League of Legends General Discussion Thread

[Cull]

If you throw your placement matches(which by the way I have very little patience for the people who do this, including myself), and have like a 60% W/L ratio, it may feel like ELO hell the first 50 or so matches. Your ELO really won't move all that much. But if you have 200/300+ games, and complain about ELO hell, then maybe its time to reevaluate your game? One easy way to do this is comparing CS if you are in a solo lane. If you don't have 100 CS at 15 min, practice that hard. Just by going 50 CS higher than the enemy you basically have gotten a kill+ advantage on the other carry.

The system works when you do as you should do in your placement matches. If you do break-even, then its a long grind, but thats the way the ELO system works I think.

[Deceptus]

"All Europeans are gay" - Speaker



"f**king gay backdoor" - enemy Lee Sin



"LOLOLOLOLOLOLOLOLOL" - Jaje & Seyruun, proud Europeans

(I *really* hate Rammus jungling now...)

[v8mark]

More Elo hell discussion: of course Elo hell is nonsense. As Sir Bruce said a few pages ago, the law of large numbers dictates that you will gravitate towards your 'true' skill level. The high variance experienced in LoL is because in any given game, you only constitute 20% of your team's output. If you've only played 20 ranked games, there's a considerable chance you'll be outside your Elo bracket because of this variance. If you've played 100 games... well, as Cull says, you really need to look at your game. You're clearly doing something wrong, and you have to find it: it could be poor farming, poor fighting, lack of map awareness, lack of teamplay, adversely affecting morale by raging, not adapting to the meta of the Elo, playing too many champs so you're not practised enough with individual champs -- whatever. But the argument that you get all the feeders which makes it impossible to rise is absolutely ridiculous, because sometimes the enemy team gets them too. Often you won't even notice; such is human nature. Unless you think the Riot matchmaking system is specifically targetting you, and giving you all the baddies, in which case I strongly suggest you look up your local psychologist, and book an appointment.

Sulla and Varis are an extreme example of a legitimate rise in Elo -- obviously both are far above 1200 Elo in skill, and they played the system by duo queueing. That reduces the variance considerably. Clearly they weren't carried with the wins they have -- if you watch a couple of those games, you'll see that it's just not the case. On a few occasions, including one that Sulla posted, they in fact carried a game that 'average' players at that Elo would clearly have lost.

Now, clearly, if you're duo queueing, the lower variance means you have to play fewer games to reach your 'true' Elo level (assuming you're both of a similar skill level). But it's simple statistics that, if you play a representative/statistically significant number of solo queue games (probably 50-75 or so) you won't be too far from your Elo. If that's a lower number than you expected, be prepared to re-evaluate your game.

Personally, I think I'm about right at the moment -- my Elo was a bit inflated at 1500 due to good luck in my initial (high-impact) games, and is now a much more representative 1400 or so. There's genuinely a huge difference between 1200-1400 and 1400-1500, and I'm probably on the cusp of those two zones right now in terms of skill.

TL;DR: You have to play a lot of solo queue games to be sure you've reached your Elo, but once you've played 50-75, you need to stop complaining and l2p.

TL;DR^2: Elo hell doesn't exist.

[VoiceOfUnreason]

I'm unconvinced. Granted, part of the problem may be that I'm working with a different definition of ELO hell than others are.

If you've played 100 games... well, as Cull says, you really need to look at your game. You're clearly doing something wrong, and you have to find it

Agreed -- now, here's the big question from my point of view: if you find the problem(s), and fix them, can you get out?

Imagine, if you will, a player who establishes as low ELO over 100 games or so, and then goes off for a couple months to practice, get coaching, whatever. Let's imagine that he actually does manage to improve his true skill level by 200 points of ELO (as in, he creates a smurf, and plays 100 ranked games there, resulting in a higher rating).

Now he takes that extra skill back to his old toon, and.... Can he rise to his new ELO? How long does it take? Are the skills that were keeping him at his higher rating enough, or does he need to learn additional skills because the "meta" is different in the lower division?

Are the answers to these questions just as valid for support/tank roles as they are for carries?




One additional thought: Varis and Sulla are 9/2 1534 ELO rating. Speaker is 336/324 1522 ELO. I'm thinking that maybe we shouldn't be over counting the evidence of the first sample.


ps - well done guys; it's an experiment I would have liked to have done if I thought I was on the right side of 1200.

pps - I'm pretty sure that Riot is getting the response they wanted from these conditions of contest, but I'm sort of puzzled at what the choice. What is the upside of playing ranked all season, beside bragging rights?

[Cull]

Ranked solo q, no matter how much higher Elos trolololol with random stuff(I DUNNO THIS, BUT JUST GUESSING FROM STUFF) is a ton more competitive/tense. Its still gaming, but its a lot more "yaH, I WIN MORE RATING", or "i lose but I LEARNED STUFF/WILL GET BETTER FOR MOAR RATING!" than "k I just played a game" that I feel like I get from normal sometimes.

The same argument can be used in reserve. Why NOT play ranked>normal? The atmosphere is a lot more tense, fast-paced for me at least. Iunno. Normal's people try out a ton of new stuff, and everybody generally is chill. Ranked is a lot more "try-hard" to use a silly term.

Thats my argument for why to play ranked. If I'm just solo playing, I prefer ranked>normal.

[v8mark]

Agreed -- now, here's the big question from my point of view: if you find the problem(s), and fix them, can you get out?

Of course. It'll take a while, but you can and you will. Just plucking some numbers out of the air here, but I think if you find yourself at, say, 800 Elo, and suddenly improve to 1500 overnight, you'll probably win 70% of your solo queue games. Now, it'll take a while to get out of 800 wth that win rate, as you only get 15-20 Elo per win or so. So to rise 700 points you need to get approx 40 more wins than losses -- which even at a 70% win rate means you need to play, what, almost 120 games? And of course, the win percentage goes down as you get higher in Elo, so maybe you'll have a 60% win percentage when you hit 1200. So that's even more games you have to play. But statistically, if you're good enough, you'll get out.

Of course, the above is only a hypothetical, and players don't improve from 800 to 1500 overnight.

It comes back to the definition I guess. There's no question you have to play a lot of games to 'escape' an Elo. The good news is that it's fun, because you'll mostly have good numbers and be stomping a lot.

Good question about the roles. But part of being a good player is picking what works in the team comp, so if you go in with the attitude of 'I'm going to carry whatever', it'll be suboptimal. Now, you might be a good enough player that you'll overcome that and win anyway much of the time. But it'll be harder.

A good support can win games as well, although I think they have less of an impact than a mechanically and strategically good carry. That's just my feeling, and is based on no evidence whatsoever. But generally the fifth player in a team (and hence, the lowest ELO) will often play support. I feel tanking is very important though -- a decisively won top lane is a big advantage in a game.

Sulla and Varis's small sample size is effectively bigger, of course, because they duo queued. But I agree with the general sentiment -- there's still a lot of variance with 11 duo queue games. However, my guess is (especially having seen some of the games) that their run of wins is probably broadly representative of their ability.

And the upside of playing ranked is that it's fun. Re: Cull's post, I've noticed absolutely no difference in 'atmosphere'. People rage a lot in normals too.

[VarisNox]

I'm unconvinced. Granted, part of the problem may be that I'm working with a different definition of ELO hell than others are.

Agreed -- now, here's the big question from my point of view: if you find the problem(s), and fix them, can you get out?

Imagine, if you will, a player who establishes as low ELO over 100 games or so, and then goes off for a couple months to practice, get coaching, whatever. Let's imagine that he actually does manage to improve his true skill level by 200 points of ELO (as in, he creates a smurf, and plays 100 ranked games there, resulting in a higher rating).

Now he takes that extra skill back to his old toon, and.... Can he rise to his new ELO? How long does it take? Are the skills that were keeping him at his higher rating enough, or does he need to learn additional skills because the "meta" is different in the lower division?

Are the answers to these questions just as valid for support/tank roles as they are for carries?

One additional thought: Varis and Sulla are 9/2 1534 ELO rating. Speaker is 336/324 1522 ELO. I'm thinking that maybe we shouldn't be over counting the evidence of the first sample.

ps - well done guys; it's an experiment I would have liked to have done if I thought I was on the right side of 1200.

pps - I'm pretty sure that Riot is getting the response they wanted from these conditions of contest, but I'm sort of puzzled at what the choice. What is the upside of playing ranked all season, beside bragging rights?

Actually, for some reason, I'm 29 points higher than Sullla at 1563.

To your point though, yes, if you find the problems affecting your play, you should be able to get out. I had an extraordinarily easy time in these games. There were few points where I felt like the game was totally lost; I felt in most every game that I could carry the team to victory. As a point, the other enemy team in our very last game were (at the time of our game) 1601, 1648, 1800, 1551, and 1450. Comparatively, our Elo ratings were 1304, 1256, 1332, 1489 (Sullla), and 1518 (Myself). Obviously, we had two pairs of Duo queue and a solo queue player for these high discrepancies; this fact was made even more obvious when the Nocturne and Sion players ganged up to blame our Irellia player. Even with these higher ratings, I felt like the enemy team played terribly. If the game continued, I would've destroyed the entire enemy team 1v5.

Now, as I reach my true Elo, this feeling of myself needing to carry should diminish to where I feel that both teams have an equal chance of winning. The thing about lower Elo levels is that they are so much easier to affect. For example, in my very first game when I played Nocturne, I did not feel endangered what so ever, and that was around the 1200 Elo level. I've seen SaintVicious, Chauster, CleverAdvisor, and other high level players basically screw around in sub 1k Elo and still win easily. I feel like I could easily do the same at that level of play - winning about 90% of my games without issue while playing seriously. Depending on how many games I've played, it may take more games to rise up due to lesser Elo gains, but I'm confident I could still do it.

Of course the ability to carry is not as plausible for tanks/supports. If you want to carry, you're the lowest Elo, and no one else wants to, offer to jungle. It's a great way to influence the game. However, your higher Elo players should be allowed to play the carries. Theoretically, they would be most suited and able to carry your team. As I stated in my previous post, if the enemies high Elo player is a carry while yours is a support, expect to have a more difficult game.

However, you can still have a great impact on the game as a support. The advantage of having a warded map should not be discounted, and many times the support is the reason that your team wins a team fight. For example, Sullla kept so many people alive in our games through the smart use of skills and warding. He also was the primary reason I was able to do well as Kogmaw in the majority of our games, since he allowed me to farm in relative safety in the first 10 minutes or so of the game.

After my games, I'm certain that the biggest factor to low Elo success is the jungler. So, if your team wants to all be carries, pick a strong jungler. Nocturne, Lee Sin, and Fiddlesticks are all strong junglers that can majorly impact the outcome of the game. In particular, these junglers have easy buff control, great ganking, and excellent team fighting ability. Nocturne in particular is a wonderful choice in solo queue, since his ult allows you to punish any overextends made by the enemy team. While this may result in a 2v1 top lane, it will allow for greater map dominance. Even if top loses, if you can completely disrupt bottom and mid, while helping top when possible, it will be a huge gain.

Lower Elo metagames are also much weaker than higher level ones. These players don't seem to understand how to stop a super carry, constant pushing to tower, or a back door, and I rarely see competent junglers. There's a reason that the best way to play at low Elo is tanky, since few players at that level seem to be able to cope with it or understand how to counter it. It's the reason I picked Kogmaw so many times - it was really easy to counter an entirely tanky team that seems to be formed in low Elo solo queue.

I guess the biggest point I'm trying to make is that the farther away you are from your actual Elo, the more of an impact you will have on each game, both good and bad. If you have a much lower actual Elo than your current rating, you will more than likely be constantly outplayed at all points in the game to the point where you will lose more often than win. However, if your Elo is supposed to be much higher, your actions and decisions will allow your team to be led to victory more often. As you get closer to your true Elo, your influence will become more negligible. The issue of difficulty occurs when you only belong a couple rungs higher. If you only belong, say, 100-200 points higher instead of 800 points, climbing that next step will be much harder since you will have less influence on the game. In comparison, that 800 point higher player will have a much easier time influencing the game toward victory until he gets within his 100-200 point difference. For example, bigfatjiji has an easy time sub 2k Elo carrying his team on his smurfs. However, once he breaches that 2k, or so, range, his games become a lot harder to win, since the players are beginning to make less mistakes he can take advantage of.

So, to your other point, I hope that the player who trains and gets coaching improves more than a mere 1-200 points. Otherwise, no matter what he plays, while he will eventually reach his new level, he will probably still have difficulty climbing up to it.

[spellman]

Hm, I should probably buy more junglers and practice them. Right now I'm only competent w/ Nunu.

Then again, I really need to get real Teir 3 Rune pages too. And I've been told some of the other junglers just don't work (or their path is super limited) without a solid set of runes.

Then again, good gods I don't have time these days for games.

[sunrise089]

As we've discussed my ELO is ~800 (ok, I did raise it from 699 or whatever the lowest was...yay...). I've reached the point where my average KDA playing a mage mid is like 10. I welcome advice in how to translate this to wins (the coaching mentioned above if you will).

I know I've vastly improved as a player since starting ranked play. I take Sullla's criticism that no one forced me to play ranked the day I hit 30 (on the flipside Riot didn't help by refusing to implement unranked draft mode), but hypothetically if I'm capable of rising to 1200elo via grinding up that might take what using the above examples? 100 games? Duo queue would be ideal of course, but no one has an elo remotely as bad as me to make a good duo combo.

[Dantski]

I'd happily duo queue with you sunrise, my lofty 1100 rating was hurt a bit by recent DC's. New internet being setup tomorrow should hopefully resolve that.

@ Sullla - Two criticisms I have from your Janna games is that sometimes you stand too far back in lane and even miss XP when not being harassed, the other point is that you used Howling Gale on dragon for minimal damage while saving it for use if the enemy team contests it would be optimal. Definitely want to save those skills if you don't need to use them.