Elo hell

Comment below rating threshold, click here to show it.

mwarme

Senior Member

01-18-2013

Quote:
Originally Posted by Boagster View Post
Now do yourself a favor. Stop ignoring the matches you win. Get a pen and paper or a word processor document. Every time you play a game, write down the result as well as the scores of every player and who disconnected. After one hundred games, compare the wins and losses and tell me if the other team didn't suffer from the same problems just as often.
After i went 1-9 in placement matches this season with a total of 14 dc's/afks on my team, I actually went and DID THIS. For the next 100 matches, i had a 63% (46 out of 73) win rate of games in which NOBODY on either team dc'd on either team. in two of those games with dc's, they reconnected within 5 mins and we won those. so. 48/75 in games with no *effective* dc/afk. those other 25 games, 5 were dc's on the other team and we won 5v4. (right here i rage because 80% of games with a dc on either side for me happened to my disadvantage. them's numbers). Of those 20 games where a teammate dc'd, we won three. my end result? 56/100. not a bad win rate for 100 games. The problem? If ALL of the games with a dc (all 25) were at my average win/loss rate for unaffected games, 64%, I would have had a 64% win rate overall. So, that 8 % difference in win rate is 8 games worth of elo difference, which is about 100 elo give or take my average 11-13 gain/loss. That's 200 elo if you consider I won those and gained elo instead of losing them and losing elo. As it is i gained about 100 elo over those matches. If i say I only were to win 56% of games as my elo rose, i'm still about 100 elo under where i would be if not for the afk's/dc's.


Comment below rating threshold, click here to show it.

Boagster

Senior Member

01-18-2013

I wasn't just talking about DCs, though, mwarme. That is a huge factor, but WaltLongmire also said:

Quote:
Originally Posted by WaltLongmire View Post
I have lost too many games where multiple teammates have had 15 or more deaths with less than 5 kills.
I wanted him to also see how many games he wins where the enemy has people with similar numbers. The losing team will often have one to three players who perform very bad. That's the nature of the game. In order for one person to do really well, at least on person on the other team has to be doing bad. Until you are a really high end player playing against other really high end players, you should not expect to see many close matches, as people at the low end tend to be really inconsistent in their performance (this includes most, if not all, of us here).

I am glad that you have done this, though. Citing the aforementioned Law of Big Numbers, you'd hopefully would have seen more DCs on the other team given a a larger sample. The only other explanation would be that you are just an unpleasant person and make people want to quit games on you, but I don't think that to be the actual case.


Comment below rating threshold, click here to show it.

PogoPogoPogoPogo

Senior Member

01-18-2013

There's some logic failure here mwarme.

Unless you are exceedingly far from the Elo rate you belong, you're win rate in any given bracket can not be extrapolated to other brackets.

It's more difficult to win at 1200 Elo than it is at 1100 Elo. It's more difficult to win at 1300 Elo than it is to win at 1200 Elo. It's more difficult to win at 1400 Elo than it is at 1300 Elo.

It doesn't matter what Elo you are, the higher the Elo bracket you're playing in, the harder it is to win.

Now, that doesn't mean that it's harder for a 1400 player to win in 1400 Elo than it is a 1200 player to win in 1200 Elo. Nope. It's of equal difficulty for these two players.


So, when you lose a game because of an AFK on your team (by the way, I've both won with an AFK on my team and lost with an AFK on the enemy team), you're ending up at a lower Elo than you were before you started that match. That means the game will be easier to win.

You can't manage a 63% win rate in 75 matches and say "because I lost 100 Elo to AFKers, I should be at 1400 instead of 1200." Why? Because let's take those 25 AFKer matches out for a second. A larger portion of those matches would have been played at whatever your top Elo ended up being (after 100 matches). And then there'd be some matches (assuming you continued to win) that would be played at Elo brackets you've never played at before.


BUT... but.. here's the thing... despite those AFKs, you actually did still manage a positive win rate overall. Play another hundred matches. First off, unless YOU'RE the one causing people to AFK, then once you've played enough matches, the AFKs will balance out. Next... as long as you have a positive win rate, you're increasing your Elo... so you're not stuck anywhere.


Comment below rating threshold, click here to show it.

Wiz el Poderoso

Junior Member

01-19-2013

Quote:
Originally Posted by Busy Izzy View Post
i've felt the same way for quite a while now, but in S3 there seems to be more unskilled players than trolls in my solo queue's. i've also had really bad ragers who have pissed me off so much i felt like uninstalling, but that's just a part of the game...like any online game.

i found better to play with friends, as they don't rage at you much, they help you, and the chemistry is right...not everyone has 5 or more friends whom they can play with, i try to play with everyone...i am part of a 5 v 5 ranked team, although we don't really have the time to play altogether, some of us can play together. playing with friends have made me a better player, and a better person.

for anyone feeling out of place, hit me up! i love to make people feel better, teach people new tips/hints, give advice, and i enjoy playing with a broad range of people no matter they're playing style...so if you want, why not add me and i'll play some games with you...maybe if the chemistry is right, you could possibly join the team...we'll see

Yeah I usually go in duo with my buddies but they don't really care as much.. they just hop on CoD and vent off on there


Comment below rating threshold, click here to show it.

Wiz el Poderoso

Junior Member

01-19-2013

I hate the whole "Provisional" system at the start of your ranked career. Once you're out, there's nothing stopping you from pairing up with lower skilled people who are just starting/going through their provisional games.

There should be 2 pools of players to get Ranked started.
1 - Provisional Pool. These are the players who have yet to attain their ranking. Provisional Pool has to play at least 10 games to get their "Highest Rating". When you start the Provisional rounds, you will start at 1200 elo (base), and be paired up with fellow players of the same rating. Each round/game you progress further and verse players of gradually increasing difficulty, but not so high as such you get absolutely stomped. For example, you are in your 5th game/10 and you currently have 1250 current rating. You are then matched up with players +/-50 of your rating.

Pool 2: Ranked Pool. These are players who have played all 10 provisional rounds to attain their final Current Rating. 10 games seems fair as your rating isn't final until the 10th game has been played. some people will say 5 games are enough, doubling the amount of games shouldn't matter. But doubling the amount of games ON TOP of facing gradually higher ranked players per round should indicate what level you are capable of playing at. This is opposing the current system where you face higher and lower ranked teams willy nilly until you reach you final "Highest Rating".

Some people will say 10 games are much too many, but why play Solo Queue and only play a couple of games? If you really want to show your skill you should be willing to slog it out in the provisions and have your final fancy rating that's more accurate than the current system.

One more addition could be Picking Roles. Once you've selected you're champion you then have 10 seconds to Type in/Select the Role you wish to play your champion as. This should be taken into account when being given a rating, and will be more accurate as to not seem like a feeder or a waste of space.

For example, you can choose from:
Top Lane
Jungle
Mid Lane
Bot Carry/ADC
Support.

Champions are given various roles in the store which should also be applied to this system, meaning you cant choose Ez and select the Support role to boost your rating. Theres so much more to expand on this but if people can get the jist of what I'm saying that would be awesome. It does seem like a lot of effort and complication, but to have a game that boasts a one-of-a-kind "fair" system is truly unique, and I'm sure it will pay off if implemented. I'm not saying do EXACTLY what I said, but along the lines of.


Comment below rating threshold, click here to show it.

frostysam

Member

01-19-2013

pro tip stay silver till s3 patch never go into bronze noobs place aka elo hell


Comment below rating threshold, click here to show it.

WaltLongmire

Member

01-19-2013

To be honest, I don't need to compile any list of 100 games to see what the problem really is. The problem really is that players on a winning team are not rewarded equally, hence the reluctance to take on roles that are guaranteed to provide less elo boost after a win. Case in point: playing proper support as Taric. When playing Taric bot with the ADC, I think of it as my job to let the ADC get the creeps and enemy kills while ensuring that he (or she) stays alive at all costs. I buy the wards, and the aura items for the team. That will generate final numbers for me that seem to be 2/3/10 with an easy win in 20-25 minutes, with maybe 12/2/5 for my ADC. I get +15 ELO, and my ADC gets +20 or more.

Add that to the fact that if you play support or jungler, you are most likely to be berated constantly by teammates that got their wish in the choosing process. God forbid if you accidentally kill an enemy on a gank with your burst combo and ult after the laner hit them once.

To sum up, most people in the mid and lower levels seem not to try as hard if they don't get the alpha roles on a team. This might change if the rewards were adjusted to more accurately reflect the value of jungling and support to the team.


Comment below rating threshold, click here to show it.

adc

Member

01-19-2013

Quote:
Originally Posted by WaltLongmire View Post
To be honest, I don't need to compile any list of 100 games to see what the problem really is. The problem really is that players on a winning team are not rewarded equally, hence the reluctance to take on roles that are guaranteed to provide less elo boost after a win. Case in point: playing proper support as Taric. When playing Taric bot with the ADC, I think of it as my job to let the ADC get the creeps and enemy kills while ensuring that he (or she) stays alive at all costs. I buy the wards, and the aura items for the team. That will generate final numbers for me that seem to be 2/3/10 with an easy win in 20-25 minutes, with maybe 12/2/5 for my ADC. I get +15 ELO, and my ADC gets +20 or more.

Add that to the fact that if you play support or jungler, you are most likely to be berated constantly by teammates that got their wish in the choosing process. God forbid if you accidentally kill an enemy on a gank with your burst combo and ult after the laner hit them once.

To sum up, most people in the mid and lower levels seem not to try as hard if they don't get the alpha roles on a team. This might change if the rewards were adjusted to more accurately reflect the value of jungling and support to the team.
Umm, what? Nothing affects your Elo gain or loss other than your current Elo and how much you were expected to win the game...


Comment below rating threshold, click here to show it.

Rillain

Senior Member

01-19-2013

Quote:
Originally Posted by AD Bottom View Post
Umm, what? Nothing affects your Elo gain or loss other than your current Elo and how much you were expected to win the game...
Play 10 games of mid compared to 10 games of support. The elo change will be different guaranteed.


Comment below rating threshold, click here to show it.

Boagster

Senior Member

01-19-2013

Uh, no Rillain, it won't. Your Elo gain may be different from game-to-game, but your role is 100% irrelevant to your Elo. How can the system even figure out what role you played - especially considering Riot does not like embracing the meta?