Which matchmaking 'issue' is the most important to you? (See post for more details!)

1) AFKs in Champion Select Lobby 4,881 36.23%
2) Duo-Queue Elo Disparities in Ranked 1,010 7.50%
3) Skilled Ranked Players in Normal Modes 668 4.96%
4) Premade Matching 674 5.00%
5) Transitioning from Normal to Ranked Mode 1,348 10.01%
6) Free to Play Champions in Ranked Mode 802 5.95%
7) Random Champions in Ranked Mode 647 4.80%
8) Provisional Matches in Ranked 723 5.37%
9) Duo Queue Prevalence in Ranked 423 3.14%
10) Level Disparities 652 4.84%
11) Team Margin of Victory 1,645 12.21%
Voters: 13473. You may not vote on this poll

After Hours with Matchmaking and Lyte

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

The Dork

Senior Member

06-09-2012

Here is another question and i'm basing this off my last match.

Why are only 2 rune pages needed to enter ranked? And how does this affect a team that needs different roles if the majority of players have only 2 pages that limit them to only 2 roles?

(just had a support leona with attack speed quints and glyphs. we lost)


Comment below rating threshold, click here to show it.

Dex7ter

Senior Member

06-10-2012

Quote:
Originally Posted by Lyte View Post
I am writing a new FAQ for matchmaking, since I often see the same questions raised again and again. Could you guys help me with which questions you want answered in the new FAQ?
1. sats in ranked like ratio k/d/a is consideration in ranked game or just only pure rating ?
2. why new players can he higer in the champion selection even if we dont know what they skill represent
3. can u have duo who is first and last in ranked ?
4. what is criteria for duo, it is fair to have 1 duo vs 5 randoms ?
5. why duo is allowed on ranked
6. why some1 is first or last in champion selection its only pure rating ?
7. why new players are threat like experienced players, why they can ban or they place in champion selection lobby.


Comment below rating threshold, click here to show it.

The Dork

Senior Member

06-10-2012

Should there be an emphasis on educating the importance of respecting the pick order, and building a well rounded team?


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

06-10-2012
217 of 362 Riot Posts

Quote:
Originally Posted by Turbobuns View Post
Should there be an emphasis on educating the importance of respecting the pick order, and building a well rounded team?
We haven't explicitly said anything about pick order before. Honestly, in the average game, being first pick or last pick isn't a significant Elo difference in the grand scheme of things so first pick shouldn't believe they are the 'best' player on the team and therefore can pick whatever role he wants.


Comment below rating threshold, click here to show it.

simpsonboy77

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

Senior Member

06-10-2012

Some questions for the FAQ.

1. Why does your starting elo not matter? (Answer is that is it relative)
2. What does ELO actually mean? Roughly how many elo points are 1 standard deviation?
3. What does elo have to do with matchmaking?
4. Why will my elo find its 'true' value after a large number of games?
5. Why did I win and gain 6 elo, but when I lost I lost 12 elo? Is the system broken? (Explain how matchmaker tries to predict the outcome.)

Lyte I have a few questions for you.
What order does matchmaker make matches? I know its a balancing act between queue time, balancing teams, and predicting outcome. How exactly does it do this though?

Hypothetically I have 100 games played on an account and I am at 800 elo. I watched some streams and played normals and got better. Since my elo was fairly stable it had a low confidence value associated with it. If I go on a win streak of 9 or 10 in a row, will the system raise my confidence value because it no longer can accurately know my elo?

Waffles or pancakes?


Comment below rating threshold, click here to show it.

CupcakeTrap

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

Senior Member

06-10-2012

Quote:
Originally Posted by Lyte View Post
We haven't explicitly said anything about pick order before. Honestly, in the average game, being first pick or last pick isn't a significant Elo difference in the grand scheme of things so first pick shouldn't believe they are the 'best' player on the team and therefore can pick whatever role he wants.
On the other hand, I personally like the idea of the system providing some kind of default order. Ideally, sure, people work out a compromise. But it's nice to have a "neutral" fallback option. I don't like to have to feud or debate with my one-shot team. I think it's also easier to acquiesce to a system-generated order without feeling as though you've "caved" or lost face.


Comment below rating threshold, click here to show it.

KingMoonfish

Senior Member

06-11-2012

First of all, I've read pretty much 60% of your (Lyte's) posts in this thread, especially the last half, and I have a question before you start the new thread. (Or perhaps include it in the faq?)

Have you or the development team considered "role checks?" It would be similar to how WoW does dungeon queue, when you enter the queue, it asks what roles you'd be comfortable doing, such as support, carry, mage, tank, basically most of the roles already listed in the champions profiles in the LoL client. You can only select a role if you have a champion that fits it; especially now that free-play rotation champs are gone from ranked. All roles would default to selected if the player has a champion for that role.

I know that you answered many, many questions on the idea of what I shall refer to as "role stigma", especially in regards to making a player select a role. I'm not saying that this system force any player to play any role at any time, only that the random team of appropriate Elo rating will always have the (here's the keyword) ability to form a complete team. It still relies on player behavior and humans being humans, but it makes sure to prevent those odd times where you really have nothing to bring to a team because of bans/picks/pick order. The cost would be easily calculated, at least, as a coded flash screen and the backend needed to support filtering of matchmaking (although I suspect a similar system is already in place here.) This, imo, will balance out fluke wins and losses, encourage team play and diversification, and make the Elo rating more confident.

After having said all that, please realize I have only watched ranked play and the like, I'm not a ranked player myself yet. Perhaps this system could be used for normals as well, though?

-And a question unrelated to elo/ranked/comp play really, but what, if any, are Riots plans to make ranked more appealing (not to say they aren't already) to people say, like me, who watch live streams, recorded games, oogle over guides and such, but is too much of a chicken to click ranked?

(Edit: In retrospect, I see a downside of people only selecting AD carries or whatever they want to play, which could be both good and bad. If player interest is equally spread [I bet it isn't] then que times would be unchanged, I suspect a good many hate support though!

Perhaps instead of a selection, or even a screen the player sees, there is merely a filter for which class is "available" to that player, ie. what he has a reasonable chance to play as? To get into even more complex numbers here, since you know the ban ratios for all classes in the game, you could set that number+1 as a minimum to be flagged for that class. For example, throwing numbers out there, let's say 3 ad carries get banned on average in a game, if the player has 4 ad carries owned, his flag for that role would be set to "on". To get even more complicated, you could base this on how often each champ is banned on average in ranked play, and use that to get a weighted calculation. For example, say Graves is banned 65% of the time, then he would count as 65% less toward that number required. [A note: you can set an average here, so champs that are banned less count more toward the required total.] Obviously these numbers are much smaller, but you get the general idea. And finally, to completely complicate it, you could base these values on banned listings for Elo ranges, ie. banned percentages for 1200-1300 Elo ranking and the like. Complicated stuff, but since the player doesn't really ever need to see it, it doesn't have to be too simple.)

Warning: I made this post at 3 a.m., it's likely to not make much sense. ;D


Comment below rating threshold, click here to show it.

isobold

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

Senior Member

06-11-2012

Quote:
Originally Posted by KingMoonfish View Post
Have you or the development team considered "role checks?"
Has been discussed a lot. The problem with roles is: there are no roles in LoL. The players have just started to play a lot with similar style. But there is no predefined role in this game and even more importantly: there shouldn't be some. What you are referring to when talking about roles, is in fact just the current meta game that has emerged over time.

If you do role-checks, you have to decide which meta game you want to favor and set it in stone for the players. Riot shouldn't be doing this and afaik they won't be doing it.
If one day 3 support mid becomes a viable strategy, nothing in the game should prevent it from happening.

Fun fact: prior to the first Dreamhack, na players were using a totally different meta game and claimed their meta to be the strongest (just take a look into the fun discussions we had in elements tier-list thread on these forums). During the first Dreamhack CLG emerged as the strongest team at that time, but the EU meta, the one we are still playing today, proved to be the most efficient meta of all (CLG was able to win using the weaker meta by having better teamplay, but their trouble winning the matches convinced most players). If devs would have implemented roles prior to that Dreamhack, they would certainly have tried to set the na meta as standard, which in retrospective would have been a bad choice. So my claim isn't just pure theory. And it's not just old stories from two years ago. In the past we have seen variations on the eu-meta to be pretty successfull as well (for ex. the double AP meta). The game should be allowed to evolve in the future as well. Considering roles in matchmaking would hinder that evolution.


Comment below rating threshold, click here to show it.

IS1dcd9a4f2ff9a3

Senior Member

06-11-2012

Quote:
Originally Posted by Lyte View Post
I am writing a new FAQ for matchmaking, since I often see the same questions raised again and again. Could you guys help me with which questions you want answered in the new FAQ?
The most important one:
- Why is it always my team that is full of baddies?


Comment below rating threshold, click here to show it.

Knyaz

Senior Member

06-11-2012

Quote:
1) Intentional Queue Dodging in Ranked to lower their Elo to stomp low Elo players
We are revising the penalty for queue dodging in Ranked Mode. We intend to remove a player's ability to intentionally queue dodge from Ranked Mode and tank their Elo.


@Lyte ;

Won't doing this just ensure said players continually reroll low level accounts and crush "noobs" there?


Im not saying its not an issue , but its never been perceived as a huge one for the game. To me its better that some people spend a day or week or w/e it takes to dodge down , than make lvl 1 accounts over and over. Basically its faster to make a new account than it is to dodge a couple hundred elo. And a low elo ranked player is still on a higher footing than a completely new person to the game.

I recently made a new account on SEA servers and starting at level 1 , there was 1-2 of us "vet's" in every game just roflstomping. I felt bad and im pretty sure we made some potential long term players quit out of frustration.