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.

Bennyandthejets

Senior Member

02-16-2012

Quote:
Originally Posted by Lyte View Post
Hi everyone, I am investigating several potential upgrades for the matchmaking systems and am interested in collecting some feedback from you the players! If you have played a game you felt you had no chance of winning, or a game where you felt a player or two was too high or too low skill for the game, help me improve the matchmaking!

*PLEASE e-mail (Matchmaking [[[at]]] riotgames.com) these details directly. Thank you!

1) What is your Summoner name?

2) What was the game mode? (Normal, Normal Draft, Ranked)

3) What was the map? (Summoners Rift, Dominion, Twisted Treeline)

4) Were you queuing with a friend?
a. If yes, how many friends? (1, 2, 3, 4?)
b. If yes, and ranked, what were you and your friends Elo(s)?
c. If yes, and normal mode or normal draft, what was the Level and win/loss of you and your friends?

5) If you are below Level 30, do you have a Level 30 account?

6) Any specific reason(s) you felt the game was unbalanced?

7) A screenshot of the end-game score
Interesting how you don't even ask how many games we have won in comparison to those of our opponents. Why is this not taken into consideration? If my LoL career consists of 300 wins & I am playing against players with 1500 wins, I am playing against players with at least 5 times my experience level. This is what should be taken into account. Not just win/loss %.


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

02-16-2012
55 of 173 Riot Posts

Quote:
Originally Posted by shaytalis View Post
What if the K-factor varied based on the champion you were using? So the first time you play Annie ranked, there is a larger K-factor than if you are playing Vayne for the 40th time.
There is potential for abuse here. For example, take a current player who is absolutely top notch with Annie. Let's say this player's true skill is around 1800.

This player could play other champs to around 1800, then start using Annie--with a high K-factor, he can now blow up the ranks towards the 2000+ range, even though he theoretically should just be a 1800 level player.

Alternatively, players who are quite strong with multiple champs could 'rotate' through the champs to maximize K-factors in every single match, propelling him far above his true skill range because he is essentially getting bonus placement matches.


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

02-16-2012
56 of 173 Riot Posts

Quote:
Originally Posted by CupcakeTrap View Post
Wouldn't allowing people to pick their desired win percentage lead to problems if choices were systematically skewed in one direction? I do like the idea, though. Different Normal Elo "profiles" would also be fun. I think it would be best in combination with a system for flagging your role prefs. e.g. a Veigar-mainer could flag "mage" as their role and use their "SRS BUSINESS" Elo profile.

It seems like this would be better than relying on MM to realize "Oh, CupcakeTrap is trying to play Riven now; better drop that Elo like a rock to find reasonable matches."
I hesitate to let people choose their 'desired' win%s. Explaining a system like that to a player is complicated and we risk overwhelming players with math when it does not necessarily add value to their experience.

Allowing people to flag their role, or the many posts suggesting 'group finders' are neat ideas; but very difficult to implement. This is not to say it is impossible, but it is an idea that we must tread carefully with and really consider the advantages and added value it provides our players.

One, we have to be careful we do not recommend or push metas or 'suggested team compositions' on our players.

Two, we have to be weary of some roles or lanes being less popular, and increasing queue lengths into unsatisfactory levels.

Three, we do not want to profile a champ as a 'jungler' or 'mage' necessarily. One of the great things about League is that you can build champs many different ways. Some players build AP Master Yi and that works. Some players build AD Twisted Fate, and some build AP Twisted Fate. We never want to stifle creativity. If we label the vast majority of champs as 'viable' for multiple roles... then I am not sure we are actually helping the situation all that much.

There are possible solutions to each and every issue; but, tackling all the problems elegantly is not easy!


Comment below rating threshold, click here to show it.

timtwins

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

Senior Member

02-16-2012

Wow Lyte. I have to say I am extremely impressed with how well you have kept up with this thread. 52 pages and you are still commenting. It's Rioters like you that make League of Legends such a great game. Thank you for all your hard work and dedication


Comment below rating threshold, click here to show it.

Terrible Guy

Member

02-16-2012

1) What is your Summoner name? Terrible Guy

2) What was the game mode? (Normal, Normal Draft, Ranked) Ranked Solo/Duo is where I notice match making problems more often

3) What was the map? (Summoner’s Rift, Dominion, Twisted Treeline) Summoners Rift in regards to question 2

4) Were you queuing with a friend? I so do both solo and duo queuing often.
a. If yes, how many friends? (1, 2, 3, 4?) me and one other
b. If yes, and ranked, what were you and your friends’ Elo(s)? I've been as low as 900 elo and done up to 1700 elo games with friends.
c. If yes, and normal mode or normal draft, what was the Level and win/loss of you and your friends?In normal mode I'm 600+ wins with 500 loses. I was way over that before, but started screwing around with friends who I know are bad, but I play with to have fun. I'm an adult and not everyone of my adult friends are all that great in this game but I still play with them.

5) If you are below Level 30, do you have a Level 30 account? N/A

6) Any specific reason(s) you felt the game was unbalanced?My problem with match making is getting the same losers over and over when redoing a game. Or being on a losing streak and after the game finding out everyone on my team is on a similar losing streak while the other team is all on winning streaks. So I'm stuck with constant trolls. Another is being stuck constantly with people just reaching level 30 and deciding to try rank for the first time. They always end up on my team and sucking when I'm in the 1000-1300 elo range.

7) A screenshot of the end-game score no screen shots at this time

I am level 30 with 1300 elo right now. I've bounced around a LOT. I can tell you that my biggest gripe with the matching making is in regards to question 6 that I answered.

I can't tell you the number of times I've been stuck with the same exact troll/afker/feeder multiple games in a row if I re-queue. It makes me shake my head in frustration. I remember not to long ago when I went on a 5 game losing streak late one night. Everytime I queued I got the same intentional feeder on my team. I mean INTENTIONAL as in running up straight mid entire time going 0/20+ by the 12 minute mark. I was pissed cause he kept being placed on my team. His history showed he did nothing but that. So as I sunk lower in elo that night due to that dip-dung, I darn near quit playing LoL.

I've been regaining elo slowly since my last massive sinking to 900 ish or so. I've basically tried to make sure I only duo queue with my one friend I know is pretty good. We never do more than 3 games in a row. If we lose any game we stop. No point in getting stuck with the same trolls or bad players again which happens way to often. We also try to play during late prime time. Just after most of the kids have gone to bed, but before the really bad foreigner players come out.

Many prepubescent kids have horrible attitudes in this game and troll majorly. They make a one mistake like give up first blood and blame the rest of the team for it. Then continue to troll the rest of the game.

The foreigners are mainly just bad. Not all, but those from countries south of the USA are typically horrible. There is a good reason for that "Morde es #1 heuheuheu" running joke. That's assuming you can even get them to speak English and follow directions or act in concert with the rest of the team.


That 1000-1300 range is dubbed elo hell for a reason.


Comment below rating threshold, click here to show it.

일부분

Senior Member

02-16-2012

I just played a game where I was 1900, and had 3 1800's and a 1600 on my team against 5 1800's

I don't think this was fair...the 1600 obviously didn't play at the skill level or understanding of a platinum and often almost threw the game when we had a lead in all 3 lanes.


Comment below rating threshold, click here to show it.

USnip

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

Member

02-16-2012

I was duo queuing a lot with a very low skilled player in normals recently. However I don't know what the expected behavior is, hence I don't know whether anything I saw was worthy of being screenshotted.


Comment below rating threshold, click here to show it.

calicfer

Senior Member

02-17-2012

Quote:
Originally Posted by Lyte View Post
I hesitate to let people choose their 'desired' win%s. Explaining a system like that to a player is complicated and we risk overwhelming players with math when it does not necessarily add value to their experience.

Allowing people to flag their role, or the many posts suggesting 'group finders' are neat ideas; but very difficult to implement. This is not to say it is impossible, but it is an idea that we must tread carefully with and really consider the advantages and added value it provides our players.

One, we have to be careful we do not recommend or push metas or 'suggested team compositions' on our players.

Two, we have to be weary of some roles or lanes being less popular, and increasing queue lengths into unsatisfactory levels.

Three, we do not want to profile a champ as a 'jungler' or 'mage' necessarily. One of the great things about League is that you can build champs many different ways. Some players build AP Master Yi and that works. Some players build AD Twisted Fate, and some build AP Twisted Fate. We never want to stifle creativity. If we label the vast majority of champs as 'viable' for multiple roles... then I am not sure we are actually helping the situation all that much.

There are possible solutions to each and every issue; but, tackling all the problems elegantly is not easy!
I think the group finder would be very cool. I think it could actually promote meta game diversity if done right. You could find random groups who WANTED double jungle, invade, kill lanes. If I knew I had an ear and that my work would go towards improving the game, I would spend some extra spare time making a report on the current group finder ideas. I'd also run polls to gather criteria on how well the system would have to perform to be "a success." I'm busy at work right now, but I could probably put some real time into it next week or the week after.

The "elo profile" bit would be easy to implement and give the Elo system more match making knowledge about players. You just let players select their "play profile" (1, 2, 3, 4, 5) in their player profile using radio buttons. Then you use their corresponding elo. This would be like having smurfs that you play differently. These "elo profiles" or "play profiles" would only be applicable to normal games. The suggested use would be for trying to que as different rolls (you don't do any special matchmaking to get them the role, just to match them against a team that would be a fair match given they get the role). The other suggested use would be to use one profile for learning new champions, trying silly strategies, or new builds.


Comment below rating threshold, click here to show it.

calicfer

Senior Member

02-17-2012

Champion ELO would be a pretty cool stat. I don't think it should be used for queing, but it would be fun to know that I function at x,y,z elo with my main champions.

The win % doesn't really tell you who you're best at. (eg. you only play support when you que with your ad carry friend who is allways under ranked, so you're soraka has an inflated win %) (also if you played certain champs only at the bottom of loss streaks).

Quote:
Originally Posted by Nnyone View Post
A quick thought based off the data you have and the players have from ranked and the k-factor you mentioned. or maybe not so quick.. hmmm...>.>

Perhaps a vague elo adjustment could be displayed for the player on his ranked stats base off current elo compared to win rate/stats with that champ. (E.G. my win rate with karthus is 80%, since its above the 50% elo tries to achieve my karthus is considered 1560 elo when my actual elo is 1200. Conversely my Master Yi is a 25% win rate so my elo on him is considered 600 ).

Reasoning behinds this is at the very least a player has a better idea of what they are better at.

The next level of this would be players having an option to queue intending to play a certain role (support, top, mid, jungle, bot lane carry) where the combined k-factor for all the champs tagged in that role is used in matchmaking as an adjusting factor. (E.G. I've got a high stats(assists etc) and winrate with janna/sona/alistar, so the MM says ok I'll consider you as a higher elo player for this next match and also try not to pair you with another player who wants to play bot lane support. When I get in to champ select MM has me flagged as wanting to bottom lane support, and even if I pull out 0 CS ashe or something non traditional but win as bot lane support its a +elo to that role for me as well as the regular rewards, and possibly adds that instance of that champion to the role pool for adjustment.

TlR
1: players can see a vague guess elo adjustment on their ranked champ stats based off win rate and stats (normal 1200, on this xyz champ ~1600 elo)

2: MM uses this adjustment in combo with a "Hey I want to do this role" selection when queuing to get a better match and lower chance of role overlap.

3: role elo adjustment continues to accrue off your games played and champs used in those roles when you choose to use the group/role queuing system solely for use in matchmaking and your eyes only stats.

4: if this addition is successful your role elo should eventually match your real elo if you specialize/practice in a role or are equally good/bad at all roles.


Comment below rating threshold, click here to show it.

I So Pwned You

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

Senior Member

02-17-2012

Quote:
Originally Posted by Lyte View Post
I'm sorry to hear you had a poor experience while leveling up in League. I am trying my best to fix scenarios like this so new players can have a more enjoyable journey towards Level 30.
Thank you for taking the time to reply to my comment, and nowadays i am very excited when i have the time to play league of legends, as i find that majority of the summoners are nice people and are nothing like the people i met at the start.