| Forum Index > LoL General |
Same rules apply, per usual. Please use the appropriate threads (QQ, Brag, Champion, etc) whenever appropriate. Keep the resident Banling content. Thanks. Happy Gaming. | ||
|
robertdinh
803 Posts
On March 20 2012 00:31 TheKefka wrote: That exists only if you played a small amount of games. Elo hell does not exist,if you are better than the people you play against eventually you will rise no matter what sooner or later. Comil has played nearly 1000 games... | ||
|
ManyCookies
1164 Posts
| ||
|
TheKefka
Croatia11752 Posts
I have no idea who that is. | ||
|
robertdinh
803 Posts
Well he is a player that after nearly 1000 games is at 1600ish rating right now, he was at 2100 a short while ago, and in skill is probably about 1300. | ||
|
BlackPaladin
United States9316 Posts
On March 20 2012 00:06 Two_DoWn wrote: Thats like saying the garbage man is a chef. Well that would be like going to a McDonald's so it's not a huge stretch. | ||
|
LightRailCoyote
United States982 Posts
Because fuck the rare statistical outlier, you're right1 | ||
TheYango
United States47024 Posts
On my main I was 1400-1500 with 800 games played with Annie in ranked. When you buffed her I shot to 2075. Then after your last Nerf to her I have went down to 1700-1800. While I agree with the sentiment that acknowledging Elo hell is not really helpful to anyone, it's a bit unfair to robertdinh that we all had a good laugh at this when it was posted 2 weeks ago, and then pretend it doesn't exist when this kind of discussion arises. | ||
|
Sabin010
United States1892 Posts
| ||
|
robertdinh
803 Posts
On March 20 2012 00:37 LightRailCoyote wrote: Hold on. You've found 1 example of a player who fluked his way up to high elo, and are using that example to try and prove your argument? And try and belittle everyone at the same time? Because fuck the rare statistical outlier, you're right1 Is a system where a 1300 caliber player can spike up to 2100 accurate? What about every case that is less extreme but is still inaccurate? What about the fact that this "outlier" was skewing results for many games up and down across all elos? How often will people run into a case like that to "even" it out? Is 1000 games enough to determine an accurate elo? Why is this guy still fluctuating so wildly then? This guy is an example of why claiming elo hell doesn't exist, and trying to deal in absolutes like claiming X number of games will get you near your real elo, is silly. | ||
|
Two_DoWn
United States13684 Posts
On March 20 2012 00:37 TheYango wrote: For reference again: http://www.reddit.com/r/leagueoflegends/comments/qg3pb/level_130_experimentstatscomplete/ While I agree with the sentiment that acknowledging Elo hell is not really helpful to anyone, it's a bit unfair to robertdinh that we all had a good laugh at this when it was posted 2 weeks ago, and then pretend it doesn't exist when this kind of discussion arises. Except for the fact that nothing in that post has anything to do with the rate at which someone gets trolls and leavers. Its just about someone being good at one champion and sucky with the rest. In the context of this argument he is a 2k player when he gets the handicap of playing an overbuffed champion, same as how I would be 2.4k if Shyvana never died. | ||
|
mr_tolkien
France8631 Posts
On March 20 2012 00:35 BlackPaladin wrote: Well that would be like going to a McDonald's so it's not a huge stretch. Mac Donald Yorick. Flipping burgers with his shovel. | ||
|
ManyCookies
1164 Posts
On March 20 2012 00:38 Sabin010 wrote: What if a 400 Elo player gets carried up to 2k, and the superior play of the 2k players he is matched up with and against inspire him to become the 2k player he was capable of being? Off topic (thank god), but it really does help playing against stronger opponents who can actually punish your mistakes. One of the best ways to improve in Starcraft is to play against people 1-2 leagues higher than you, so your mistakes have a corresponding punishment and subsequent immediate feedback. | ||
|
fishinguy
Russian Federation798 Posts
| ||
|
mordek
United States12705 Posts
| ||
|
Mogwai
United States13274 Posts
On March 20 2012 00:35 robertdinh wrote: Well he is a player that after nearly 1000 games is at 1600ish rating right now, he was at 2100 a short while ago, and in skill is probably about 1300. I'm going to postulate that: 1. he's not skillwise a 1300 and 2. there's a perfectly logical explanation for his drastic rises and falls in elo I could just say, well, look at bigfatjiji, he's 2.2K, but he's a 1400 skill player, but we all know that's bullshit. because this guy you're talking about is obscure, we have no idea how skilled he is or why his elo bounced around, but I refuse to believe that he just lucked his way into 800 elo. assuming let's say 15 points/game, and that for no good reason his team has a 50% chance of winning for him (should be lower if he's at an artificially high elo), the statistical chances of him ending up 800 elo up at any point in time are 1 in 9,007,199,254,740,992 (technically worse, but who cares). Statistically, I don't buy it. | ||
|
spinesheath
Germany8679 Posts
Take a couple of math classes. The probability of him being a 1300 player who goes up to 2100 and stays around that elo for hundreds of games is roughly... 0.0000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000000001. Well, maybe he actually got THAT lucky. But I don't care about it then. Sometimes, unlikely things happen. Doesn't mean the system has a significant flaw. | ||
|
NovaTheFeared
United States7229 Posts
On March 20 2012 00:40 robertdinh wrote: Is a system where a 1300 caliber player can spike up to 2100 accurate? What about every case that is less extreme but is still inaccurate? What about the fact that this "outlier" was skewing results for many games up and down across all elos? How often will people run into a case like that to "even" it out? Is 1000 games enough to determine an accurate elo? Why is this guy still fluctuating so wildly then? This guy is an example of why claiming elo hell doesn't exist, and trying to deal in absolutes like claiming X number of games will get you near your real elo, is silly. Elo hell is disproven by mathematics, but tilt is real. | ||
|
Slayer91
Ireland23335 Posts
On March 20 2012 00:46 mordek wrote: I take Slayer's stance and claim that I'm terrible instead of judging others ELO. Only difference is I'm terrible and he's good XD I'm actually terrible its just that people aren't very good at this game right now, it's relatively new. | ||
|
robertdinh
803 Posts
On March 20 2012 00:46 Mogwai wrote: I'm going to postulate that: 1. he's not skillwise a 1300 and 2. there's a perfectly logical explanation for his drastic rises and falls in elo I could just say, well, look at bigfatjiji, he's 2.2K, but he's a 1400 skill player, but we all know that's bullshit. because this guy you're talking about is obscure, we have no idea how skilled he is or why his elo bounced around, but I refuse to believe that he just lucked his way into 800 elo. assuming let's say 15 points/game, and that for no good reason his team has a 50% chance of winning for him (should be lower if he's at an artificially high elo), the statistical chances of him ending up 800 elo up at any point in time are 1 in 9,007,199,254,740,992 (technically worse, but who cares). Statistically, I don't buy it. Ok but just because you don't want to accept his actual skill level because it goes against what you believe doesn't mean that he isn't actually 1300 caliber in skill. I mean some of the arguments people have tried to make is that playing many games gets you close to your real elo, and that the system isn't inaccurate. I am showing you a case of a player that is currently 1600ish with almost 1000 games played, whose top elo was 2100+, and from personal experience I will say that he is actually about 1300 in skill. You can choose to make assumptions as to why this is the case, but at this point you are just being obtuse if you can't consider the fact that the system is just grossly inaccurate for some players, and very accurate for some, and everything inbetween. | ||
|
TheYango
United States47024 Posts
On March 20 2012 00:40 Two_DoWn wrote: Except for the fact that nothing in that post has anything to do with the rate at which someone gets trolls and leavers. Its just about someone being good at one champion and sucky with the rest. In the context of this argument he is a 2k player when he gets the handicap of playing an overbuffed champion, same as how I would be 2.4k if Shyvana never died. Well it kinda does insofar as playing FotM champs and stuff that people think is OP gets you less trolls than if you play stuff that people think is weak. It's not inconceivable that people who play Twitch every game get trolls more often than you or I do, because lots of people just go into troll-mode when someone on their team picks Twitch, even if that player might actually be good at Twitch. That said, I think this angle of looking at it makes the discussion even more pointless. | ||
| ||
|
|
BSL 21
Replay Cast
BASILISK vs TBD
Team Liquid vs Team Falcon
OSC
CrankTV Team League
Replay Cast
The PondCast
CrankTV Team League
Replay Cast
WardiTV Invitational
MaNa vs Gerald
Rogue vs GuMiho
ByuN vs Spirit
herO vs Solar
CrankTV Team League
[ Show More ] |
|
|
EPT![[image loading]](http://img.xcitefun.net/users/2010/02/150582,xcitefun-rainbow-roses-1.jpg)
