Welcome to the Forum Archive!

Years of conversation fill a ton of digital pages, and we've kept all of it accessible to browse or copy over. Whether you're looking for reveal articles for older champions, or the first time that Rammus rolled into an "OK" thread, or anything in between, you can find it here. When you're finished, check out the boards to join in the latest League of Legends discussions.

GO TO BOARDS


So is the elo system a good indicator of skill level?

Yes 20 64.52%
No 11 35.48%
Voters 31 .

Official Statistical Analysis of Elo system

12
Comment below rating threshold, click here to show it.

TiDjOu

Member

01-08-2013

Quote:
Ms Secret:
If you place a high Elo player in low Elo enviroment, they will win more then they lose.

What you did, was take high Elo player who were in a high Elo enviroment and predicted that they win as much as they lose.

Elo working as intended.


this


Comment below rating threshold, click here to show it.

KirbyCake

Senior Member

01-08-2013

his conclusion fails
he should state that good players are no better than OTHER good players GG


Comment below rating threshold, click here to show it.

SmokingPuffin

Senior Member

01-08-2013

There is really so much wrong with the analysis presented in the OP that I don't know where to start.


Comment below rating threshold, click here to show it.

Snowywarriors

Junior Member

01-08-2013

I agree here, everything is right and since the best players in the world were used, they should be beating other high players. Wouldn't you expect the best player to beat another high skilled player?


Comment below rating threshold, click here to show it.

efdf

Senior Member

01-08-2013

u suck at statistics


Comment below rating threshold, click here to show it.

efdf

Senior Member

01-08-2013

and critical thinking


Comment below rating threshold, click here to show it.

bliu23

Senior Member

01-08-2013

Highly rated players win more until they get to their Elo.

Put a highly rated playing in an Elo way below them, and they win a greater % of games.

Edit: Er, I didn't read Ms Secret's post, ignore this.


Comment below rating threshold, click here to show it.

Joseph Dom

Senior Member

01-08-2013

The way to see if it works is to take players that have a high number of games played and track win rate and ELO change. If the ELO is stable and the player has about a .5 win rate it is a good indicator. If a player has about a .5 win rate and ELO has a large range then it is not a good indicator.


Comment below rating threshold, click here to show it.

j AiDeHua

Junior Member

01-08-2013

League of Legends is a team game. Remember the saying there is no 'I' in team? Yeah, it applies here too. When you initially hit level 30 and can start playing ranked games you are placed around 1250 ELO.

As you most likely know when you first start playing ranked you have a 'bonus pool' of ELO that is pulled from whenever you win or lose a game. The amount in the pool gradually goes down over time. However usually the first game you play you win or lose 50-60 ELO after the game. Now there are approximately 10 placements matches and since this is a team based game a really bad person can be paired with four other really good players. This can happen an indefinite consecutive number of times and that player can get carried through all their placement matches and be placed at a relatively high ELO 1500-1800. In opposition really good players can also be paired with really bad players and lose an indefinite consecutive number of times.Take into account that the League of Legends community is really large and you have an enormous amount of lurking variables, meaning that you can have either good or bad players on your team (as the lurking variables).

So, really the ELO system isn't a very good representation of skill in Solo Queue or Duo Queue. Solo Queue and Duo Queue are just matches where you play with a 'carry more' kind of attitude because you do not really know who you will be paired with or against. The only real definitive representation of skill in ELO is through ranked team games such as 5v5s or 3v3s where there are almost no lurking variables to be found.

ELO is a really difficult system to understand but it is very much like the Starcraft ladder system. The only difference is that SC is a 1v1 oriented game for the most part, most of the glory is in 1v1s.

What I'm trying to say is ELO is not totally a definitive measure of skill, because in a team game you can always get carried to victory.


12