Ultrasoldier

aim

Properties1
tagsaim, mechanics, motor_learning, sensitivity, input_latency, aim_trainers

13 min read

Aim

treactis

I don’t think I’m an expert at aiming. But I’ve played many different FPS titles over the years. That has given me enough experience to form my own view.

Sensitivity

I play VALORANT at 0.1 sensitivity and 1600 DPI. That’s about 82 cm per 360° turn. For a long time I played at 0.05, which is about 163 cm per 360°.

I switched because a very low sensitivity held me back in a few situations:

  • quickly checking an angle on the opposite side
  • moving my spray onto a target far off to the side
  • turning away from flashes in time, which mattered most

The higher sensitivity had two bonus effects. My hand gets tired much later, because moving the mouse takes less effort. My mouse skates also wear out more slowly.

Getting used to the new sensitivity was quick. It took 1–2 Deathmatches. After that it was smooth sailing.

Lower sensitivity still has real benefits, though. Long-range fights are much easier. 1v1 fights are usually easier too. When an enemy micro-strafes (makes tiny, quick side-steps), a lower sensitivity gives you more control to follow them.

Consistency and warm-up

I believe consistency is the key to improving. In FPS games, and especially in VALORANT, that means a regular warm-up and practice routine. The higher you climb, the more this matters. For me, it’s what makes progress faster and steadier.

Aim trainers

I used to dislike aim trainers. That changed after I researched them properly and started using KovaaK’s seriously.

I don’t think KovaaK’s made me a much stronger aimer. It made my aim much more consistent. That happened together with the rest of my routine, so I can’t fully separate KovaaK’s effect from everything else.

Finding the right scenarios took time. I tried a lot of playlists before I found ones that fit me.

My warm-up playlist focuses on balls that move randomly. In most scenarios, I try to predict where one ball will go next. I place my crosshair on its path, as close to it as I can, and hold it there. This trains reaction and reading the ball’s path.

In some scenarios the paths are too easy to predict. There I practice fast target switching instead. I mix two approaches. Sometimes I pre-place my crosshair and tap when the ball arrives. Other times I react, switch to the ball, and follow it until I hit it.

It’s worth taking time to notice which parts of your aim are weak. Examples are micro-corrections, flicks, tracking and target switching. Once you know your weak spot, you know what to practice.

My current KovaaK’s warm-up playlist share code: KovaaKsBackflippingMidCrucible

In-game warm-up in the Range

Warming up in the Range was just as important for my aim. My routine:

  1. Get 50 kills on strafing bots with bot armor on. I focus on precise tapping and on dead-zoning, with the highest accuracy I can manage.
  2. Do 1 medium and 1 hard challenge to practice fast target switching and flicking.
  3. Sometimes, if I feel like it, do one more full round with the Phantom.

Dead-zoning means shooting during the short moment of a counter-strafe when your gun is fully accurate. A counter-strafe is tapping the opposite movement key so you stop instantly. Dead-zoning matters because you keep moving during the fight, and a moving player is harder to hit.

You don’t have to tap single shots on the bots every time. Bursts of 2–3 shots are fine too. Imagine the most realistic gunfight you can. Your only goal is to kill the target as fast as possible.

Deathmatch

After the Range, I sometimes play Deathmatch. I’m not a big fan of it, though. The fights there often feel unrealistic, because many players there don’t really know how to play the game.

That’s my whole warm-up routine. I don’t have a separate aim practice routine yet. By that I mean playing KovaaK’s or Deathmatch just to improve, like after matches or on days I want to practice without playing ranked. So I can’t say anything about that yet.

Scientific Evidence

Aiming in an FPS is a fast pointing task. It happens under pressure. Like any pointing task, there’s a tradeoff between speed and accuracy.

Being good at aim isn’t only about moving the mouse faster. Better players do move and react faster. But they also land closer to the target more consistently. They need fewer small corrections. They fire a little earlier in the motion.

There’s no single “best” sensitivity. There’s just a wide range that works well.

Input latency is the delay between your mouse movement and what you see on screen. It makes aim measurably worse, even at very small amounts. But it affects people differently.

Aim trainers like KovaaK’s are good at measuring your aim consistently. Nobody has proven that practicing on them makes you better in real matches, though.

A lot of common aim advice goes further than the evidence supports. “Use this pro’s exact sensitivity.” “Changing your sensitivity ruins your muscle memory.” “Aim trainers are the guaranteed path to better aim.” These are folklore. This note tries to separate the folklore from the evidence.

Aiming follows the same math as any pointing task (Fitts’s Law)

There’s a well-known rule in movement science. The further away and smaller a target is, the longer it takes to move to it accurately. This holds true in FPS games too. It’s just measured as an angle instead of a straight-line distance.

Small, far-off targets cost more time. Every extra millisecond is time an opponent can shoot you first.

This matters for a simple reason. It reframes a lot of “aim problems” as positioning problems instead. If you’re already aimed close to where the enemy will appear, you need less correction. Less correction means less time exposed. That’s before your actual aim skill even comes into play.

Researchers (Looser, Cockburn & Savage) checked whether this rule really applies inside FPS games. It does. Their measurements were within about 4% of results from plain, non-game pointing tasks. So it’s reasonable to think about FPS aiming using this same well-tested theory. Games aren’t some special case here.

There’s a wide “good enough” sensitivity range, roughly 20–80 cm per 360°

A 2022 NVIDIA study (Boudaoud, Spjut & Kim) ran the first real experiment on mouse sensitivity in FPS-style aiming. It tested 13 experienced FPS players, who played 10–20 hours a week. Each did 500 attempts per session, across eight sessions. The study tested six fixed sensitivities, plus each player’s own preferred one.

  • The best performance happened somewhere between roughly 20 and 80 cm per 360° turn. That’s a wide range, not one exact number.
  • The differences between sensitivities were statistically solid. Speed and accuracy peaked somewhere inside that range.
  • Both extremes cost you time. Too low, and you physically can’t turn fast enough. Too high, and you need more small corrections to land the shot.
  • Harder shots — smaller, further targets — work better with lower sensitivity. Easier shots work better with higher sensitivity.
  • Against the “changing sensitivity ruins muscle memory” idea: the first and last sessions both used each player’s own sensitivity. As a group, performance didn’t drop significantly between them, even with sensitivity changing in the sessions between. Individually, about half got slightly worse. Nearly as many got better.

A few limits are worth keeping in mind. Only 13 people took part. Only stationary targets were tested. There were no moving or tracking targets. Turns were small, up to about 25°. Big turns, like checking the opposite side, were not tested. Scoped aim was not tested either. It was one lab, one study. This is the best evidence we have. That’s not the same thing as strong evidence.

Being better at aim means speed and control together

A 2022 study (Donovan et al.) analyzed 32 pro and semi-pro players in Aim Lab. It did this by tracking their actual mouse movements.

The better players reacted faster. In the easier task (Gridshot), they also moved faster. But speed wasn’t the only difference. They landed more consistently near the same spot. They fired a bit earlier in their swing.

All players also changed strategy depending on the task. With bigger targets they went faster and looser. With smaller targets they went slower and tighter.

In practice, this means “aim practice” isn’t only about “moving faster.” Two other things matter too. One is landing more consistently. The other is knowing when to commit to a shot and when to keep correcting.

System delay (latency) hurts aim below 20 milliseconds — but not the same for everyone

A 2021 study (Spjut, Boudaoud & Kim) compared two latency settings: 12ms and 20ms of click-to-photon delay. That’s the time between your click and seeing it register on screen. The test used 8 people. Each did 400 attempts per latency setting, so 800 in total.

  • Cutting that 8ms of delay led to a 182ms improvement in how fast people completed the task. That’s a much bigger effect than the delay itself. One likely reason is that small delays add up with each correction you make. The paper doesn’t prove that part, though.
  • But the effect wasn’t the same for everyone. Only 3 of the 8 people improved by a statistically significant amount.

So reducing system latency is worth doing. The payoff can be outsized. But how much it actually helps you specifically is something only your own testing can answer.

Note this is about local system delay — mouse, then PC, then monitor. That’s different from network ping.

Aim trainers measure consistently, but don’t yet prove they help your real game

A 2024 study (Rogers et al.) tested how consistent KovaaK’s is as a measurement tool. It used 10 FPS players across four tasks: micro flicking, macro flicking, strafe tracking and wall peeking. Each player was tested twice, 3–5 days apart. The consistency was excellent.

That’s a genuinely useful result. But it only proves KovaaK’s is a reliable ruler. It doesn’t prove that training with it makes you better at VALORANT. The researchers themselves point this out. Aim trainers strip out things that matter in real games: pressure, teammates, and unpredictability.

So claiming aim trainers definitely transfer to real games overstates the evidence. Claiming they’re useless also overstates it. The honest answer sits in between. They’re unproven for transfer. They’re genuinely useful as a way to measure your own aim consistency over time.

Movement penalties on accuracy are specific to each game

Stopping before shooting — counter-strafing, hard stops, and similar techniques — exists for one reason. Most games penalize your accuracy while moving. Exactly how big that penalty is, and how it works, differs per game. It can also change with patches. It’s not a universal rule.

How this compares with the experience above

Claim from treactisVerdictWhy
0.05 (≈163 cm/360°) was too slowAgrees in partThe study’s slowest setting was 160 cm/360°, almost the same as the old one. It fell outside the best range, because players couldn’t turn fast enough. But the study only tested small turns. The main reasons for switching (opposite-angle checks, turning from flashes) are big turns, which weren’t tested.
Lower sensitivity is better for long rangeAgreesThe NVIDIA study found harder shots (smaller, further targets) worked better with lower sensitivity.
Lower sensitivity counters micro-strafesNot testedThe study only used targets that stood still. Nothing here measures moving or strafing targets.
Doubled sensitivity, used to it after 1–2 DeathmatchesAgreesThe study found performance barely changed even with sensitivity changing. This is a real-world match for that. It’s still a personal report, not a measurement.
Less hand fatigue, slower skate wearNot coveredNone of the studies here measure these.
KovaaK’s made aim more consistent, not strongerFits, but doesn’t prove transferDonovan et al. found that landing more consistently is one clear sign of better aim. But the whole routine changed at once, and consistency was judged by feel.
Pre-placing the crosshair on the ball’s pathAgrees in directionThis follows Fitts’s Law: less distance to cover means a faster shot. It’s the same logic as crosshair placement. Mixing approaches by scenario matches Donovan’s finding that players change strategy per task. Whether it carries over to VALORANT crosshair placement is untested.
Train weak areas separatelyAgrees in directionNo FPS study has tested this directly.
Warm-up routine, dead-zoning, DeathmatchUntestedThese are coaching common sense, not measured findings. Dead-zoning details depend on VALORANT’s movement rules.

How strong is this evidence

Solid, well-established (general motor science): the speed-accuracy tradeoff. The fact that skilled aim means less scatter around the target. The idea of small in-flight corrections. This comes from decades of movement-science research outside of gaming.

Single small study, FPS-specific: the 20–80 cm/360° sensitivity range. The sub-20ms latency effect. The “better players show better control” finding. KovaaK’s reliability. Each of these comes from one study with 8–32 people. This is the best evidence available. Treat the exact numbers as a useful direction, not as precise facts.

Coaching common sense, not measured: crosshair placement. Pre-aiming. Warmup routines. How to split practice between tracking and flicking.

Folklore not backed by evidence (some contradicted): that changing sensitivity ruins your muscle memory (the NVIDIA study found the opposite). That copying a pro’s exact sensitivity is optimal for you. That your aim-trainer rank reliably predicts your in-game aim.

Overall, the evidence here is mixed. Squeezing all of this into one confident label would misrepresent it.

  • gamesense — crosshair placement is where aim and game sense overlap. Knowing where to aim ahead of time is a decision. Actually landing the shot is a motor skill.
  • system — Windows power and USB settings that affect local system delay, the kind of latency covered above.

Sources

Graph View

Backlinks