Crosshair ScaleCROSSHAIR SCALE · CALIBRATION TOOL · PUBLIC V2.5
Training data stays on this device · Switch device
// Universal FPS mouse calibration
Find the sensitivity
that fits you
Profile your reaction, clicking, tracking and flick control, then use a blind three-setting comparison to decide whether a change is actually supported.
Steady!
Got you
STEP 00
Translate every FPS into one physical unit
Sensitivity numbers are not comparable across games. Use a reproducible coefficient where available; otherwise use your measured physical turn distance.
Current physical turn distance16.3cm / 360°Calculated from a common public coefficient · eDPI is comparable only within one game
Important: these inputs do not change the web crosshair speedDPI and in-game sensitivity are used only for cm/360°, conversion and suggested values. A browser cannot recreate an engine's exact feel; test 05 is what evaluates a change.
Convert to target game
Keep 16.3 cm/360° →Equivalent target sensitivity3.182Uses a common reproducible hip-fire coefficient; this is not an official full-feel recreation.
This game can be convertedThis game can be converted with a common reproducible mouse-turn coefficient; it is not a developer-guaranteed full-feel recreation.
Game selection changes the calculationFormula mode uses
cm/360° = 914.4 ÷ (DPI × sensitivity × yaw) while measured mode avoids opaque or inconsistent game coefficients. Test 05 simulates only current, −5% and +5%, using your landing error and time to estimate a direction.This tool estimates only a first adjustment direction for hip-fire/base view; it does not claim a universal best sensitivity. ADS, scopes, FOV, acceleration curves, equipment penalties and vertical multipliers require separate in-game checks. The aim tasks observe both speed and accuracy based on the classic Fitts speed–accuracy relationship; verify the result in at least three matches.
FIELD TEST
Four skill profiles, one comparison
The first four tests describe ability; only the fifth informs sensitivity. Use your normal setup with browser zoom at 100%.
Target visibilityDual outlines are enabled by default. Use high contrast for colour-vision differences, dim displays or small screens.
Confirm your test setupKeep the same device and settings for comparable results.
- Fine pointer detected
- Pointer lock supported
- Raw mouse input will be requested for flick tests
Reaction
LIVEReady
Reaction
Click when the field turns green; early clicks do not count.
During the test, LIVE in the top-right shows time, hits and accuracy.YOUR REPORT
Your calibration report
Your first training data will appear here.
Basic skill profileNot tested—Does not decide sensitivity
Start hereComplete test 05 for a sensitivity verdict
The first four tests profile basic skills; they do not decide sensitivity.
How to read these scores
Reaction：Lower milliseconds means a faster response.
Target clicking：More hits and higher accuracy are both better.
Smooth tracking：A higher percentage means steadier tracking.
Flick control：A higher score means landings are closer to target centre.
Coach recommendation
Complete test 05 before a sensitivity recommendation is generated.
This is only the next setting worth trying. Change one variable, verify for at least three matches in the same game, and revert if it does not help. Equal cm/360° guarantees only equal physical turn distance.
VALORANT · current → suggested1.00016.3 → 16.3 cm/360°
Convert to Counter-Strike 23.182
With the same DPI, keep about 16.3 cm/360°. Hip-fire / base view only.
METHOD & EVIDENCE
How is this recommendation calculated?
This is not a sensitivity guess from one score. Physical conversion, skill profiling and sensitivity comparison are separated, with every decision rule disclosed.
Use cm / 360°
Use a reproducible coefficient where available; otherwise measure one full turn. Raw game numbers are not directly comparable across games.
Compare only against yourself
Current, −5% and +5% are hidden as A/B/C. Each setting gets six identical target positions.
Accuracy, pace and stability
Accuracy uses mean landing error, pace uses median acquisition time, and stability uses median absolute deviation (MAD).
No clear winner, no change
An alternative must lead by at least five points, avoid meaningful accuracy or stability loss, and win a majority of matched-target pairs.
Limits
The first four tests only profile ability. Reaction time includes display, browser and input latency. This tool suggests only the next hip-fire/base-view setting to try; it does not determine a best DPI or replace ADS/FOV checks and at least three in-game matches. PC and mobile scores are not directly comparable. The disclosed weights are conservative project rules, not an industry standard.
FAQ · READ THIS FIRST
Common questions
Find out why a change was not recommended, why totals can match, and what the input values actually do.
01Why was no change recommended?
No alternative passed all four checks at once: a five-point total lead, no meaningful accuracy loss, no meaningful stability loss, and enough support from same-target comparisons. Keeping the current setting is a valid result.
02Why can different input sensitivities produce the same verdict?
DPI and in-game sensitivity are used to calculate cm/360° and the suggested game value; they do not directly change the web crosshair speed. Test 05 compares the browser baseline with −5% and +5%, so different inputs can still produce the same keep-current verdict.
03What are the DPI and game-sensitivity inputs for?
They calculate physical turn distance in cm/360°, cross-game conversion and the final suggested value. They are not web mouse-gain controls.
04What do the total and three subscores mean?
The total is weighted from three subscores: accuracy 50%, pace 30% and stability 20%. They represent closeness to target centre, acquisition time and repeatability of landings. All run from 0 to 100; higher is better.
05Why can equal totals have different subscores?
The displayed total is rounded, so 71.7 and 72.0 both appear as 72; the decision still uses unrounded values. A gain in one subscore can also be offset by a loss in another.
06Is data quality a skill rating?
No. It only checks whether samples are complete and valid and whether the run varied too much. It decides whether a recommendation is safe to make.
07How are better / similar / worse calculated?
All three settings use the same six target positions. Each landing receives a paired score of 62.5% accuracy + 37.5% pace. More than one point above current is ‘better’, more than one point below is ‘worse’, and the rest is ‘similar’.
08Why not give one “best sensitivity”?
A browser cannot fully read a game's FOV, acceleration curve, ADS, scopes, equipment effects or device latency, and there is no single value that suits everyone. This tool gives only the next relative direction worth trying.
09When should I retest?
Retest after changing device, posture or browser; after an interrupted or unusable run; or after changing a setting and verifying it for at least three matches. A close but complete result does not require immediate repetition.
10Can desktop and mobile scores be compared?
No. Mouse and touch use different input models, trial counts and device latency. Compare only your own records from the same route, device and similar conditions.
11Are my results uploaded?
No. Tests, history and share cards are processed in this browser. Clearing browser data or changing devices does not sync them.
ARY // AYAN · HIDDEN TRANSMISSIONAyan, keep your crosshair steady.The original calibrator, Ary, has been found.