Matchmaking Changes [1-16-2013]

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

Grey Tigress

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

Junior Member

01-19-2013

Firstly, I think 30 second wait time is no problem whatsoever. It's all relative, but I remember 15-30 minute queue times in some popular MMOs. That we get closer matchmaking consideration only sweetens the deal.

I do have two major concerns though:

1. Average. I read one of your posts saying that the -average- number of wins from a team is compared to that of another. How does this not lead to teams with huge win disparities among its individual members? For instance, yesterday I (with relatively high win #) was matched with teammates all with wins below 300, whereas the enemy team had only one below 800. In an exaggerated scenario for clarity: is it considered balanced for someone with 1000 wins to be teamed with someone with only 200 to fight a pair of 600 win opponents?

2. Counting wins seems to disadvantage players who are more skilled for their experience than average, and vice versa. Elo is already a dubious system for team-based matchmaking instead of its intended 1v1 scenario, so what's the rationale for further reducing the impact of individual performance? For example, a player who has 500 wins but is more skilled than most players with that many games of experience is now more likely to be paired with with people who have about that many wins, regardless of their own skill. Since this is a team-based game, that the enemy team MIGHT have the same problem does not ameliorate this scenario.

Wouldn't it be better to continue focusing on rating, but increase the queue time for more precise matchmaking?

Thanks for working so hard guys! Keep up the good work!


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

01-19-2013
12 of 46 Riot Posts

Quote:
Originally Posted by FAQdat View Post
Is that change live now? Additionally is this change only for na, or for other servers as well.
All servers, and yes it's live for now so we can collect more data.


Comment below rating threshold, click here to show it.

Yordle Swordler

Senior Member

Comment below rating threshold, click here to show it.

Ryios

Senior Member

01-19-2013

Language checks would be harder. Most people that que and speak spanish know english and choose not to speak it.

What I think they should do is add a language selection to your profile, then use something like google translate.

So if I have english checked and someone on my team has spanish checked then when I type something the translator should convert it to spanish for the guy that speaks spanish and everyone that speaks english would see what I typed without the translate, and vice versa.

It's not perfect, but

"doing good you are"
makes more sense to me than
¿Está haciendo el bien //which might be wrong (translated)...


Comment below rating threshold, click here to show it.

Ryceman

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

Senior Member

01-19-2013

Pretty sure I've been getting worst match ups to the point where my teammates just seem down right stupid.


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

01-19-2013
13 of 46 Riot Posts

Quote:
Originally Posted by Cyanide Velocity View Post
This is a great change, especially for those of us who leveled off of bots for a very long time.
Right, the feature takes into account number of wins specific to the queue you are playing in, so if you leveled up in Co-op vs Bots to Level 30 you don't face 1000+ win players immediately when you switch into Normals.


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

01-19-2013
14 of 46 Riot Posts

Quote:
Originally Posted by Nicksiren View Post
Riot/Lyte I have a big problem with this. I have only really played 3v3s to get to my ~320 wins, but now I have wanted to try 5v5s.

Why are the number of wins of ALL queue types being taken into account when playing a certain maps instead of ONLY the number of wins for the mode you are queueing for?

I suck hard at 5v5s but I am fairly decent at 3v3s. When I attempt 5v5s, why should I be matched up against players who have similar number wins as I do, except they have wins ONLY in 5v5s? Many players do not play literally anything other than 5v5s, whereas I'm the opposite. This can be vice versa for the other maps too, for players who may play only 5v5s but try a dominion/TT game and get matched up against people like me who only have wins in our respective game mode. Why are you not simply just weighing specific game mode experience instead of an amalgamation of all types?

PS: I've had a running issue with honor ribbon problems for awhile now and have not yet had a response to any of the threads I've made for the past 3 weeks. Could you please check it out? Sorry for tangent.
http://na.leagueoflegends.com/board/...8#post33605958
It does take into account only the number of wins in the specific queue you are playing in.


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

01-19-2013
15 of 46 Riot Posts

Quote:
Originally Posted by PissedoffBuddha View Post
I have to disagree with most people. I feel this change isn't fair. I worked hard in normals to get higher in normal elo. I was only 1300 in ranked but my normal was up at around 1575. I was finally playing with tough opponents and food teammates. These people had 2K wins and I only have 370, but I have so few loses that I was up there. Now I am back playing with really bad teams and its not fun. Don't matchmake people based on how many hours they play, that doesn't mean they are good. Match people based on skill.
The matchmaking system still prioritizes skill ratings. It's not just using Number of Wins.


Comment below rating threshold, click here to show it.

Ryios

Senior Member

01-19-2013

I just played a match where everyone had about 300 wins... Yet we had two guys just off doing w/e they wanted not engaging in team fights and then trying to win them after 3 of us died.

Wins doesn't mean everything. However I feel the probability of getting a good team with equal qins is better than a bad team with bloated wins.

I think it should only count your wins over the last 12 months. Tons of people have 1000+ wins from 2008, haven't touched their account in a long time, or only have 30 or so wins since they got back in the game.


Comment below rating threshold, click here to show it.

NEVER HAM ENOUGH

Senior Member

01-19-2013

Quote:
Originally Posted by Lyte View Post
In the latest game update, we rolled out a second beta test of a new matchmaking change that is live in all territories. This change will increase your average queue time by about 30 seconds, but we hope the benefits are worthwhile.

--

The matchmaker will now take into account the Number of Wins a player has in every matchmade queue (Normal SR, Normal TT, Ranked Solo/Duo, etc).

This means that for the first few minutes, the matchmaker will attempt to find teammates and opponents that have a relatively similar Number of Wins as you in addition to trying to find teammates and opponents that also have a similar skill rating. However, this does not mean it's impossible to get matched into a game with a large difference in Number of Wins--it's just more rare.

We need your help! Please give us feedback on the changes.

PS - RiotKiddington was the dev that did most of the work.
Can you please revert this for Dominion ques? I literally was in que with a group of friends for 18 minutes once and 14 minutes later because of this (usually takes us under 5 minutes).