Valorant Dev Tracker

It's going to vary a ton player by player and config by config. Reflex helps the most when you reach what we typically call a "GPU Bound" state. GPU bound is when your GPU "falls behind" your CPU. I have a GTX 1080 Ti at home and I still hit GPU bound scenarios because I play at 4K resolution with all my graphics settings cranked to max.

The important thing to note is that reflex will never increase your framerate. It's only impacting input latency. If you have the driver, there are also new settings in Video->STATS where you can turn on graphs for input latency caused by the GPU/CPU.

Some players may see little to no gain, for some players it will be night and day. If you have a higher end GPU you might try out boost as well but note that this will keep your GPU at max power usage so it's a tradeoff.

In general, I'd recommend all players leave reflex on. It can't hurt and there may be game scenarios where it can help smooth out input latency (e.g. ...

Depending on your gpu and video settings turning it on can have dramatic effects on input latency! I've seen scenarios where it reduces input latency by 8+ms! For comparison that's about as much as changing from a 60hz to 144hz monitor.

I'm super excited to give it a whirl on my home machine tonight! I play at high res so I hit moments in the game where reflex should smooth it out.

Boost will boost your GPU's clock speed 100% of the time. On NVIDIA GPU's if you fall below 40% GPU utilization, your GPU will start underclocking to save power & reduce heat generation. (fans will spin slower).

When you turn on boost you'll keep your GPU at 100% of it's clock speed for the trade offs mentioned. It can reduce input latency because it can reduce the time it takes for frames render start to finish. It won't necessarily increase your framerate though but you could see input latency wins especially on newer GPUs.

Boost is not necessary to get the gains from Reflex. Reflex by itself doesn't really have any downsides so we default it to on but leave it up to the user to enable boost or not.

Would be pretty neat to have a skin line named after me, but it's just coincidence.

Yeah, that's gonna be a sub from me dawg

That was her winrate on patch 1.06, and had been for quite a few patches before that too despite our smaller changes. Post-1.07 she’s looking a lot healthier, although still strong, in spite of sentiment floating around that she’s too weak or gutted now

We're going to do some tests to see which server is best for Pakistan between Bahrain and Mumbai. If it's Bahrain, no worries, Pakistanis still just start getting lower ping. If it's Mumbai we'll have to figure out next steps.

In terms of effort, it depends. Some things seem like they should be easy but are hard, and some things that you'd think would be hard are really easy. I'm not closely integrated with the team doing this work so can't speak to specifics; all I know is that it's being worked out.

We look at winrates in all mmrs, not just dia/imm. We also do actually care about people complaints (I field surveys every patch to understand how players feel about the state of the game, as do many of my colleagues), because we want to make sure the game is not only balanced, but enjoyable to play. But yeah, to your point, sentiment is not balance. If we have an agent that's balanced but miserable to play against, we do want to change that, but that's more of a pure design consideration than a balance one and we approach those problem sets very differently.

That was me that said that in the upper thread! We use a few metrics:

Game Winrate (non-mirror): What's this agent's winrate in games where they're not playing against themselves? i.e. if one team has a Sage and the other doesn't, what's Sage's winrate? We don't look at mirror matches here because every mirror match gives an agent a 50% game winrate (one wins, one loses), so the higher an agent's pickrate, the more their winrate converges to 50%.

Round Winrate: What's this agent's chance to win any given round on defense? On offense? We can look at this one in mirror matches and non-mirror matches, since we mainly use this to understand if an agent's power is manifesting heavily on one side of the map.

Pickrate: What % of teams pick this agent?

We look at all of these metrics (and a bunch of others, these are just the simplest) across all of our maps and all of our MMRs. For example, Sage's winrate i...

I'd say we have taken a more conservative approach with Sage since Closed Beta which has lead to the issue today. We were way off on what we thought her power level was early on and we've made measured changes over multiple patches to address that.

The issue today is that... well... we've hit her multiple times over many patches where it may of been better for us to be less conservative earlier and rip the bandaid off quicker. It sucks for Sage mains to constantly see her in patches. We think she's in a much better state today and is still impacting the game in a healthy way.

Let me know what you find. K, thx.

Those Sage mains are crushing it!

It's understandable why people have these types of responses to patches. Making changes that are sup-par is an easy way to burn trust with players. We are bound to make mistakes, and I personally have made changes back on LoL that took MONTHS to regain trust from the players

On VALORANT, we're taking a more conservative route to balance early on in the game's life-cycle. It's important that the game stays competitive and the time and effort that you all are putting into the game isn't wasted. Each week we are learning something new about how you all are playing the game, and we need to take that into account rather than making sweeping changes every two weeks.