Elo hell? Elo is a horrible rating system? Someone help.

Comment below rating threshold, click here to show it.

Sweetea17

Junior Member

09-18-2012

Quote:
Originally Posted by PogoPogoPogoPogo View Post
There are still other explanations.


And as I said, there's no ranked matches in his 10-match history. In fact, he hasn't played a ranked match in the last 7 days even. Moreover, his Elo has remained a constant 1493 since 09/05/12, so unless he's gained and earned the exact same amount of Elo in the same day, he probably hasn't even played ranked in the last 13 days even.

And while this still isn't definitive or conclusive, it suggests that he definitely hasn't been playing much recently. And that makes it all the more likely that he was able to inflate his win percentage by queue dodging back when that provided an Elo penalty. Again, there's no direct evidence, but to me, that's what it points too. Get him to go play a ranked match today. If he wins, watch him get 10-14 Elo.
Yea, I'll be watching it for a month or so and I'll update the first post with each players elo then average it out and see whats up, its just insane, those numbers man.


Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

09-18-2012

Well, for one, see if he even plays ranked matches in the next month...

And for two... quit using the chess argument in this case.


Comment below rating threshold, click here to show it.

Sweetea17

Junior Member

09-18-2012

Quote:
Originally Posted by PogoPogoPogoPogo View Post
Well, for one, see if he even plays ranked matches in the next month...

And for two... quit using the chess argument in this case.
Its not the chess argument, Its only what I used to understand when it comes to elo, Its like taking propellers out of the water because they worked so well in water doesn't mean it works the same on land.
I'm talking about the trains with propellers, not air crafts.

I already plainly told you, not arguing, trying to understand.

elo is this game, isn't the same meaning in the context that I'm used to understanding fully. Thought I already stated that.


Comment below rating threshold, click here to show it.

FINGER IN BUTTHO

Member

09-18-2012

So the elo system as designed for 1v1 games and is being implemented in a 5v5 game?

This means in each game you are contributing 20% to your team.

This means, for your elo to be accurate, you ought to be playing 5x as many games


Does this make sense? It's not about each game, you should really be thinking in sets of 5


Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

09-18-2012

Quote:
Originally Posted by Sweetea17 View Post
If you look at The Elo system for what it was made for, It makes complete sense and all of it falls into a nice wall of logic that explains it.
But, this isn't 1v1 chess.
You're pointing out that Elo is designed for chess, a 1v1 game, and not for LoL, a 5v5 game, and therefore, Elo will not work for LoL. People have made that argument (and I've got my counterpoints to that argument).

But that argument doesn't fit in this thread.

The reason people point out chess when talking about LoL's Elo is because an individual player could get stuck with 4 bad teammates, and despite the individual player playing exceptionally well, he'd get stuck with a loss and lose Elo when he played well enough individually to win (and earn Elo).

The reason the chess example simply does not work here at all is because there is something else going on. The player your citing is playing well enough to deserve wins AND he's is GETTING those wins. Whatever is causing him to be at this low Elo (relative to his winrate) would cause him to be equally low if he were playing chess.

Suppose we're talking about an online chess server. Your Elo gains and losses for matches played works exactly the same as it normally would. You gain more Elo for beating higher ranked opponents, gain less for beating lower ranked opponents, lose more for losing to lower ranked opponents, lose less for losing to higher ranked opponents. Okay, that's fair, right?

Additionally, there's Elo decay on this chess server. If you don't play ranked matches often enough, your Elo will fall over time.

And finally, there's something equivalent to an Elo dodge penalty if you queue into a match with an opponent, but you quit the match before the first move is played. It doesn't count as a win or a loss, but you do lose Elo for it.

Now, if we take our chess player (we're gonna call him Timmy), and he's only playing ranked chess matches against extraordinarily low ranked chess players, then he's going to only be gaining a point or two of Elo per win, while he'll be losing a LOT the few times he does lose. This could explain an extraordinarily high win rate despite a relatively low Elo, even in chess.

Perhaps that's not the case with Timmy. Suppose instead Timmy legitimately got that Elo playing people of similar Elo ratings, but Timmy hasn't played a ranked match in 6 months. He'd Elo decay down, eventually to this medium Elo level, despite his extraordinarily high win-rate. This is a way for Timmy to have this same Elo in chess.

And then there's the final option. Let's say Timmy just likes to queue up for matches then dodge them before the first move is played. He does this for whatever reason, probably because he wants to stay in a certain Elo bracket. Now, he crushes that Elo bracket, so when he does play, he dominates, and he gets a legitimate amount of Elo, but then he just dodges a few times to knock his Elo back down. Once again, we've got a situation in chess where a player could manage this win-rate but remain at a relatively low Elo.


Now, the first example can happen in ANY real chess rating system. It'd be the most common explanation in chess for this scenario. But it's actually probably the hardest to obtain in LoL...

The second and third examples assume some online chess match-making websites with rules regarding Elo outside the actual winning and losing of games. LoL has these rules (though they've somewhat recently disbanded the dodging penalty... it remains as possible explanation for the player in question in this thread however). Now, you may not like these rules that impact Elo outside of actually winning and losing, and maybe you think they should be removed, but it does exist as an explanation for a low Elo despite an very high win-rate. In fact, if one of these rules are in play here, if the player in question has tanked his Elo (purposely or not) in either manner, it makes his win-rate essentially completely irrelevant.


Comment below rating threshold, click here to show it.

Sweetea17

Junior Member

09-18-2012

Quote:
Originally Posted by PogoPogoPogoPogo View Post
You're pointing out that Elo is designed for chess, a 1v1 game, and not for LoL, a 5v5 game, and therefore, Elo will not work for LoL. People have made that argument (and I've got my counterpoints to that argument).

But that argument doesn't fit in this thread.

The reason people point out chess when talking about LoL's Elo is because an individual player could get stuck with 4 bad teammates, and despite the individual player playing exceptionally well, he'd get stuck with a loss and lose Elo when he played well enough individually to win (and earn Elo).

The reason the chess example simply does not work here at all is because there is something else going on. The player your citing is playing well enough to deserve wins AND he's is GETTING those wins. Whatever is causing him to be at this low Elo (relative to his winrate) would cause him to be equally low if he were playing chess.

Suppose we're talking about an online chess server. Your Elo gains and losses for matches played works exactly the same as it normally would. You gain more Elo for beating higher ranked opponents, gain less for beating lower ranked opponents, lose more for losing to lower ranked opponents, lose less for losing to higher ranked opponents. Okay, that's fair, right?

Additionally, there's Elo decay on this chess server. If you don't play ranked matches often enough, your Elo will fall over time.

And finally, there's something equivalent to an Elo dodge penalty if you queue into a match with an opponent, but you quit the match before the first move is played. It doesn't count as a win or a loss, but you do lose Elo for it.

Now, if we take our chess player (we're gonna call him Timmy), and he's only playing ranked chess matches against extraordinarily low ranked chess players, then he's going to only be gaining a point or two of Elo per win, while he'll be losing a LOT the few times he does lose. This could explain an extraordinarily high win rate despite a relatively low Elo, even in chess.

Perhaps that's not the case with Timmy. Suppose instead Timmy legitimately got that Elo playing people of similar Elo ratings, but Timmy hasn't played a ranked match in 6 months. He'd Elo decay down, eventually to this medium Elo level, despite his extraordinarily high win-rate. This is a way for Timmy to have this same Elo in chess.

And then there's the final option. Let's say Timmy just likes to queue up for matches then dodge them before the first move is played. He does this for whatever reason, probably because he wants to stay in a certain Elo bracket. Now, he crushes that Elo bracket, so when he does play, he dominates, and he gets a legitimate amount of Elo, but then he just dodges a few times to knock his Elo back down. Once again, we've got a situation in chess where a player could manage this win-rate but remain at a relatively low Elo.


Now, the first example can happen in ANY real chess rating system. It'd be the most common explanation in chess for this scenario. But it's actually probably the hardest to obtain in LoL...

The second and third examples assume some online chess match-making websites with rules regarding Elo outside the actual winning and losing of games. LoL has these rules (though they've somewhat recently disbanded the dodging penalty... it remains as possible explanation for the player in question in this thread however). Now, you may not like these rules that impact Elo outside of actually winning and losing, and maybe you think they should be removed, but it does exist as an explanation for a low Elo despite an very high win-rate. In fact, if one of these rules are in play here, if the player in question has tanked his Elo (purposely or not) in either manner, it makes his win-rate essentially completely irrelevant.
his spread is ridiculousness, All I'm saying is this
this "elo for 5v5 not teamed solo play" is the best it has, Its lets sooooo many variables, So many More Variables then turn based chess, The math behind chess can be broken down into so many ways and not just the elo but actual planning,
but chess is a very very old game and it makes sense that all the math have already been done on this, Now league is still in its toddler stages, and riot is always improving, but His TOP rating is his rating right now. If it were Decay his Top elo should be higher, no?

I'm sure in 10 years or so, there will be some other systems in place working with the elo that we know now.

Also a question mark is on my topic right? Its means to inquiry, not to argue.
Take what your reading and read it, don't read in between the lines, 92 percent of communication is nonverbal.
This is only 8 percent, Stop interjecting arguments, It doesn't belong here, its not what it is,
If you think that is what is it, You are mistaken.

If its how you feel about it, so be it, But I myself is not stating its not an argument.
I want Facts.


Comment below rating threshold, click here to show it.

Zaelious

Member

09-18-2012

Perhaps they should have elo gain/loss scale with k/d/a. That could fix a lot of issues if you're stuck carrying noobs constantly. It would help with all placements, your team won but you got a noob who went 0-1-0..? He gets 20% of the elo he would have gained, etc.


Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

09-18-2012

Quote:
Originally Posted by Zaelious View Post
Perhaps they should have elo gain/loss scale with k/d/a. That could fix a lot of issues if you're stuck carrying noobs constantly. It would help with all placements, your team won but you got a noob who went 0-1-0..? He gets 20% of the elo he would have gained, etc.
No. Stop. Go find one of the other idiotic threads that thinks this is a good idea.

Quote:
Originally Posted by Sweetea17 View Post
his spread is ridiculousness, All I'm saying is this
this "elo for 5v5 not teamed solo play" is the best it has, Its lets sooooo many variables, So many More Variables then turn based chess, The math behind chess can be broken down into so many ways and not just the elo but actual planning,
but chess is a very very old game and it makes sense that all the math have already been done on this, Now league is still in its toddler stages, and riot is always improving, but His TOP rating is his rating right now. If it were Decay his Top elo should be higher, no?

I'm sure in 10 years or so, there will be some other systems in place working with the elo that we know now.

Also a question mark is on my topic right? Its means to inquiry, not to argue.
Take what your reading and read it, don't read in between the lines, 92 percent of communication is nonverbal.
This is only 8 percent, Stop interjecting arguments, It doesn't belong here, its not what it is,
If you think that is what is it, You are mistaken.

If its how you feel about it, so be it, But I myself is not stating its not an argument.
I want Facts.
Sweetea, I don't think English is your first language, is it?

You're making multiple different points here that are polar opposites. On the one hand, you're talking about not wanting interjecting arguments, and you want to talk about facts. On the other hand, you're replying to my points about why, in this case, we can't talk about the fact that Elo is designed for chess.

I mean, you flat out don't understand my point about chess at all, try to refute my point, then try saying that chess has nothing to do with it.


First off, you have to allow for hypotheticals in this argument. Let me simplify your argument.

Your argument is Player X has Y Elo despite Z win-rate. Therefore, the Elo system in LoL is broken. Whether you put a question mark behind it is irrelevant. Whether you state it as a fact or as a question, I'm giving you an explanation why Player X can have Y Elo despite Z win-rate and the Elo system is still working perfectly fine as intended. As I said, queue dodging penalties, Elo decay, etc., etc., etc. There are perfectly good reasons for this to exist.

Now, as for his "top" Elo, top Elo has always been pretty buggy and I've seen it often display as people's current Elo (especially on 3rd party sites). But if you don't believe me that it's been buggy, that doesn't even matter because super recently, Riot said "Hey, we're resetting your TOP Elo to match your CURRENT Elo, whatever it is today (whatever day they made annoucnement) and the end of season reward you get will be based on what your TOP Elo is from today forward." Therefore, Elo decay is another superbly excellent explanation for this summoner's Elo/win-rate.


As for the chess argument, I don't know how to explain this at all any clearer whatsoever. If the player in question had the KDA he had, but his W/L ratio were flipped, then the chess argument might be more relevant in this case. The chess argument regarding LoL's use of Elo revolves around the fact that 4 other players supposedly have too big of an impact on your Elo because your Elo is based on your W/L, not your actual performance. Here, however, we have an example where the individual player's performance is crushing, and his W/L ratio is ALSO crushing. The fact that his W/L accurately reflects his individual performance means that the chess argument is 100% completely null, void, and absolutely retarded in this instance. You simply can not make the chess argument in this case. I'm not saying there's not room for the chess argument to be made regarding LoL's use of Elo. What I'm saying that with the example of this player here, the fact that Elo was designed for chess, a 1v1 game, is 100% irrelevant.

100% irrelevant.

Stop trying to bring the chess argument into this thread.

It's the wrong argument to make here.

Whatever is causing this guy's Elo to be so low would happen EXACTLY if he were playing chess, had the same W/L record, and was subject to queue-dodging Elo penalties and Elo decay. What is causing this player's Elo to be so low is queue-dodging or Elo decay. Period. End of story.


Comment below rating threshold, click here to show it.

OriginalJack

Senior Member

09-19-2012

Quote:
Originally Posted by Sweetea17 View Post
Now league is still in its toddler stages, and riot is always improving, but His TOP rating is his rating right now. If it were Decay his Top elo should be higher, no?
Just FYI the top rating you see on lolking is the top rating since they announced the new badges a week or two ago. It's not the top rating for all of season 2, the Elo decay theory is still a possibility.

Also, I think it's pretty clear something strange is going on here and we aren't really going to be able to get the full story with the limited info lolking has. There's no way of knowing for sure what happened without seeing his full ranked matched history for the entire season.


Comment below rating threshold, click here to show it.

iZaKLY

Junior Member

09-19-2012

whatever. mad respect for the dude. Whatever the reason, he deserves a higher elo lol. like seriouslt 148 wins 4 losses jarvan. Wtf..