear it. WHY is Valve doing this?

[Ragnoraok]

İlk olarak Technomancer tarafından gönderildi:

Let's say we accept as true that Valve is rigging the matchmaking. The question is why? Is it to make people think they're better than they are so they don't quit? If so, it isn't working; 'forced fifty' conspiracy theories are endemic in the lower ranks. Is it to hold certain people back from greatness? Doubt it, Valve benefits from the truly great players climbing to the top, as pro players earn them massive income from Battlepass sales. Is to hide the effects of smurfs and boosters? Doesn't make sense; 'bad actors' take money out of the game (by making people quit), while spending nothing themselves (since they don't care about their account).

So, lets hear it. WHY is Valve doing this?

Its because it keeps players engaged more than a regular merit-based matchmaking system.

Here is an article that discusses how matchmaking that is optimized for engagement is better businesswise than mm optimized for skill.

http://web.cs.ucla.edu/~yzsun/papers/WWW17Chen_EOMM

there are different ways to keep players engaged, but one of the easiest way is to make sure that bad players win more games than they should. Also, frustrating players into playing more games is better for player retention than having them stagnate and gradually lose interest.