BMS BENCHMARK METHODOLOGY

WHAT ARE WE
MEASURING?

BeatMyScroll 2.0 is a six-axis browser mechanics benchmark aimed at the motor skills that show up repeatedly in mouse-heavy shooters. It is not a replacement for an in-game aim trainer, and browser/device latency still matters, but every test is designed to isolate a useful mechanical trait.

01

THE SIX SKILLS

Reaction — 20%

Five randomized visual-delay trials measure the time between the field becoming actionable and the primary pointer press. The benchmark uses the median, not the fastest trial, so one lucky response does not define the result.

Flick Acquisition — 25%

Twenty targets appear sequentially at random positions. The raw result contains average acquisition time and hit accuracy. The normalized score rewards speed, but accuracy scales the result down when a player throws misses at the problem. Flicking receives the largest rating weight because fast acquisition with control is central to many shooter engagements.

Tracking — 20%

A target follows a multi-frequency continuous path for fifteen seconds. The browser samples whether the pointer is inside the target over time. The raw result is the percentage of active test time spent on target. This is continuous control, not a click test.

Target Switching — 15%

Multiple threats remain visible while only one is active. After each successful hit, another target becomes active. Average transfer time and miss rate are combined, approximating the repeated visual-selection and aim-transfer pattern common in multi-target fights.

Micro Precision — 15%

Small targets score where the pointer lands relative to the center. Touching the edge is not treated like a perfect hit. Misses reduce the center-quality result, while average acquisition time contributes a smaller speed component.

Click Burst — 5%

Five seconds of primary-click cadence on a fixed pad. This test exists because raw clicking is still a motor skill, but it receives the lowest BMS Rating weight so spam speed cannot overpower aim mechanics.

02

FROM RAW RESULT TO 0–100.

Each event first produces a raw measurement: milliseconds, percentage, clicks per second, or a combination of timing and accuracy. That raw result is converted into a normalized 0–100 skill score using a fixed scoring function.

Fixed functions matter because the meaning of your score should not suddenly change just because a new player joined the site. Community percentiles are displayed separately from the normalized skill score.

Server-recomputed, not cheat-proof. When a synced result reaches Cloudflare, the backend recalculates the normalized score from the submitted raw measurements instead of trusting a browser-provided 0–100 score. A browser benchmark can still be manipulated by a determined client, so public rankings should be treated as competitive community records rather than a secure professional tournament anti-cheat system.
03

BMS RATING.

The weighted overall score uses:

  • 25% Flick Acquisition
  • 20% Reaction
  • 20% Tracking
  • 15% Target Switching
  • 15% Micro Precision
  • 5% Click Burst

The weighted 0–100 result is mapped onto a 0–2000 BMS Rating centered around 1000. A weighted overall score of 50 corresponds to roughly 1000 rating; every additional overall point is worth about 20 rating.

Class thresholds are Recruit, Bronze, Silver, Gold, Platinum, Diamond, Ascendant, Immortal and Demon. Until all six tests have been completed, the site displays a provisional rating based only on the completed axes.

04

BEST RESULTS, NOT AVERAGE GRIND.

Each skill stores your best normalized result. Repeating a test can improve that axis, but a weaker retry does not erase the personal best. The global BMS leaderboard only includes complete six-test profiles.

This design makes the profile a mechanical ceiling snapshot rather than an average of every warm-up attempt. A future ranked mode could use session averages or limited attempts, but the current benchmark is deliberately a best-performance profile.

05

HARDWARE MATTERS.

Display refresh rate, browser scheduling, mouse polling rate, operating system input handling, touch hardware and end-to-end latency can all affect browser measurements. Reaction time is especially sensitive to the complete input/display pipeline.

For the most comparable shooter-oriented profile, use a desktop or laptop with a mouse, keep browser zoom at a normal level, avoid background load, and run the full suite in one focused session. Touch is supported for several tests, but a touchscreen result should not be interpreted as directly equivalent to a mouse benchmark.

Losing tab focus aborts the active test rather than letting background throttling produce a misleading result.

06

WHERE DID SCROLLING GO?

The original scrolling game is still available as BeatMyScroll Classic. Classic measures scroll endurance, scroll speed, RUSH performance and survival. Those records remain separate from BMS Rating because scroll mechanics are not a meaningful component of shooter aim.

Keeping the systems separate protects both identities: Classic can stay ridiculous, while the mechanics benchmark can stay interpretable.

RUN THE FULL BMS BENCHMARK →