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.


Is the "ELO" match making system enough?

The "ELO" system of basing matching off WINS and LOSSES is perfect. 0 0%
Taking Average stats into account for matching is the best idea. 10 83.33%
I am content with the fact i think if youre good you can win no matter what. 2 16.67%
Voters 12 .

Improvement to Ranked Match Making

Comment below rating threshold, click here to show it.

Rej Kay



Improvement to Ranked Match Making

Lets state the obvious problems

1. Every player starts out with 1200 ELO when playing a ranked match
2. Resulting from the stated above novices easily get match made with players who
are of "same skill level".
3. this "skill level" is the players current ELO rating.

Now obviously ranked matches are ruined by leavers, feeders, and just plain novices.
But the game itself records each players performance accordingly yet none of the stats
are ever reviewed and included into the match making procedure.

When match making players into teams the game needs to take in account ALL stats
the player has accumulated over the played ranked games when matching with random summoners, instead of just matching summoners with similar ELO numbers.
The Improvement

Base all ranked match making on the players ranked stats and not ELO alone.

The system should be based like this.

Step 1
the match making system determines if the player has ANY ranked game experience. If the player has NO ranked game experience he is then matched with other players who have NO experience and NO ONE ELSE. After that match then the player starts to accumulate Ranked stats. and then is pooled with other "experience" ranked players. The first match is crucial to where the player is to be placed in future matches.

This eliminates potential unskilled players from preventing players in "elo hell" from getting out of it plain and simple. Unskilled players come in at 1200 ELO and keep good players from ever breaking 1200 because they are stuck at 800-1100 elo.

Step 2
Once the system has determined if the player has any ranked game experience it reviews the players current ranked stats. This is to include how many wins, losses, and leaves the player has, his or her current K/D/A. Once all averaged stats are determined the player is then placed with players of similar statistics.

The system matches players that are essentially at there same level of game play ALL AROUND.

This will keep "unskilled" players together and the better players essentially in the same groups. If you deserve to be put with bad players you get stuck with other bad players until your current averages prove you are.

Step 3
After the system has matched you with players that have similar ranked stats as you, you are then placed in a pool of players that have similar ELO. And then the match begins

after the calculation of ranked stats the ELO will put you with players around your ranked game total experience. and essentially wins and losses actually matter.

The proposal

With all that said and done. That would bring me to conclude that the
match making system should be based off not just one Stat which is
currently "ELO" but the overall stats of the player and THEN "ELO".


Quite simply this is how it should be done instead of a plain WIN/LOSS system. Its flawed i get it, but you need to fix it. Competitive players lose interest everyday when they see an obviously rookie player make it to play ranked and costs them the match.
Start taking into acct a players average ranked stats and match them with other players like them .... not a cluster fk of players that have the same ELO beacuse ELO does not determine a players skill .... HIS AVERAGE STATS DO.

No matter what you cant Argue with average stats, and this best reflects
the players average game play level. not his Win/losses.

Comment below rating threshold, click here to show it.


Senior Member


I like the current one, but I'd also like to get "Taking Average stats into account for matching is the best idea." mixed in.

Since, if you go 0/10/2 5 matches in a row then I'd say you do not belong above 1200, but you could still possibly win all those 5 games and gain like 60 Elo.

Comment below rating threshold, click here to show it.

USC Zeratul



I think the people should be able to upvote or downvote you, so you will play with people that have also been upvoted/downvoted.
This way, if there is someone who is in the low ELO's and doesn't deserve it, he might get upvoted and play with people that are ~200 ELO higher, and if he does well, then his ELO gain should be higher from his wins.