http://www.youtube.com/watch?v=MxYsi5Y-xOQ

; )

But seriously, elo system works in the vast majority of cases- only a few incredibly, incredibly unlucky people actually play at enough inorganic disadvantage (afk's leavers, etc) to prevent them reaching the right ELO. If you want to get out of hell, learn from your mistakes, accept what you are doing wrong, and resolve that you will do better. Often, players complaining about ELO hell cannot even do this, because they lack the mechanical skill to use their champions or the game knowledge to even approximate an informed decision. But, they blame teamates, because their lack of knowledge is occlusive to their ability to evaluate it (Dunning Kruger effect).

If you really want to get out of hell, step out of ranked, practice, practice, practice until you can stomp normals, and then try again. You will be pleased with your results I think, and by that I mean you may just manage to stay above 1200. If there is anything people have misconceptions about, it is that everyone can reach a higher elo. No, because ELO is symetric. Most people should be at the average.

This is again wrong. High elo players will take advantage of every little mistake you make. They do not trade mistakes. You mess up, they will take advantage of it. They will take advantage of things you didn't even know we're mistakes.

There isn't a play style difference between the higher elo and low elo. It's how you capitalize on mistakes being made. If a 1900 Ap mid plays like. 1500ADC he will get jumped on at every mistake possible.

The only saving grace you have is you are assuming the enemy jungler is playing smart and is able to pick up the slack, forcing the lane to be a bit more passive from the fear of ganks. You can bet though when they have ward coverage if you step out too far they will take any little edge they can get


There are 2 aspects to this. First I didn't communicate my though correctly. What I meant to say was I have never seen a more accurate system for judging individual skill(over a period of time) in a 5v5 game, than in the elo system we have here.

Furthermore, no system will ever judge individual skill in a 5v5 environment better than in a 1v1 environment even if the system was designed to evaluate individual performance during 5v5 games it will still more accurately and more quickly get a reading in a 1v1 environment. Once I offered a $50 riot card to disprove that idea. But it was a trick question. It simply goes against game theory. 1v1 will always be faster and more accurate. That doesn't mean you can't get a reading in 5v5, just it will take longer to isolate your impact, and there will always be a little noise as far as pinpointing your rating. But it can figure out your rating within 50 points. So it's not broken.


I'm not gonna go into the game theory of finding you chance of winning based on how you develop an edge in that situation. You have constructed a situation that it's safe to say you will lose(specifically the part that says the opponents are equal skilled)

What's your point though? Is this your proof that elo isn't a good representation of your skill?

I'm going to teach you a poker term today known as EV. Expected value. In short, its how much you are expected to win in certain situations, over a period of time if your big hands get drawn out against weaker hands, say you go all in with AA 5 different times and each time you are against a weaker hand (55, 94, JJ etc) and you lose all 5 times. This is known as running negative EV, -EV. same thing the other way. If you win a few times when you should of lost more often that is running +EV. Over time this EV will always balance itself out, when you can't control any more action, you will lose 20% of your 80/20 splits. Win 60% of your 60/40 flips. Over a short period you might be +EV or -EV but you will always end up at EV(in poker it generally takes about 200,000 hands to have an accurate reading of your skill. To balance EV out)

Here's how it applies to league. Your teammates and opponents you have no control on whether or not their internet dies. House is on fire, baby fell out of the tree again, whatever. You get into a game and whatever is gonna happen will happen. This might cause you to unfairly lose elo in that game.

This is known as running below EV, or running -EV. In the short term you can have some streaks where you get a dc on your team a few times. You can also lump trolls, spontaneous ragers, god awful players(truly god awful. Not just you are upset he lost lane and you weren't able to catch him up) into this system of EV. And you simply have to accept it will balance out. Keep track of how many games below EV you are if you have to. But over a 100 games you wiki see you will be pretty close to EV as far as that stuff goes.

The one big difference between poker and league. While you will always have an equal number of bad beat beats to suck outs(percentage wise). This is not the case in league. Because you control an aspect(yourself) if you never dc, rage, troll. Play like Helen Keller, you will always be +EV. Bad things over the long run will to them more than you. Causing you to gain elo from the total pool of dcs

I'm going to go out on a limb and say you misread that part of my post. I never said high elo players "trade" mistakes, I said they capitalize for profit on them. Low elo players "trade" mistakes, and thus, the difference in play-style becomes apparent, you'll find more low elo players "trading" instead of farming, watching their map, warding, the list goes on. High elo players don't do this and we have both stated this from the beginning.

Again, we're not on the same page, but since I sense that you seem to be trying to understand where I'm coming from, I'll start over and try to be clearer. My main qualm, while related to accuracy, is not with the accuracy of the system. My attempt with the hypothetical question was to get you to see what it looks like for players at 900-1300 elo. The scenario I put together is a common occurrence at this elo and it should not be. This does not inhibit people of significant skill(or lack thereof), if you belong at 1800 elo you will likely reach that point quickly, if you belong at 700 elo it works the same way. The problem is, if you're at 900-1300 elo and you belong at 1400, or if you belong at the elo you're at, but you want to get better and raise your elo so that you're around players who also trying to improve, this scenario will often inhibit you. I will repeat, it should not inhibit you.

I will ask you to answer this, and only this instead. Should the system allow you to be dragged down by players who to put it fundamentally, are not playing? If the answer is no, then we are in agreement and the system needs "tweaking". If the answer is yes, then we are in disagreement and shall glare angrily at each other until the other submits.

If you actually ready his post that you just replied to, and understand any basics of statistics (which he just kindly explained to you in kid talk), then you know that this question is irrelevant, and those people aren't actually bringing anyone down in the long run (100 games or less, most likely, even).

High elo players are making less mistakes, you can call it not trading mistakes or whatever, you don't play different at low elo than you do at high elo, you wait for mistakes and capitalize, you just might have to be a little more patient for those mistakes at higher elo

This concept does exist. Not near the degree you think though.

The thing is, if you are 1400 caliber, you just won't get stuck at 900, or even 1200. I know this because I experienced it as I climbed the ladder from 1100. Whenever you drop too low. You'll begin winning at a very high rate.

To get within 200 points of your elo is very easy. It doesn't take much time at all due to the higher influence you have when you are outside 200 points.
Even getting within 100 points does not take long. You have significant influence still, less than before but you will still have at least 60% winrate.

As you approach your true elo, it becomes less and less about your skill and more towards the luck of your teammates. This is why a 1400 player will hover 1350-1450, maybe 1325-1475. But this is natural in any ranking system.

This question is provoked by a fundamental flaw in your understanding of game theory, what you are asking about is called "noise" . The term covers a lot of different things. But what you described is an instance of noise.

What will happen as you develop a better understanding of game theory is that you can't take a certain amount of noise away from the system. It is a natural byproduct of a 5v5 environment. At a certain point when you continue to tweak and tweak the system, you will cause more damage and unintended consequences from the tweaks than you solve from the tweaks

if your good, you have bad teammates, if you bad, it gives you good teammates. this is the truth, everyone will contradict me, but its a fact.

i got 2 irl friends of mine who play with locked camera, one of them only jungles, when he does, he doesnt check lanes HP/items/summoners/wards, he just walks in with locked cam hoping gank will work, which half the time it doesnt, he goes negative everygame, his ranked stats are all negative. and he is at 1560 elo. why? he is a very bad player, but when i watch him play he is always with players who go 10-0 in their lanes. As where I think 9/10 rank games i've done a lane loses its tower in 5 mins and already has 6-8 deaths

What I have gotten from this thread is , that this is not the first time this has come up. I also see there are some intelligent individuals taking the time to reply to your concern. What makes you different than me is that you're not taking the time to read the rational. I am an average skilled player, with bad league habits. Habits which hinder me from leaving the elo I am in. My team game/ward-map awareness and so on. I could say I mush bronze players and can get good k/d, which could be intact true. The element that seems to miss you is that there's a higher elo; that other players are actually in. That higher elo are of those of a better skillet of mine and the players I play with. If I get better, or better if I get to the same skill of those higher elo players, it's statistically inevitable that I reach their elo and join them in matchmaking. There will be "noise", and my own mistakes that may drop my elo, but my skill will always bring me to the ranking I deserve to be in. That's how the curve works.

In short, all players are facing the same obstacles we hate, yet there still exists this higher elo, which we aspire to reach. To reach it we must become better players.

Selcopa seems more than willing to help you analyze your games and help you reach the plateau you so crave. I hope you don't squander that opportunity.

I've actually observed this. Not so much in lol but rather on the forums.

Because I have good ideas. And the forums keeps sending me idiots with bad ideas and no understanding of game theory.

And the bad players come and me(and other good players) match up with them and carry their thread

-Off Topic- @Selcopa, You mind adding me. I definately need help with my game. I know what I am doing wrong most of the time... but there's those games that you speak of carrying where I have no clue what to do next.

ELO isn't perfect, but if you can't dig yourself out of ELO hell, you deserve to be where you are.