How to pick a drill

Feeling slow in a match can mean several different things. It can be the time between seeing a target and pressing, it can be a failure to read where a moving target will be, or it can be a repeated input drifting off the beat. Practising all three at once tells you nothing about which one improved.

So this tool splits the actions apart and measures them one at a time. Inside a game you also get the map, the weapon and the opponent mixed in. Here only one thing changes per run. The drills are generic input actions, they read nothing from any game client, and this tool is not affiliated with any game or studio.

Click aim

A circle appears at a random spot. We time the gap between the frame it is drawn and the frame you click, and a click outside the circle counts as a miss for that round. After ten rounds you get the fastest value, the median and the mean. The mean alone is easy to wreck with one slow round, so the median sits next to it.

Target leading

A target crosses the board at a steady speed. You have to hit where it will be when the shot lands, so you click ahead of what you see. The app takes the current position, adds velocity multiplied by travel time, and reports the distance between your click and that aim point as a share of board width.

Rhythm input

A steady beat runs. Press space once per beat, and we measure the distance from your press to the nearest beat in milliseconds. Early and late presses cancel each other out in a plain average, so the mean is taken over absolute values. The drill works with sound off, since the tile lights up on every beat.

Consecutive jump timing

A short window opens at a fixed interval. Press space while it is open and the chain continues, press outside it and the chain resets. At the end you get the longest chain and the mean distance from the center of the window. This is the generic version of pressing repeatedly inside a timing window, not an imitation of any specific movement technique.

Where the numbers come from

Reaction time is the gap between the moment the target is painted and the moment you press, read from the browser performance clock so a system clock change cannot skew it. Display refresh and input device latency are still baked in, so treat these numbers as your own trend on your own machine rather than something to compare across hardware.

What is stored

When a drill ends, one headline number and a date go into browser storage. Thirty runs per drill are kept and the oldest is dropped after that. The server only learns that a drill started and that a drill finished. It does not learn which drill, and it does not learn your numbers.

Clearing site data clears the history too, and nothing syncs to another device. The clear button below the chart wipes it immediately.

What this app does not do

There are no benchmarks. We have never measured what a good millisecond figure is, and we do not state things we have not measured. Ranks exist only among aim runs with Compete on, built from records players actually set. We do not count attendance either, and nothing is said about the days you skip.

Why there are no game names

Drill names and screen text carry no game titles. A game name appears only where it is needed to identify whose setting is being discussed, and then only as description. No art, audio or maps are used, no config file is read, and no input is sent into any game.

You do not need to run all four every time. Pick the one that matches what actually went wrong in your last match. Late on a stationary target points at click aim, shots landing behind a moving target point at leading, and broken movement points at jump timing.

Running the same drill at the same time of day across several days is what makes a trend readable. Fatigue moves these numbers a lot, so comparing two runs from the same hour is more confusing than useful. One run, recorded and left alone, works better.

What you are matching

Moving between games, the thing to match is not the number in the settings menu but the physical distance your hand already knows. That distance is the travel needed for a full 360 degree turn, written cm/360. When two games share that value, the same hand movement turns you the same amount.

The math

Divide 360 by DPI multiplied by sensitivity multiplied by yaw, then multiply by 2.54. DPI is the count your mouse reports per inch of travel, yaw is the angle one count turns at sensitivity 1, and 2.54 converts inches to centimeters. The target sensitivity is whatever makes DPI times sensitivity times yaw equal on both sides.

Why we ship no yaw table

Per game yaw values are only shipped here when a studio publishes them. On 16 September 2026 we opened five sources and none published a value, so the built in table is empty. The table on screen lists what was opened and what came back. If you do not know your yaw, set both sides to 1 and compare DPI and sensitivity alone.

Numbers swinging wildly usually means too few rounds. Ten is small enough that one mistake moves the mean a lot, so read the median first and treat the mean as context.

If the leading gap will not come down, stop tracking the target with your eyes and instead pick the spot it is heading for and wait there. How far ahead to aim is speed multiplied by travel time, so faster targets sit further out.

If your rhythm offsets all lean the same way, that is bias rather than jitter. Leaning early means pressing slightly later, leaning late means pressing slightly earlier, and the number drops.

If space does nothing, click the board once and try again. Key input only reaches the board while it has focus.