How many games do you expect to play before you reach your true Elo?

< 10 Games 180 1.63%
11-40 Games 819 7.42%
41-80 Games 1,162 10.52%
81-120 Games 1,737 15.73%
121-160 Games 664 6.01%
161-200 Games 929 8.41%
200-399 Games 1,703 15.42%
400-599 Games 1,085 9.82%
600-799 Games 456 4.13%
800-999 Games 170 1.54%
1000+ Games 2,139 19.37%
Voters: 11044. You may not vote on this poll

Help Riot improve matchmaking! Looking for examples of bad matchmaking

First Riot Post
Comment below rating threshold, click here to show it.

Pentastik

This user has referred a friend to League of Legends, click for more information

Member

02-28-2012

this is my game result
solo que at ranked

https://docs.google.com/spreadsheet/...Gc&output=html

dead zone 1000~1300elo..

depend on which team have afk or bad duo or seeding player(in batch)

EDIT 1. updated to 44 results. today met afk 4times.. awesome


Comment below rating threshold, click here to show it.

Lyte

This user has referred a friend to League of Legends, click for more information

Lead Social Systems Designer

Follow RiotLyte on Twitter

02-28-2012
98 of 173 Riot Posts

Quote:
Originally Posted by II Peter Pan II View Post
this is my game result
solo que at ranked

https://docs.google.com/spreadsheet/...Gc&output=html

dead zone 1000~1300elo..

depend on which team have afk or bad duo or seeding player(in batch)
Wow, thank you for the detailed feedback here.


Comment below rating threshold, click here to show it.

GaiaEffingKing

This user has referred a friend to League of Legends, click for more information

Member

02-28-2012

All I can say is perhaps you could meet with Blizzard. The MMR system they use for Starcraft 2 (1v1 AND 3v3/4v4), though not perfect, is by far superior to the system in place in LoL.

Also, I have to agree with many people here that starting at 1200 elo was/is a horrible idea. I wish I saved screenshots but i've played several games where I have over 200 wins and all 4 of my (terrible) teammates have under 20 wins.

Matchmaking needs to factor in a LOT more than it does currently; win ratios, number of wins, average KDA in games, etc.



But backt o my original point: if you were to translate and implement Blizzard's MMR matchmaking system for SC2 into LoL, you would see a huge improvement.


Comment below rating threshold, click here to show it.

Leon

Senior Member

02-28-2012

Yeah, so i HATE playing normals (I'm 2k elo just wanted to mention that right away.) with anyone lower than me. I always get 1900+ on the other team which becomes a stomp fest for them because the people i'm playing with obviously aren't that great.

But, when I queue with other 2k elos I'm ALWAYS getting unranked on the other team and it's just a stomp fest for us.

I ESPECIALLY remember when i was normal queuing with a four people less than 1400 I had somebody from diginitas on the other team and everyone else was 1700+


Comment below rating threshold, click here to show it.

WWWWWWORP

Senior Member

02-28-2012

I'v heard people say that the type of elo system doesn't work for team games but I actually think it does it's just the games are so long is hard for the average person to get enough of them in for it to match the true elo. One thing you might try is hiding the elo unless a person is at a certain rank kinda like in a different game has it. It sucks for the people that want to see their elo but generally you can figure it out just by looking at your wins/loss and for the casual player it might help them feel less punished by losses idk?

I just would like to be able to get out of having to play with some teams in champ selection without it costing me a game loss in ranked. Making back elo from a dodge is very hard so I have not been doing it as I dropped to 700 elo in a about two weeks when I was just sick of trolls. I'v stopped dodging and it's brought my rating back up to 1100ish but I still get those moments in champ selection and thinking to myself this is going to be the least fun 30 mins of my life. I would also be fine with a dodge not costing as much elo as you can make back a few elo lost fairly fast but not a whole complete loss.

Like generally you'll get a team of 3-4 people who are on the same page and then you end up with one person who is just trolling so my suggestion would be to make some way a majority of a team could get the champ selection canceled by vote it would make trolling much less effective at least in champ selection.

Really the biggest issue I think in ranked is that as a team game with rating on the line it makes trolling more appealing to the types that would normally troll. Having a really bad game is one thing but having a really bad game with someone making it bad intentional is just not fun. So I would personally love to see a 1v1 type map with this type of a game.

TlR I want a 1v1 map for ranked


Comment below rating threshold, click here to show it.

5071

This user has referred a friend to League of Legends, click for more information

Senior Member

02-29-2012

I'm not sure if this is relevent to this thread, but a while ago I played a premade 5v5 normal game with some friends. Two of my friends were brand new to LoL and three of us were level 30's playing on lvl 1 smurfs (We did this in order to prevent being matched against other 30's, who would stomp the new players on my team, the experienced players played as support Galio with unexperienced Teemo top, support Soraka with unexperienced WW bot and Ezreal for the first time mid so we wouldn't damage the game for any new playerrs on the other team). We got matchmade againt a team of 5 level 30's. 5x lvl 1's vs 5x lvl 30's. One of the two friends had such an unenjoyable game that he quit half way through and refuses to try the game again (He got an AFK kill at the end of them game due to an auto-attack on an enemy who dived the fountain). The other friend is now level 10 Here's a screenshot of the end-screen (you can check all of the summoners names to check levels if you don't believe me). Definitely bad matchmaking...

Stupid micro pic, see here: http://i.imgur.com/e3BOx.jpg


Comment below rating threshold, click here to show it.

OPTLawyer

This user has referred a friend to League of Legends, click for more information

Senior Member

02-29-2012

Quote:
Originally Posted by II Peter Pan II View Post
this is my game result
solo que at ranked

https://docs.google.com/spreadsheet/...Gc&output=html

dead zone 1000~1300elo..

depend on which team have afk or bad duo or seeding player(in batch)
Wow, that's really awesome. Nice spreadsheet

I believe the screen shot I've included here is a huge part of the problem (it's from this guy's stats). The other team had an average of 16 wins while his team averaged over 204. This is just extremely unbalanced. It appears that the current matchmaking only averages out ELO on both sides, but with the placement games, early ELO is not indicative of skill level...


Comment below rating threshold, click here to show it.

Au11

Junior Member

02-29-2012

I had put in a support request about this and I was told to give examples to thbis forum. So here is another. I just played another game where I was stuck with 3 level 30s (1 big feeder) and a really bad level 29. We were up against 5 level 30s. Again it was an unfair match because of the experience diffence between the teams. This is easily shown by the number of wins by the team members: My team had me with 857 wins and the other 4 members had 229, 606, 191 and 165 for a total of 2048. The opponents had 916, 262, 699, 727 and 552 wins for a total 3156. Thats a 150% experience difference. This is like taking a college football team vs a high school team. We surrended at 23 mins with a kill score of 14 to 29 and they destroyed 5 turrets and we had none. This was ZERO fun and is one reason why my kids dont play this game nearly as much anymore. I dont think that their team was premade (and our wasnt) so the matchmaking system should have swapped players around to make a more equitable game.


Comment below rating threshold, click here to show it.

Au11

Junior Member

02-29-2012

While we are all discussing this subject. What exactly is the math that LOL uses to determine matchmaking? It would make it a lot easier for the user group to help if we could see it.


Comment below rating threshold, click here to show it.

NeverPull0ut

Senior Member

02-29-2012

I think it would be really cool if draft pick normals were based more significantly on player skill and Elo, whereas blind pick normals were much more loosely chosen. As a ~1950 player, I usually get matched up with 1600+ players trying their hardest in blind pick games when in reality I just want to mess around with some item builds and champions. There is currently zero option in this game to just have fun with "troll" builds, roaming, strange summoner spells, etc. without ruining the experience of other people. For example, even if I queue with 5 people, and we all run promote and surge, we still usually win. The other team leaves the game feeling awful and unskilled because we were blatantly trolling and still won. On the same token, if I am only with a few friends, and we want to do something fun (AD Sona, support Ashe or something along those lines) and have a bad game, we will most likely get reported by our teammates for trolling even though we were 3/0 in lane and just weren't able to carry hard enough based on the comp.

TL;DR -- Create a mode for less serious, more fun games. Or turn blind pick normals into this. I don't play normal games to tryhard, UNLESS I am doing draft pick.