|
Baa?21244 Posts
On September 29 2012 03:43 clickrush wrote:
requizen the games at 1.4k arent really different from 1.2k
having perosnally experienced 1.0k to 1.7k they all feel the same
the only thing different is that once you hit a new medal tier you get more rage. ie 1400s -> im silver im so much better htan this bronze scrubs im messiah
1500s-> shit im gold im in TOP THREE PERCENT OF LEAGUE PLAYERS IN THE WORLD bow down to meeeee
so a lot more rage around those 2 numbers
rest is all the same
|
On September 29 2012 03:39 Scip wrote:
Wait so what are we discussing here I'm slightly confused.
Is it that your elo gain slows down as you approach your skill level elo-wise?
Or that a lot of games are not enjoyable because of griefers/afkers?
I don't really have any idea what's going on and I think I'm not the only one
Both, I suppose. I'm frustrated at those types of games, and I blame many, but not all, of my losses on those types of games and players. I feel that the 1100-1300 rating, because of the high number of newly ranked players or people trolling on smurfs/alts, is volatile, and deserving of the "ELO Hell" moniker. There are more blatantly one-sided and troll/bullshit games at this level than higher or lower, from what I've seen on streams or spectating.
So my argument is that ELO hell does exist because of these types of games, and it is frustrating as fuck. It's also hard to get new players or other frustrated individuals like yourself in the right mindset to overcome a disadvantage, leading to more games that are lost that shouldn't have been. I'd be fine with a 50% W/L (about where I am now) if every game was populated by people trying hard and not feeding/giving up after 10 minutes, and I could actually influence games more than I currently am.
|
On September 29 2012 03:34 Requizen wrote:Show nested quote +On September 29 2012 03:25 bmn wrote:On September 29 2012 03:14 Requizen wrote:
Can a 2k player carry "lost" games at 1100? YES. I don't know how many times I have to agree with that before you stop thinking I'm arguing against it. But I don't have a 2k player's awareness, technical proficiency, or plain game sense. I just don't. I can't carry games where each damage dealer on the enemy team has 4+ kills over me, despite how well I did. I try, I don't think I've agreed to a surrender in about a month (except for double explosions).
[...]
If I was matched at 400 ELO for the next 100 games or so, I'd probably win 99% of my games too. Skill differential does that to a game. It's frustrating, as I said, to be at a point where you outskill the people you're playing with, but not enough to overcome the veritable mountain that your team puts in your way by feeding or making bad calls like roaming the enemy jungle by themselves when we can't see any of them. What is the evidence that you outskill the people you're playing with? They play with exactly the same feeding teammates as you do. Why are they at the same ELO if you're clearly better? Unless you're really the only person in your ELO bracket who isn't a smurf or trolling, it just doesn't add up. And of course you won't win 99% of the games if you are a 1400 player at 1300. But you don't need to, and that's not what everyone is talking about. Carrying yourself out of low ELO means that you were substantially better than everyone else there, not just a little better. And I still don't believe your assertion that your skill has no influence on the majority of your games, that sounds like a really lame excuse. My assertion wasn't just in the negative way. There are won games that have nothing to do with your skill level either, I've been in enough to accept that. Games where I'll jungle, none of my ganks will get a kill, and I'll be behind gold and levels because of it, but the enemy team throws lanes anyway, teamfights like shit, and loses. I've also had games where I dominate my lane in both CS and kills, but it didn't really matter because our AD got 11 kills to my 3 and single-handedly 1v3s bot and jungle because Vayne with a BT and PD can do that. In both of those games, my skill didn't matter, but we won anyway. And that, as well, isn't skill based, it's a mostly luck victory because the trolls and afks were on the other side. Those aren't fun games for me either. I don't bitch about lost games where it's close, we try our hardest, and the enemy team outplays us and gets an advantage. It's not a bad game, and it's a learning experience. Likewise, I don't gloat about games where the enemy team tries to single file into my fed Rengar. It's not indicative of my skill, and I didn't get anything worthwhile out of it. In fact, I'd rather those games didn't even count, because it's not even a win in my book. I don't want to prove myself against people that are playing like shit or intentionally feeding, and I don't want to sit back and accept a loss because those players are on my team. That's all I'm saying.
There's a huge difference between the statement "my skill has no influence on the game" and "we won but it wasn't because I played well".
The former means that your team would have the same *chance* of winning regardless of whether you're afk or playing like Forellenlord.
The latter just means that someone else on your team did well enough to compensate for your lackluster performance. Your skill still had an influence on the game, it just wasn't sufficiently good or bad to change the binary outcome.
I really don't understand what you are asking for. You're saying that your own skill level should decide the game, but the skill level of your teammates should not?
If you get better and worse teammates equally often, you're probably at just the right ELO.
If you want a game where the only thing that matters is your own performance, and not the performance of any other humans, multiplayer won't offer that.
|
People with less than 10ish games on ranked go in a separate queue, so the effect of new accounts and smurfs at 1200 isn't as strong as you might think (the truly new player who jumped into ranked too early should fall enough elo before they start queueing in the "real" ranked queue)
|
United States47024 Posts
On September 29 2012 03:45 Requizen wrote:Show nested quote +On September 29 2012 03:39 Scip wrote:
Wait so what are we discussing here I'm slightly confused.
Is it that your elo gain slows down as you approach your skill level elo-wise?
Or that a lot of games are not enjoyable because of griefers/afkers?
I don't really have any idea what's going on and I think I'm not the only one Both, I suppose. I'm frustrated at those types of games, and I blame many, but not all, of my losses on those types of games and players. I feel that the 1100-1300 rating, because of the high number of newly ranked players or people trolling on smurfs/alts, is volatile, and deserving of the "ELO Hell" moniker. There are more blatantly one-sided and troll/bullshit games at this level than higher or lower, from what I've seen on streams or spectating. So my argument is that ELO hell does exist because of these types of games, and it is frustrating as fuck. It's also hard to get new players or other frustrated individuals like yourself in the right mindset to overcome a disadvantage, leading to more games that are lost that shouldn't have been. I'd be fine with a 50% W/L (about where I am now) if every game was populated by people trying hard and not feeding/giving up after 10 minutes, and I could actually influence games more than I currently am.
That's not an argument for "Elo hell exists". That's an argument for "solo queue is frustrating".
Games don't get any less frustrating at higher Elo. The idea of Elo hell implies that this is somehow confined to some specific region of the ladder, when its universal to solo queue.
I think Tree once said something along the lines of "As you go higher up in Elo, the players you play with get better, but you get better too so you still think the mistakes they make are retarded and everyone is terrible".
|
On September 29 2012 03:45 Requizen wrote:Show nested quote +On September 29 2012 03:39 Scip wrote:
Wait so what are we discussing here I'm slightly confused.
Is it that your elo gain slows down as you approach your skill level elo-wise?
Or that a lot of games are not enjoyable because of griefers/afkers?
I don't really have any idea what's going on and I think I'm not the only one Both, I suppose. I'm frustrated at those types of games, and I blame many, but not all, of my losses on those types of games and players. I feel that the 1100-1300 rating, because of the high number of newly ranked players or people trolling on smurfs/alts, is volatile, and deserving of the "ELO Hell" moniker. There are more blatantly one-sided and troll/bullshit games at this level than higher or lower, from what I've seen on streams or spectating. So my argument is that ELO hell does exist because of these types of games, and it is frustrating as fuck. It's also hard to get new players or other frustrated individuals like yourself in the right mindset to overcome a disadvantage, leading to more games that are lost that shouldn't have been. I'd be fine with a 50% W/L (about where I am now) if every game was populated by people trying hard and not feeding/giving up after 10 minutes, and I could actually influence games more than I currently am.
You CAN influence games a lot. I'm at 1300 (stuck at 1200 for a long time), there's nothing hellish about it.
If I play well, I win a lot more than I lose. If I play badly, I lose a lot more than I win. It's always very obvious in retrospect.
In many games I feel like I played well and we still lost. That is fine. It's a team game, you are one of ten players, whether you play well or not is not the only thing that matters.
I've seen several smurfs in games I played. They were good enough to carry almost all of those games with ease, including more than one 4v5. I can't carry these games. That's because I'm not actually better than other 1300 ELO players.
|
On September 29 2012 03:47 thenexusp wrote:
People with less than 10ish games on ranked go in a separate queue, so the effect of new accounts and smurfs at 1200 isn't as strong as you might think (the truly new player who jumped into ranked too early should fall enough elo before they start queueing in the "real" ranked queue)
I have close to 400 games and slingshot between 1100-1204 in the past couple months. Last week I had a game with a teammate who had SIXTEEN games. Am I where I deserve to be? Most likely. Are they? Most likely not. 10 games with a +39/-34 split is NOT a valid beginner gate.
|
Well, it's easiest for smurf to get out of elo hell since they get like 50 elo by game.
|
Are we really doing the ELO hell thing again?
If you win you get higher rating and if you lose you get lower rating. If you are better than your rating then you should do well and win more often. I don't see the problem.
|
On September 29 2012 03:52 WhiteDog wrote:
Well, it's easiest for smurf to get out of elo hell since they get like 50 elo by game.
I meant people smurfing on low ELO accounts, not people playing on new accounts. If you win consistently, you'll get out either way.
|
Anyone know the details of the Riot elo system (and how it differs from chess, for example)? Let's Assume a standard system. Let's assume that overall team elo = average of individual elos and whatever other standard assumptions. Not exactly true, but close enough.
If you're actually 1400, and your in a game of all 1200s, then you're expected to win ~55.7% of games. Which is to say, your skill advantage is enough to turn around 1 in 18 games.
If you're actually 1800 in a game of all 1200s, then your win rate should be ~66.6%. You still only change the outcome of 1 in 6 games.
If you're 2400, win rate goes up to ~79.9%.
So yes, you're mathematically not expected to affect the outcome of the majority of games. It is in fact impossible. Because you're expected to win about half your games anyway.
Even at +600 points above your level, you're only expected to turn 1 in 3 losing games around. Elo hell is really just another way of saying "doesn't understand math"
|
Solo Q is jsut a matter of finding the least braindead guy on your team and the most braindead guy on the enemy team. Stick with your reliable ally and farm the shit out of their weakest link.
I play support, and ADC is rarely the best guy on my team. So I do what I can for him during lane phase, then I tag along with the smart guy afterwards.
|
On September 29 2012 04:01 azndsh wrote:
Anyone know the details of the Riot elo system (and how it differs from chess, for example)? Let's Assume a standard system. Let's assume that overall team elo = average of individual elos and whatever other standard assumptions. Not exactly true, but close enough.
If you're actually 1400, and your in a game of all 1200s, then you're expected to win ~55.7% of games. Which is to say, your skill advantage is enough to turn around 1 in 18 games.
If you're actually 1800 in a game of all 1200s, then your win rate should be ~66.6%. You still only change the outcome of 1 in 6 games.
If you're 2400, win rate goes up to ~79.9%.
So yes, you're mathematically not expected to affect the outcome of the majority of games. It is in fact impossible. Because you're expected to win about half your games anyway.
Even at +600 points above your level, you're only expected to turn 1 in 3 losing games around. Elo hell is really just another way of saying "doesn't understand math"
I'm pretty sure the numbers will not work out like this at all. You make several assumptions that aren't warranted, and saying that disagreeing is "not understanding math" sounds facetious.
There's no reason to assume that the team mean -- as opposed to the median, max, or min -- is the most relevant number.
It's not just one entity playing against another; the 1800 player can exploit mistakes by any of the 1200 players on the other team, and he can help any of his 1200 players make better plays (calling dragons, ganks, etc.). And due to multiplicative stats (and snowballing) in the game, advantages aren't linear either, so it's particularly easy to carry advantages to the end if you also are more skilled.
|
On September 29 2012 03:54 RebelSlayer wrote:
Are we really doing the ELO hell thing again?
If you win you get higher rating and if you lose you get lower rating. If you are better than your rating then you should do well and win more often. I don't see the problem.
Yea I'm surprised people don't realize how simple it is.
|
On September 29 2012 04:03 ticklishmusic wrote:
Solo Q is jsut a matter of finding the least braindead guy on your team and the most braindead guy on the enemy team. Stick with your reliable ally and farm the shit out of their weakest link.
I play support, and ADC is rarely the best guy on my team. So I do what I can for him during lane phase, then I tag along with the smart guy afterwards.
How is that working out for you in terms of your elo-level and enjoyment?
Just curious, I get bored if I have to support more than 4-5 games in a row.
|
On September 29 2012 03:44 NeoIllusions wrote:Show nested quote +On September 29 2012 03:42 Gahlo wrote:On September 29 2012 03:39 Scip wrote:
Wait so what are we discussing here I'm slightly confused.
Is it that your elo gain slows down as you approach your skill level elo-wise?
Or that a lot of games are not enjoyable because of griefers/afkers?
I don't really have any idea what's going on and I think I'm not the only one How reasonable it is to expect a player to carry every game 100 elo below their hypothesized true elo, despite any actions their teammates make. Not unreasonable at all. In reality, he probably can't carry all the games but a hefty majority should not be a prob. Granted you will need a large sample size of games. Triple digits easily.
I think it's a lot less than you think. Mathematically, it's closer to turning around 1 in 17 losing games. Of course this makes a lot of assumptions. You can affect outcomes more in certain roles than others, for example. In the long run, it's noticeable, but it's probably at a level that most humans are incapable of distinguishing from luck or "elo hell"
|
On September 29 2012 04:08 bmn wrote:Show nested quote +On September 29 2012 04:01 azndsh wrote:
Anyone know the details of the Riot elo system (and how it differs from chess, for example)? Let's Assume a standard system. Let's assume that overall team elo = average of individual elos and whatever other standard assumptions. Not exactly true, but close enough.
If you're actually 1400, and your in a game of all 1200s, then you're expected to win ~55.7% of games. Which is to say, your skill advantage is enough to turn around 1 in 18 games.
If you're actually 1800 in a game of all 1200s, then your win rate should be ~66.6%. You still only change the outcome of 1 in 6 games.
If you're 2400, win rate goes up to ~79.9%.
So yes, you're mathematically not expected to affect the outcome of the majority of games. It is in fact impossible. Because you're expected to win about half your games anyway.
Even at +600 points above your level, you're only expected to turn 1 in 3 losing games around. Elo hell is really just another way of saying "doesn't understand math" I'm pretty sure the numbers will not work out like this at all. You make several assumptions that aren't warranted, and saying that disagreeing is "not understanding math" sounds facetious. There's no reason to assume that the team mean -- as opposed to the median, max, or min -- is the most relevant number. Demonstrably players will often make decisions that others follow. It's not one entity playing against another; the 1800 player can exploit mistakes by any of the 1200 players on the other team. He doesn't have to do it himself; he can make suggestions to other players (e.g. calling baron after you died; asking people to watch out for a gank you expect to happen, etc.).
Whatever, you could perhaps suggest a better alternative than the mean. I don't think it's that terrible of an assumption, especially when the difference is small. Making helpful suggestions to other players affects your "true elo" and thus is already factored in.
|
On September 29 2012 04:03 ticklishmusic wrote:
Solo Q is jsut a matter of finding the least braindead guy on your team and the most braindead guy on the enemy team. Stick with your reliable ally and farm the shit out of their weakest link.
I play support, and ADC is rarely the best guy on my team. So I do what I can for him during lane phase, then I tag along with the smart guy afterwards.
Imo the thing you should abuse at lower ELO's (dunno which supports you prefer) is that both ad carries usually have weak cs, so focus on getting his tower early and try to roam around (might be hard if he is clueless and dies 1v2 bot) and try to set up potential fights near drake or invade with your own jungler.
|
On September 29 2012 04:01 azndsh wrote:
Anyone know the details of the Riot elo system (and how it differs from chess, for example)? Let's Assume a standard system. Let's assume that overall team elo = average of individual elos and whatever other standard assumptions. Not exactly true, but close enough.
I think Riot uses a different distribution than standard logistic; graphs of percentile elo rankings look log-normal.
|
God I hate how they came out with Rengar and Khazix consecutively. They are way too similar. And I really think that the "Head of Khazix" thing needs to be taken out. No other hidden passive is that game changing especially over something so small. Just buy bonetooth when he hits lvl 16. Co-ordinate your team for a huge suicide gank on him. Even if half your team dies you just got permanent full bonetooth stacks. Yeah that's retarded IMO.
|
|
|
|
|
|
EPT
