EvrMoar

I don't know where you've seen this, but we don't see this at all.

One of the hardest data points you see, in ranked, is that more time = more skill for players. In most cases, if you were to map time invested:skill you would see a linear line for the community. The problem is the entire community is all getting better at the same time. So to rank up you not only need to get better, but you need to get better faster then the people at your rank.

I don't have hard data to argue more cheaters, trolls, new players, etc. But I will say more players are more likely to play the start of a new act/episode because of new content coming out. So if anything I would say you are more likely to run into those cases at the beginning of an Act/Episode then the end.

If you do have data to support this I'd love to see it tho!

Players will get 0-5 RR on average for a performance bonus, with 3-5 being average, and yes it's based on their average performance.

So if a player normally performs middle of the pack, with average ability usage and kills - if that player has a 30 kill game where they used their abilities effectively they would get a performance bonus. performance bonus can go above 5 tho, but it's very difficult too get it that high.

To start, I really like to avoid doing breakdowns of players match history if I can avoid it. There is a lot more intricate math going on, in the background, to determine your MMR/Rank/RR Gains. But I think this one is pretty easy to understand so I definitely wanted to put in my 2 cents.

If you look at your match history yes you have two games where you broke 30 kills. But you also have way more games where you only got around 10 kills. You also only have 0.9 kills a round. That means you are dying more then you are getting a kill in a round. You also have a win rate that is below 40%. Unfortunately that means you aren't performing at your Rank, and in fact are losing more then you are winning. In any ranked system that means you are probably ranked too high and you are on your way down.

Now that sucks, but you also haven't played A TON of games(but you've played a decent amount). It could just be a bad streak of games and if you had more time maybe you would start...

I don't determine what is or is not punishable, I don't make the rules or enforce them. I just want to be honest because it isn't my job. There are also legalities in certain countries involving this kind of stuff. I just forward stuff like this to player support and if they've broken rules they know what punishment fits the crime and how to investigate the issue.

That being said I don't want to run into this type of behavior, also I know lots of us don't want this type of behavior in our game. This is pretty damning evidence they were being toxic, so take that however you will :)

Not all Rioters enforce, or make the rules. But collectively, as a whole, we our vocal about issues we run into. We also are aware, and vocal, about the issues the community brings up on social media and what is happening in game. My experience so far at Riot is that we are anything but quiet. We will make sure issues are heard and we don't stand for toxicity in our game. Every Rioter is e...

Your reports do matter, and we have improved report features coming online soon:

https://playvalorant.com/en-us/news/dev/valorant-behavior-detection-and-penalty-updates/

Also there are plenty of Rioters that also forward these issues to player support, me being one of them. So even tho a Rioter doesn't say anything, you can guarentee if we see it on twitter, reddit, etc. we try to report it(If we have the time of course). This doesn't mean spam us with every issue, if we tried to be player support we just wouldn't have the time to do anything else. My first recommendation is still to report in game, and...

This isn't the case for Valorant. I, personally, won't push to add anything like that and I only have minor changes I want to investigate for our MMR system currently. Now we are digging through data, as well as how the new Ranked Rating system is working, to plan changes for the future!

Same :(

You're best metric of if your rank is at your MMR is if you are gaining around the same amount of Ranked Rating per win as a loss. That means your Rank has converged with your MMR and you are at your rank you currently belong at.

That being said, some players are climbing up, some players are being pushed down, so everyone is on their own Ranked journey. So because you have a lower rank player in your game doesn't mean you are going down, it could mean they are on their way up.

I hope that helps, and thank you for the kind words!

We didn't change the match making system. You are getting the exact same matches as you would have last episode/act. In fact we didn't even touch players MMR.

So the only explanation, if you really feel like there are more smurfs, is not the match makers fault but realistically that more smurfs are being made. We are still refining how we find out who is smurfing in data, but right now it's actually a very small % of the playerbase.

I'll definitely get some flak for saying this, but I think there is a perception players are smurfing more then it's actually occurring. In shooters it's not uncommon for a player to have a really good game, and instantly have a bad game in their next one. Just because a player may have a lower rank, and do well, doesn't mean they are smurf. They could just be having a good game, or are in your match because they are actually climbing up rank. Players are on their own journey all the time, so players will climb up in rank. Everyone will ...

Because players stop playing the game, and new players rise up. The top players are constantly shifting, even monthly. You need to soft reset to to clear out high ranks where a player may have stagnated at their rank, and may not really be engaging with Valorant or even a top player to begin with.

Creating a restriction where "You only have to play 1 game a week" or docking Rank can only get you so far. At the end of the day a player could coast, for multiple seasons, at the top rank without much effort if we didn't soft reset.

The other issue is cheaters. If we let people climb too fast, you could get an account onto the leaderboard. We need a few games to actually determine if you're cheating or not. We are good at detecting cheats but sometimes it takes some games.

Also if we let TenZ get the 1 spot, right off the bat, the leaderboard would literally be 10 people, and all of their alt-accounts, sitting in the top 250. Radiant would be full of alt-accounts...

Rank is always an illusion until you put in enough games, even in the old system. You weren't Radiant, in any season, right off the bat. It took some games to get there. No matter what ranked system you play, you have to put in some games, to prove you belong at a rank and climb there.

While you may have been playing other Immortal/Radiant players, while being plat/diamond. You were getting +35-50 per win and only -10 for a loss. Do you think it's fair if we, after 1 game, just gave TenZ the #1 spot on the leaderboard just because we think he belongs there? Unfortunately every Elo/Rank system needs you to play games, to figure out where you belong. We can guess all we want, but until you actually give us enough matches to have a big enough sample size, we won't know your rank until you put in some games.

Also we are pretty generous in letting you move upward, and even increased how fast you can increase within a patch of the initial new ranked system launch.

...

!pin

I read this before you changed it and I wanted to give my opinion, just because I'm the designer of it.

So I understand that you feel like it's not about personal skill, but we actually have two MMR's that make your singular MMR. A performance based MMR and a Win/Loss MMR.

In lower ranks performance has a bigger impact on your MMR. This is because it's very easy to see who is probably not Iron based on how well they are aiming, how often they win duals, how they use their skills, etc. Players are usually at low ranks because they aren't consistent, and part of that reason is because they don't know every map, every call out, how to play multiple agents etc. As you climb in ranks the difference in aim ability, reaction time, agent usage, etc. starts to shrink. At the highest levels, team play and game sense become the main factor in winning a match, and that's when you are solely being rated based on your wins/losses.

Just to be clear, your performance MMR ex...

I think I understand where you are coming from, wanting RR to drive MMR, but the whole point of Ranked Rating was to smooth your Rank swings and give you an easy to understand point system.

You aren't matched based on Ranked Rating, and I sometimes see the argument where you should be.

"I want to beat silvers, then golds, then plats, etc. to get to my actual rank."

In concept, and not diving deep, that seems like an okay idea. But there is a reason this isn't used in most games. Things like Microsoft developing their True Skill Elo System(which most big games use and license), as well as the Elo system in chess. The reason being is that by beating a Silver player, it does not mean you are actually Plat or wherever you think you belong. It only means you can beat a silver player. Could you imagine us throwing the whole player based into Iron/Bronze and giving them the same RR per win/loss and matching them based on their RR to climb? You would have teams of ...

That curve is before I changed distribution. I don't have an image of after unfortunately. Now the highest points are in silver, and curved down both ways from there. Instead of having an insane amount of players in iron.

Here is a great example of what the curve was like before Episode 2/December:

https://www.esportstales.com/valorant/rank-distribution-and-percentage-of-players-by-tier

I can't confirm if that's the exact, correct, data from december/early Jan but it looks close. Maybe at some point I can get the okay to release what it is now vs what it was, but we are going to be making adjustments each act so I don't want people to focus on it too much.

My goal for distribution was...

Our match maker is actually pretty good in match making, of course that's just data. I think part of the problem is FPS games can be very unpredictable. Having a bad match, where you are just not landing shots, can really throw the fairness of a match. So while we expect you to perform at a certain level, sometimes people have bad days. The same can be said about people having good days.

I've seen the feedback on Russian comm's, just so you know the team is aware of it. It will take time but I think we are on a good place and only will get better. Thanks again for reading and you have definitely been heard.

Thanks for reading! I don't expect a ton of people to really want, or need, to know the inner workings of ranked. But I think it's important to have discussions, because it helps find the real problems. If you are frustrated, on top of not knowing why, it creates a huge barrier between you and the dev team in finding a solution.

Thanks for the kind words!

You can say your last act rank, but really it's just your MMR. Your MMR is how the game finds your matches, and what it uses to weigh you against other players. The game will push you to your MMR, so the better you do the more you raise your MMR and the more RR you will gain per win.

You will start to catch up with your MMR, and then your win/loss amounts will even out.

Yes! Right now you are being tested at a higher rank. So when you win a game at that rank you earn more, and if you lose you will lose less.

If you keep winning, you'll climb to that rank. If you were to lose more then you win, the system will start to test you lower and you will climb to a rank slightly below the one you are currently being tested at.

After placements your first 20-30 games will be testing you to find your true rank, after that you have to keep improving and climbing the old fashioned way :)