EvrMoar

Thanks! My job is definitely a change of pace from what I was doing before. I was a systems designer on COD and the last thing I worked on was Warzone. It's nice being able to dive into such an awesome product, with an awesome dev team, and really engaging with the community to try and make a difference - I've had a blast so far.

There are so many learnings from ranked, elo systems, match making, so we definitely can't sit still and think we know what is best. Trying to constantly improve it and get feedback is super important.

Top radiant is probably most effected by win/loss streaking. There are all the factors I listed above, and as teams start to get mixed up and elo ratings start to get mixed, the system is pretty good at seeing who should be rising up and who shouldn't. Of course, this is win/loss base and not performance, but even just using win/loss alot of things can happen.

I unfortunately do not know enough about combat score and what stats it tracks. My guess is that the combat score doesn't take into account every individual players skill, where as the elo system does.

So the combat score may be saying "you get 5 points for a kill"(I made that number up). Then the elo system is going to go "You are X elo, opponent was Y elo, do some calculations - that kill was worth 3.14"(again made up numbers).

Obviously we don't want anyone to be able to figure out how we do things, so there is probably always a level of transparency we can't offer(this also can prevent figuring out exploits). But maybe this is an area we can improve visibility on someday.

Also I could be wrong because I don't know how the combat score works, unfortunately. I'm still learning and asking questions about the system as I'm diving in. I'll try to figure out what it's doing so I can talk about it more informed in the future. The above text was purely a guess.

Unrated, because it requires you to play it before going into ranked, is definitely used as a "guideline" on where to test you in your placements. Due to it being unrated it's very loose, but it's the only info we have on you that directly translates to the bomb game mode.

That being said I believe the system is very good at figuring out where you belong during your placements, or at least shortly after. I actually think the changes requiring games won over games played will help the lower ranks and placements(creating more fair matches).

Because some players don't have to try in their first 20 games to get to ranked, or you have people afk'ing/voting to surrender, it really throws off their initial testing/variance area when stepping into ranked. Now that we will be requiring unrated wins, players will have to try in unrated. This will give us a better idea of where they may belong so we can test them better in placement. That also means you will have less players ...

I can confidently say we haven't changed gains or losses for ranked, or any major changes that would result in this post or what you talked about.

I think the major factor that may be playing a role for you, and possibly this post, is the variance I talked about. Think of variance as how confident the system is that you are at the correct rank. The more confident, the less change you will go up and down. The less confident, you will swing a little harder then others. If you aren't seeing lots of movement in ranked it's probably because you've played alot of games, you play regularly, and you probably have a pretty even win/loss ratio; if you were in a lower rank where performance is factored in you may also be performing at that expected rank on average. So until you start win/loss streaking or out performing in a low rank, the game is going to be confident you belong at that rank and your movement may slow down.

If I were to be honest I think a big problem that lea...

Hey guys, for those of you who don't know me I'm the new competitive designer on Valorant! So, I'll try to explain this as best as possible - because I've only been at Riot for a month and have just learned how most of this stuff is working.

First off, remember that the arrows are only reflective of your current rank not your teammates. Because you may be sitting higher in immortal elo then your teammate you may gain or lose more based on match outcome. I don't know the exact elo calculation numbers on the backend so I'm going to make up imaginary numbers as an example. If you had 13 elo, but your friend had 6 elo, and lets say the enemy team(on average) had around 6 elo, you would most likely lose more elo because you lost to opponents way below your elo.

So yes, higher ranks are based mostly on the win/loss, but how ELO systems usually work is gathering the average rank of both teams and you gain/lose elo based on the match outcome. In essence a lower skilled pla...

I don't want to give too big of a response because I think I can give you peace of mind easily. I'm new, only been at Riot a few weeks, but I've already had conversations about both afk and smurfing. I don't have anything to share, but these are issues that we will constantly be working on. Hope that helps, I want the best competitive space we can get(and that's also my role on the team :P ).

Yea these things are tricky. Luckily I get to work with a social team that specializes in these things! I don't want to overly speak for them, because this is their area. I do want to work with them, especially when it comes to ranked, to find a solution. We will find something, I'm sure if it. If that solution doesn't work out, I'm sure we will go back to the drawing board and try again. We need to get it right for everyone, I want the best ranked experience just as much as everyone here!

Yea that's true, I like the idea that you can work your way out of it for sure. You want to let people learn from their mistakes, there are exceptions obviously. This is more of the social teams space, but being that it effects competitive we want to work together to find solutions!