Guide

Game mouse sensitivity, ADS tuning and aim assist systems explained

Harbor Crest’s closed beta used one sensitivity slider for hip-fire and aim-down-sights. Scoped rifle fights showed 41% overflick rate on 30 m targets — players trained at hip-fire cm/360 could not translate muscle memory when zoom doubled effective magnification. Hip-fire duels felt sluggish at the only sensitivity that made ADS usable. Controller players quit ranked after three sessions citing “sticky” aim that fought them on micro-adjustments. The weapons were balanced; the input layer was not.

Mouse sensitivity, ADS multipliers, FOV scaling, and aim assist are the systems that convert physical stick or mouse motion into consistent reticle travel across zoom levels and platforms. They sit upstream of recoil and spread and downstream of input lag — if sensitivity changes when you enter ADS, players cannot separate aim error from weapon error. This guide covers cm/360 and eDPI math, hip vs ADS curves, FOV compensation modes, controller deadzones and assist zones, persistence and profile design, the Harbor Crest refactor, a technique decision table, pitfalls, and a production checklist.

Mouse sensitivity fundamentals: cm/360 and eDPI

Players compare settings across games using centimeters per 360° turn (cm/360) or effective DPI (eDPI = mouse DPI × in-game sensitivity). Both express the same physical relationship: how far you move the mouse to rotate one full circle in-game.

eDPI = Mouse DPI × In-Game Sensitivity

Lower eDPI means more mouse travel per degree — finer aim, slower turns. Competitive PC shooters cluster between roughly 20–80 cm/360 depending on role and desk space; battle royale players often run higher sens for looting rotations. Document your game’s sensitivity formula (linear, power curve, yaw-only scaling) in the settings tooltip so converter tools work.

Raw input and acceleration

Ship raw mouse input on PC by default — Windows pointer acceleration corrupts muscle memory when players alt-tab. Expose an optional acceleration curve only in accessibility presets, never silently. Poll mouse at the same rate as gameplay simulation or document the decimation rule; inconsistent poll timing shows up as micro-stutter in hit registration replays as “I was on him.”

Hip-fire vs ADS sensitivity

When magnification increases, the same mouse delta moves the reticle farther across the screen unless you compensate. Three common patterns:

Unified sensitivity — one slider; simplest UX, worst for mixed-range combat
ADS multiplier — ADS sens = hip sens × multiplier (typical range 0.5–1.0)
Independent ADS slider — separate cm/360 for scoped mode; clearest for players who convert from other titles

The industry default for tactical shooters is a 0.7–0.85 ADS multiplier relative to hip, tuned per optic zoom tier. A 2× red dot and a 4× ACOG should not share one multiplier unless your optic system normalizes zoom internally.

Per-zoom and per-weapon profiles

If weapons change FOV or optic attachment at runtime, sensitivity must key off effective zoom, not weapon ID. Cache the active zoom factor in the ADS FSM enter event and apply the matching row from a ZoomSensitivityTable. Optional per-weapon overrides belong only in simulation builds — ranked clients should use one table to prevent pay-to-tune advantages.

FOV and sensitivity scaling

Horizontal FOV changes how many pixels represent a degree of rotation. If sensitivity stays fixed when FOV changes, cm/360 drifts and players retune after every settings tweak.

Common scaling modes:

Mode	Behavior	Player expectation
Fixed (0% scaling)	Sens unchanged when FOV changes	Old-school PC; cm/360 shifts with FOV
Horizontal match	Scale sens so horizontal cm/360 stays constant	Most PC converter tools assume this
Monitor distance match	Match perceived speed at a reference screen depth	Scoped players migrating from other shooters

Expose the mode in advanced settings with a one-line explanation. Default to horizontal match for PC tactical shooters; console titles often use fixed because TV viewing distance varies. When ADS FOV narrows independently of hip FOV, apply scaling on the combined effective FOV, not hip alone.

Controller deadzones and response curves

Analog sticks report noisy values near center. A deadzone zeroes input below a radial threshold; too large and players cannot make pixel-level corrections, too small and crosshair drifts.

Axial deadzone — separate X/Y cutoffs; cheap but square-shaped dead region
Radial deadzone — circular cutoff; standard for shooters
Outer deadzone — caps stick travel before max turn rate; softens max spin

Map stick deflection through a response curve (linear, exponential, dual-zone) before multiplying by turn rate. Competitive defaults: 5–10% inner radial deadzone, slight exponential on outer 30% for fine aim. Let players adjust inner/outer independently in advanced settings.

Look sensitivity tiers

Split horizontal and vertical sensitivity only when genre demands (vehicles, flight). For ground shooters, keep Y = X unless accessibility requests otherwise — asymmetric defaults confuse muscle memory. Offer separate ADS look multiplier mirroring the mouse ADS table.

Aim assist architecture

Console and hybrid PC titles use aim assist to close the gap between stick precision and mouse fidelity. Treat assist as layers, not one magic magnet:

Slowdown (friction) — reduces turn rate when reticle nears target hitbox
Rotational assist — adds yaw toward target velocity vector while firing
Magnetism (snap) — pulls reticle toward target center on acquisition; use sparingly
Bullet magnetism — server-side cone widening near targets; never visible in UI

Define assist zones by distance bands: strong inside 15 m, fade 15–40 m, off beyond engagement range where spread dominates. Disable or attenuate assist while ADS for precision weapons if hip assist is generous — players feel the fight between systems.

Cross-play fairness

PC mouse lobbies should disable rotational assist entirely; mixed lobbies need disclosed rules and input icons on scoreboard. Never let assist override explicit counter-aim input — cap assist torque per frame so skilled players can pull off target. Log assist contribution in debug builds to tune complaints objectively.

Settings persistence and UX

Sensitivity settings are identity. Store them in cloud saves keyed to account ID, not local machine only. On first launch:

Offer hip + ADS sliders with live range preview (spinning target dummy)
Show cm/360 readout updating as sliders move — players convert faster than abstract numbers
Include “film sensitivity” presets (low / medium / high) that set starting eDPI bands
Warn before applying pro-player configs that assume specific DPI

Reset-to-default must restore your tuned tables, not Unity editor defaults. Version settings blobs when curves change so old saves migrate with a one-time recalibration prompt.

Harbor Crest refactor walkthrough

Telemetry tagged overflicks when scoped crosshair overshot and corrected back within 120 ms — classic ADS sensitivity mismatch, not recoil fault.

Pass 1: Split hip and ADS

Hip default: 25 cm/360 equivalent at 800 DPI (in-game sens 0.42)
ADS multiplier slider 0.5–1.0, default 0.78 per 1.5× optic
Per-zoom rows: red dot 0.85, 2× holo 0.78, 4× ACOG 0.65
FOV scaling: horizontal match on hip and ADS independently

Pass 2: Controller curves and assist zones

Radial deadzone 0.08 default; advanced inner/outer sliders
Rotational assist fade curve 100% < 12 m, 40% at 25 m, 0% beyond 35 m
ADS assist multiplier 0.6× hip values for marksman class
Training range added with moving bots and live cm/360 display

Scoped overflick rate fell 41% to 14%. Controller ranked retention week-2 improved 28% to 44% without changing time-to-kill. Support tickets mentioning “aim feels wrong” dropped 62% to 18%.

Technique decision table

Approach	Best for	Trade-off
Single unified sensitivity	Arcade shooters, casual mobile ports	Scoped play always feels wrong or hip-fire too fast
ADS multiplier only	PC tactical with 1–2 optic tiers	Insufficient when zoom range is wide
Per-zoom sensitivity table	Multi-optic tactical and sniper hybrids	More QA matrix; must sync with attachment system
FOV-linked scaling	Games with adjustable FOV slider	Players must understand scaling mode to convert settings
Layered aim assist	Console and cross-play ground shooters	PC purists may reject; needs transparent mixed-lobby rules
Training-range live cm/360	Any competitive title	Small engineering cost; large reduction in settings churn

Common pitfalls

One slider for hip and ADS — Harbor Crest’s 41% overflick baseline.
Ignoring zoom in sensitivity math — attachments change FOV without updating sens table.
Windows acceleration on by default — breaks cross-session consistency on PC.
Assist fighting recoil — players cannot diagnose whether stick or spread caused the miss.
Max assist in mixed PC lobbies — drives input-based matchmaking exploits.
No cm/360 display — players YouTube-copy settings at wrong DPI.
Sensitivity changes mid-season without migration — invalidates muscle memory and pro configs.
Vertical sens reduced silently — feels like input lag on vertical tracking.
Per-weapon ranked overrides — pay-to-tune perception even if unintentional.

Production checklist

Document sensitivity formula (linear, yaw scale, FOV scaling mode).
Ship separate hip and ADS controls with sensible defaults per platform.
Build per-zoom sensitivity rows tied to optic magnification, not weapon ID.
Enable raw mouse input on PC; hide acceleration behind accessibility.
Apply FOV scaling when hip or ADS FOV is player-adjustable.
Tune controller deadzone defaults with hardware variance tests.
Layer aim assist with distance fade; cap torque per frame.
Disable or attenuate assist in PC-only lobbies per policy.
Persist settings to cloud; version blobs when curves change.
Add training range with live cm/360 and overflick telemetry.
Validate ADS enter/exit does not spike sensitivity for one frame.
Regression-test with input lag budget after any sens pipeline change.

Key takeaways

cm/360 and eDPI are the portable language — show them in UI, not just abstract sliders.
ADS needs its own curve — magnification changes effective sensitivity unless you compensate.
FOV and sensitivity are coupled — pick a scaling mode and document it.
Aim assist is layered and distance-banded — not a single sticky toggle.
Harbor Crest cut scoped overflicks 41% to 14% with split curves and per-zoom tables, not weapon nerfs.