General DiscussionELO HELL

Actually wink isn't in elo hell. Confirmed 3k trash for real.

sry but i acually did the test aswell, its legit

Elo hell is real ... after i get some wins .. enemies are not getting stronger. Team get stupidier picking idiotic heros and feeding ... Like allways. Same was with PD experiment - sooner or later he got such noobs he was 40/4 and rest of team 0/9-14 ;D

ps. im better then previous posters :d

What makes you think you're better than the previous posters?

My winrate is only from solo matching while they all played in same stack ;D Just check buddy list. Stacking players have no voice vs elo hell.

I don't like explaining obvious facts to stubborn morons like OP. Melody, Sam, Havoc, or any other smart dotabuffer, can you do it plz :3

"/ps this playdota experiment proves nothing !"

please elaborate, how does it not prove anything? give me your acc and i'll happily boost you for 10 keys / 100 mmr (payment first)
im not verified on epicnpc anymore because ihavent updated my bank card info there

The playdota experiment prove nothing, but maths does

If you are in elo hell, the probability of having total uncarryable trash is higher in opponent team then in your's, as they are five potential people for it, while you only have four in your team for it.
So basically, if you are in "elo hell", as long as you are either 1st/2nd/3rd/4th/5th best player of the that particular game, probability says that your team will be better than opponents' one.

Which ofc is contradictory with elo hell theory, so elo hell doesn't exist.

Edit for the guy just before me, the playdota experiment proves nothing, cause it's just one example.

It's like if you were saying "adding two numbers will give you a bigger number", and showing as example 2+2=4;
it just missed the 2+(-2) or -2 + (-2) for instance

^Though the playdota experiment isnt the only one, its just the only one that got publicity

Lets not forget the fact that people boost others' accounts for keys (wave is an example)

we also did the Hael experiment on DotaBuff too

An infinite number of experiment doesn't prove anything.

With the same example i used before
2+2
2+3
2+4
etc...
you can go on infinitly, you will still miss the case in which it decreases.

There's only one way to prove by experiment something, proving every single case (which was used once to find the littlest number of number you need to have at the begining of a sudoku grid for it to have an only solution).

Now do you feel that those few experiment were showing everysingle case possible ? definitly not

"So basically, if you are in "elo hell", as long as you are either 1st/2nd/3rd/4th/5th best player of the that particular game, probability says that your team will be better than opponents' one."

Claniraq proved that for every team there is finite number of elo. Elos of teams should be matched as close to 50/50 rate as possible. Currently if there is a stack of 4 ppl that have 3 bad and 1 good player their elo will balance counting them as one player with medicore elo. Then if enemy team has 5 decent players to compensate for 4 man stack they have to get 1 very good player. Then elo will be falsely "balanced". It is possible for a very good player to overcome such a handicap but it doesnt have to be the case.

Elo hell for highly skilled player will kick in sooner or later. Here is a DB of player with 5k mmr that was about to prove elo hell doesnt exist. http://dotabuff.com/players/110289963 currently he is only losing - A LOT.

There is also finite number of possibilities where handicapped player is able to overcome difficulties. Thats why elo hell for players in 3-4k is mostly seen.

Dota system was created to balance games for solo players to 50/50 rate. Thats why sooner or later solo players hit a brick in the wall. This doesnt work vs stacks, bcoz organised group will always have adventage over solo players. Which results in stacking for higher winrate ;D

If not, then why there is no solo queue :) ? There could be 3 queues for solo, duos/tries, 5man. But there is non. Because balancing elo hell for 50% would be harder ;D

Conclusion ? Go stack! There will always be enough imigrants from LoL and HoN to pwn :)

Then where is elo heaven ?

"pojzon"

ur entire comment gave me cancer, do me a favor and dont write again

There is no such thing as ELOHell

I raised my MMR by 800 in one week and occasionally fed because sniper = lost game
I still raised it despite morons,peruvians, and myself being on my team

Now im sitting around 4.8k laughing at this thread

PS: never play slark again please

Another one of these threads.

elo hell is knowing that you are going to feed with sand king yet still picking it anyway

Elo hell is getting Anti Mage, Lone Druid and Morphling on the same team while telling yourself you can still win this. You can freely add 2 of most useless heroes in your opinion, and you get the picture.

yes, eventually your team gets worse to compensate for you. sadly, godlykha/claniraq has never reached tbat level of play. and for those who have, they certainly would prefer it that way over fucking free wins.

his proposed model was retarded and failed to solve anything. imagine if matrice solo queued and the 9 next best players are 4.7-5k, the current system will give him the 4.7k players and the other team 5k players. despite that he's still climbing. claniraq thinks basically the ten players in a game should be randomly distributed, which results in everyone above a certain skill level getting auto wins thanks to the absurdity of it, due to their team always having the higher mmr. he justifies it by saying that the game is unfair that the best players in the world have weaker teammates and that the game is unfair for them - funnily it doesn't concern him at all.

From people having 1k mmr difference playing the same, to people not playing for months and not living up to their mmr, to people getting out of normal matchmaking and barely maintaining a 50% keeping them at a mmr that is higher than what people "Think it should be", to people who are 3.5k playing better than 4.5k, and vice versa. It is a flawed system. You are turning a team game into a solo queue one man army tactic that is completely unnecessary. 5k-6k players usually have a big difference between them in consistency and swiftness of execution. But anything under 4700 ish is a wash out. I've even had games were I would get all towers and have 20 kills in the first 20 mins trying to win the game for the team. They would complain that I am "Trying to be a hero" and they would spend their time counter acting my play. MMR doesn't account for that. People who think the mmr system is fine are part of a typical human nature phenomenon where they accept what they have and justify it with good points they heard from anonymous sources.

They could of even had the players on the losing team that resulted the most TD, HD, XPM, GPM or vicinity of kills and team fights to give them +5-10 points rather than the whole team losing. No one is going to win every game and it shouldn't force you to win every game. Because not even that, but if there are people trying to get their MMR up just like you that is a conflict of interest if you are on opposite teams.

You queue up for solo queue to be collectively judged as a team. Only to be simplified into radiant or dire victory.

Don't even get me started trying to explain why the CS:go system doesn't work with Dota's

^ you said you just started dota 2 after lol, and stil having hard time,but here,we see, you are such smart person .

Or shall i call you Hael?

New acount? Well you had to get it, to continue tradition.

Nah man, I main Blitzcrank

Let me get some Tl;Dr on the forum here.

MMR is a complete (mostly) 4 digit number that shows your winrate in ranked pub matches. It shows nothing else. Most games have a symetrical foundation on which things start from so a win/Loss rating makes more sense. In Dota, the fate of games rely way too much in players hands.

Yes, picking a hero that can handle tough situations and manipulating them in your favor does work (to an extent ). MMR does not measure the attitude and dedication of other teammates and could make your work seem countered. Not only that, When you tell multiple people to do certain things in a certain situation you are causing a conflict of interest if those two happen to be on different teams. Sometimes you might be on the same team, but that is where the luck factor comes into mmr rating. Also, if people have stated that people on your team are retarded and its your job to be smart and play correctly to increase mmr, you are increasing your mmr for them as well then they don't deserve it, putting them in a precarious position or a lucky one. so your solo efforts can hinder other or help others.

But see its called solo queue in a team oriented game, and we have some weird ass anti-social system where Solo and party are separate things. Why couldn't you just have a single ranked MMR? People complain that it would be too easy to boost it up where you're not suppose to be. But I feel that would just make the rating of ranked a bigger dynamic. Higher up players would want to really differentiate themselves from the others. People already boost up the account through mmr boosting and account buying. People were partied together and playing a lot of solo ranked matches anyways. You should build a system giving incentive to not be a team player. You could also say, since you were anyways, "Well people who buy accounts just go back to their normal mmr". Yea, and more accounts are bought and new ones pop up. Its going to be an ongoing thing because of the ills of F2P gaming. You can't have a system this simplistic on a f2p game that requires little to get in. The mmr only predicts people roughly and people don't hold onto that title for long. The game is volatile with both players and patch updates. Not everyone that plays dota is an avid dota player even 4700+ players. I know some 5k players that play once every 3 weeks and forget things.

If this game is about community then why would it not be controlled by the community? How hard would it be to make it just like Dota 1 where its a user game created community that picks and chooses their battles. Thats how friendships are made and a real community is born. There is hardly anything to do with guilds and the interface is a lonely place where you watch games, look at items, or press the play button. majority of the interaction is through steam. Why would people want to subject themselves to oppression of a potential abuser, griefer, mute system, low prio system, and no concede feature because (they don't trust people to do their own thing).

Call me old school, but with games that are competitive like this, I would chose how I did it on Starcraft: Brood War. I would join a channel or Guild Channel and ask for 1v1s. If we had this type of Irc type guild mentality, you could challenge people left and right. Make a game, play it, and even REMATCH if you want to do a fucking bo3. Holy shit, how cool would that be. You could even manage to use the takeover feature to repeat a teamfight that happened in the game to learn something. But we are back to square one with some fat humans eating off what they are given. We shut up and take it in the ass, while we worry about rating (Solo only) cause the party rating isn't a number thats meaningful enough to make your dick look bigger.

i found hael guys

i agree though the wc3 interface was better, but still if the only way to play dota was by making/joining a lobby, MMR would be pointless

also can you stop writing walls of text