← Back to fpstrain

DPI, sensitivity, cm/360 explained: how pros pick their numbers

By Mustafa Bilgic, FPS gaming enthusiast (Adiyaman, Turkiye) — last reviewed 2026-05-06.

Three numbers determine how your mouse maps to the FPS camera: DPI, in-game sensitivity, and the per-game yaw constant. Together they produce a single comparable quantity — cm/360 — that pros, coaches, and aim trainers all use. This guide gives you the math, the per-game constants, the pro distributions, and the common pitfalls.

If you read nothing else. DPI alone is meaningless. The single useful number is cm/360: how many centimetres of mouse motion produce a full 360-degree in-game rotation. Hit a stable cm/360 across all your FPS games and you eliminate the largest source of cross-game aim regression.

1. DPI is the mouse's resolution, not its sensitivity

DPI (dots per inch) is how many position counts the mouse sensor produces per inch of physical movement. A 1600 DPI mouse moved 1 inch produces 1600 counts. That's it. The mouse itself does not know how those counts map to camera rotation — that mapping happens in the game.

Mouse sensors run at fixed internal sampling rates. Above ~3200 DPI, optical sensors of the 2024-2025 generation produce essentially identical accuracy to lower DPI; the "higher is better" claim from manufacturers is a marketing artefact. Pros overwhelmingly use 400, 800, or 1600 DPI. ProSettings.net (prosettings.net) public distributions confirm this for CS2, Valorant, Apex, and Overwatch 2.

Why 800 DPI is the default pro choice

2. In-game sensitivity is a multiplier on per-count rotation

Each FPS engine assigns a base "yaw per count" constant. The in-game slider multiplies that constant. The result is the number of degrees the camera rotates per mouse count.

For most Source-engine derivatives (CS:GO, CS2, Apex Legends), the yaw constant is approximately 0.022 degrees per count at sensitivity 1.0. So a 800 DPI mouse moved 1 inch at CS2 sensitivity 1.0 rotates the camera by 800 * 0.022 = 17.6 degrees per inch.

For Valorant, the formula uses a different scaling. At Valorant sensitivity 1.0, each mouse count rotates approximately 0.07 degrees if you pair it with the 800 DPI default — the published Riot tools and well-known sensitivity converter sites (Mouse-Sensitivity.com, Aiming.pro) implement this constant. Many pros use Valorant sensitivities between 0.3 and 0.5, paired with 800 DPI.

Engine / GameYaw constant approximationNotes
CS:GO / CS20.022 deg / count at sens 1.0Source/Source 2; verified by sensitivity converter sites and the public CS2 source on the m_yaw cvar
Apex Legends0.022 deg / count at sens 1.0Source-engine derivative
Valorant0.07 deg / count at sens 1.0Riot proprietary; widely used 0.07 ratio
Overwatch 2~0.0066 deg / count at 1.0 multiplierDifferent multiplier scale; conversion via percent-based UI
Fortnite~0.005555 deg / count at sens 1.0Note Fortnite uses two slider sub-values (X and Y)
Call of Duty (MW3 / Warzone)~0.0066 deg / count at sens 1.0Multiplier scale; ADS multiplier is separate

Constants verified against the Mouse-Sensitivity.com converter, the Aiming.pro converter, and engine source where public. We use these in our local sensitivity converter calculator.

3. The cm/360 formula

The number you actually want is centimetres of mouse motion to rotate the camera 360 degrees.

cm/360 = (360 / (DPI * yaw_constant * in_game_sensitivity)) * 2.54

The 2.54 converts inches to centimetres because DPI is dots per inch.

Example 1: CS2 pro typical

cm/360 = (360 / (800 * 0.022 * 1.0)) * 2.54 ~= 51.95 cm

So a CS2 pro on 800 DPI, sens 1.0 needs about 52 cm of mouse motion for a full 360. That is on the high (slow) side of pro distribution.

Example 2: Valorant pro typical

cm/360 = (360 / (800 * 0.07 * 0.4)) * 2.54 ~= 40.83 cm

About 41 cm/360 — close to the Valorant pro median.

Example 3: Apex Legends

cm/360 = (360 / (800 * 0.022 * 1.5)) * 2.54 ~= 34.63 cm

4. Pro distributions by game (ProSettings.net 2024)

The numbers below are aggregated from public ProSettings.net listings of Tier 1 competitive players in 2024. They are typical, not prescriptive. Genetics, hand size, desk size, and personal preference all matter.

Gamecm/360 25th percentilecm/360 mediancm/360 75th percentile
CS232 cm40 cm56 cm
Valorant33 cm41 cm50 cm
Apex Legends30 cm40 cm60 cm
Overwatch 2 (hitscan)22 cm30 cm40 cm
Fortnite28 cm40 cm50 cm
Call of Duty30 cm45 cm60 cm

Trends: tactical FPS (CS2, Valorant) cluster narrower; tracking-heavy games like Apex have a wider spread because some players prefer arm-aim (slow cm/360) and others prefer wrist-aim (fast).

5. How to pick your cm/360

  1. Decide arm vs wrist style. Arm: 40 to 60 cm/360. Mixed: 30 to 40. Wrist: 20 to 30.
  2. Pick a DPI from {400, 800, 1600}. 800 is the safe default.
  3. Solve for the in-game sensitivity that produces your target cm/360.
  4. Use the same cm/360 across all FPS games. Aim transfer collapses if you change sens between games.
  5. Adjust at most once per month and only with measurable evidence.

Use our sensitivity converter calculator to do the algebra automatically.

6. Common mistakes

Changing sensitivity weekly. The motor system needs a stable input mapping for consolidation. Constant changes erase prior gains.

Treating DPI as sensitivity. A higher DPI does not mean a more sensitive setup if you compensate with lower in-game sensitivity. They cancel.

Copying a pro's full setup. A pro's cm/360 might suit a different desk, hand, and pad. Copy the cm/360, not the slider value.

Using mouse acceleration. Acceleration breaks the constant cm/360 mapping and is generally avoided by pros (Windows pointer enhancement off, in-game acceleration off).

Different in-game ADS sensitivity multipliers. If your hipfire is 40 cm/360 but your scoped ADS uses a 1.5 multiplier inadvertently, your scoped flick has a different motor mapping. Consider using the "match in cm/360" ADS option (in CS2, the "Use the same multiplier" setting; in Valorant, scoped sensitivity multipliers).

7. Windows mouse settings

Two Windows settings matter:

Verify with a tool like MouseTester. The published goal is "raw input" — every count translates to one increment of game rotation, no acceleration, no scaling.

8. Polling rate, latency, and end-to-end input timing

Polling rate is how many times per second the mouse reports counts to the OS. Common values: 125, 500, 1000, 4000, 8000 Hz. Returns diminish past 1000 Hz for clicking tasks; 4000 to 8000 is perceptible mainly for very smooth tracking. CPU load increases noticeably at 8000 Hz.

End-to-end latency is the total delay from mouse motion to pixel update. The components are: mouse polling (~1 ms at 1000 Hz), OS input stack (~1 to 3 ms), game render queue (~1 to 16 ms depending on FPS), and display response (~1 to 8 ms depending on monitor). Total typically falls between 8 and 40 ms on tuned setups; 50 ms or more on stock setups.

For deeper latency analysis, see refresh rate research.

9. Hardware that matches the math

If you target 40 cm/360 on a Razer DeathAdder V2, the desk needs at least 40 cm of usable pad width to allow a full smooth 360 motion without lift-off. Most pros lift the mouse mid-flick and reset, so 30 to 35 cm is enough in practice — but a wide pad reduces forced lift-offs. Suggested gaming peripherals (Amazon Associates, no extra cost):

10. When to change cm/360 and when not to

Do change cm/360 if: your hand size or grip changed; you switched from wrist-only to arm aim; the cm/360 was demonstrably wrong (you cannot complete 180 in one swipe at all). Do not change cm/360 if: you are tilted from a single bad session; a streamer changed; or your benchmark dipped 5%. Sensitivity stability is one of the highest-leverage decisions in FPS aim.

11. Selected sources