Matchmaking Changes [1-16-2013]

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

Avengè

Recruiter

02-01-2013

Quote:
Originally Posted by Captain Duke View Post
Please check my last few games, especially the last 2 ones, I can feel thats something is wrong:
(Premade with 2 friends, normal non-draft queue, looking for some fun and thats what I got and I'm pretty sure the randoms are duo queued)

Match 1:
Me - 1140 wins
Friend 1 - 687 wins
Friend 2 - 758 wins
Random Player 1 - 355 wins
Random Player 2 - 186 wins *sigh*
vs
Enemy 1 - 764 wins
Enemy 2 - 427 wins
Enemy 3 - 200 wins
Enemy 4 - 739 wins
Enemy 5 - 677 wins

Match 2:
Me - 1140 wins
Friend 1 - 687 wins
Friend 2 - 758 wins
Random Player 1 - 138 wins
Random Player 2 - 255 wins
vs
Enemy 1 - 491 wins
Enemy 2 - 473 wins
Enemy 3 - 885 wins
Enemy 4 - 102 wins
Enemy 5 - 545 wins

Okay so how do you guys think this is nearly closed to balanced ? Its not balanced for my team nor for the enemy team. What is the thought process envolving those changes ?
Something like "OKAY, this guy has 1140 wins, lets throw two buddies with 300 less wins with terrible stats so the games get TOTALLY balanced for their enemy and pop their queue quicker" ?

I found only frustation on the last games I played premading with 3 people or less, when we premade as 4 its okay because we can ignore the random one feeding hard and just carry out the game without his help.
Please check my games, those randoms had terrible stats and were picking weird positions.

I know that normals queues are made for fun, but thats too much for me.
I've realized that most players with over 1000 wins are worse than those with around 400-700ish range. And bad players are something you have to learn to deal with because some people just happen to get carried a lot.


Comment below rating threshold, click here to show it.

Death4Rent

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

Senior Member

02-01-2013

I have found that my games that i get into are still way lopsided. It is ether very even in wins or the entire other team has inbetween200-400 wins and all my teammate have 1000+ I also think that the que system should take into account of how many losses u have. I person who has won 1000 games but lost 1500 to get there is not as good as a person who has won 1000 games but lost 900 times. it just depends on who has played longer for number of games won.


Comment below rating threshold, click here to show it.

Captain Duke

Junior Member

02-01-2013

Quote:
Originally Posted by AvengŤ View Post
I've realized that most players with over 1000 wins are worse than those with around 400-700ish range. And bad players are something you have to learn to deal with because some people just happen to get carried a lot.
That is true indeed, I've seen many of them doing nothing or just feeding on lots of my games but the point is that the matchmaking changes are supposed to match players similar to your wins, because thats how they see your 'skill' level, to play with or against you and that is not happening at all


Comment below rating threshold, click here to show it.

NYwildlife22

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

Member

02-01-2013

ty for listening to me riot :3 <3


Comment below rating threshold, click here to show it.

ClaymoreNo47

Senior Member

02-01-2013

I read all of your posts and i had a question. You admit that there is a potential difference between a 1700 elo player that has 1000 wins and 100 wins. Yet you think its ethical to have potentially a game filled with 1700 elo players with 100 wins gain/loose elo off eachother and a concurrent game with 1700 elo players with 1000 games win/loose elo and potentially gain rank divisions off eachother. There needs to be unity, how is this balanced? What if potentially you have really low wins and there's so many players and you play at peak hours that potentially you only ever face/ get paired with very low win players all the way up that only played with other low win players. Could these players potentially have gained alot more elo as a whole than they would otherwise if they just randomly got paired with players base purely on their elo?

This new system is very troubling to me because it only really adds to the experience at low to mid elos. Last season i got to 2150 and really i could care less about any stats, wins or anything, I expected a certain kind of performance from my teammates unless they happen to have been elo jobbed but that's another issue. If you reached diamond last season legitly i really don't see how matchmaking skill discrepancies applied to you.

Also is it a goal worth working towards to design a system who's goal is to better track progress and performance but does nothing if not hinder it, the higher you get in the ladder?


Comment below rating threshold, click here to show it.

Dream of Glory

Senior Member

02-01-2013

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.

---

UPDATE :::

I'm crunching the data these days and finding some interesting things. I think for the sake of transparency, I'm just going to reveal the numbers I'm seeing.

One, the average game has a much closer win differential than before, so we know the feature is at least doing what it's supposed to. For example, we're seeing ~5% more games in Normal SR where teams only have a 0-49 number of wins difference between teams. In Ranked SR we're seeing ~30% more games with only a 0-49 number of wins difference between teams.

Two, we're seeing that the average game has a similar Elo differential than before, but I think this data point is misleading and we need to deep dive this a bit more. For example, we're seeing lower numbers of games with 20-30 Elo differences between teams which is an unintended effect.

Three, if we define a lopsided game as games with a huge KDA differential or something like 10,000 gold difference, we're not seeing more or less of these games post-patch. It might just be that players who are getting the lopsided matches come and post about it, and the players who are getting fair matches don't.

Taking all three points into account, we believe the Number of Wins feature is a net positive and don't believe it's creating more lopsided games than before. However, we have a few improvements we want to make to the feature--we will most likely increase the weighting for skill rating (Elo) so that skill rating is always much higher priority than number of wins. This will fix Issue #2 I pointed out above, while retaining some of the benefits of adding Number of Wins to the matchmaker.
Thanks for explaining things to us <3


Comment below rating threshold, click here to show it.

TheC√ł√łkieMonster

Senior Member

02-01-2013

Quote:
Originally Posted by Makair View Post
GRAMMAR NAZI ALERT!

itself*
.. unacceptable comment


Comment below rating threshold, click here to show it.

TheC√ł√łkieMonster

Senior Member

02-01-2013

Quote:
Originally Posted by AvengŤ View Post
I've realized that most players with over 1000 wins are worse than those with around 400-700ish range. And bad players are something you have to learn to deal with because some people just happen to get carried a lot.
Quality > Quantity

For instance...if you add up the accounts from the Mac Address of my 3 computers... you will find that since beta I have played near 5000 games (combines of normal, ranked and custom) and that while I love this game, I do not fancy it a job.

Yes Im a nerd but I work from home with a simple run on its own business and a single parent.

----------------

Even if you have:

1000
1000
500
700
2500

vs

54
12
19
400
245

You still have a variable there becasue those with more games, play more champs, yet those smaller numbers may play well together, be smurf/twink accounts, or play more serious than those with more wins.

Personally... I like playing people who are as hardcore as me because we have more in common


Comment below rating threshold, click here to show it.

Blue101

Senior Member

02-01-2013

this change sounded good, but I've seen no difference in the quality of my games. it just takes a bit longer to get into one....

I for one would not be sad to see the change reverted. however I guess 30 seconds isn't a big deal.


Comment below rating threshold, click here to show it.

Biginla12

Junior Member

02-01-2013

Do you lose league points if you dodge a ranked game?