The training data page has four sections: overview, per-mode progress (grouped by mode, difficulty, weapon and guide settings, duration and rule version, with score trends), last 30 runs, and data management. Overview shows cumulative practice totals, your latest session score and shooting accuracy. Shooting accuracy is total hits divided by total shots, excluding Tracking samples. To assess changes, compare progress groups and recent records with matching practice conditions.
What Each Metric Actually Measures
| Metric | What it actually means | Can you read it alone |
|---|---|---|
| Score | A total weighted by hit location, combo and each mode's own rules | Compare completed sessions with matching mode, difficulty, duration, rule version and applicable weapon/guide settings |
| Accuracy | Hits ÷ shots fired | No — bigger targets give higher accuracy |
| KPS | Hits per second = hits ÷ actual practice seconds | No — alone it just encourages you to spray shots |
| Combo | Consecutive hits, reset to zero on a miss or a missed target | Fine as a consistency reference, but it does not reflect intensity |
| Reaction time | The gap between the target appearing and your shot | No — never off your single fastest run |
Accuracy Has to Be Read Together With KPS
Those two sit on a seesaw: slow down and accuracy rises while KPS falls; speed up and it flips. So read them together:
- High accuracy, low KPS — you are playing it safe; there is room to speed up
- Low accuracy, high KPS — you are flailing; go back to deceleration work
- Both climbing — that is real progress
- One up, one down — you just changed your approach; your level did not move
One more thing: easy difficulty has bigger, slower targets, so accuracy is naturally higher, which means any comparison has to fix the mode, the difficulty and the duration.
Read the Trend, Not a Single Run
A single run has far too many sources of variance: sleep, how clean your mousepad is, whether the previous run went well. Very little of one run's score can be treated as "your level". The trend line plots your score over the last 14 runs; read the direction:
- Trending up — it is working, keep going
- Flat — you are repeating things you already know; raise the difficulty or change your main
- Trending down — first check whether it is fatigue, lost sleep, or a settings change you just made
- Sawtooth — your performance is unstable; what you need to train is consistency, not your ceiling
Runs you end early do not enter the trend, because the duration differs and there is nothing to compare. Do not get into the habit of quitting halfway — that run's data is wasted.
Reaction Drills: Read Average Time and Consistency Together
This site’s reaction drill measures time from the target appearing to a successful hit. That includes spotting the target, moving the mouse and firing, so it is not a measurement of pure neural reflex speed. Your fastest result is a useful record, but it does not describe your consistency on its own.
The results screen shows average target-to-hit time, not a median. The reaction drill also shows the standard deviation of successful hit times. With comparable modes, difficulty, settings and sample counts, a smaller standard deviation usually means more consistent timing. Read average speed and consistency together; two figures cannot determine whose overall aim is better.
Look at One Metric Only After Each Session
Closing the page right after practice wastes it. Spending two minutes on this beats playing three more runs:
- Open your training data and look at today's rows in the last 30 runs.
- Find the worst item today, not your total score.
- The same item sitting at the bottom for three days straight is a structural weak point — make it tomorrow's main.
- Set exactly one goal for tomorrow, specific enough to be a number, like "micro-adjust accuracy from 62% to 68%".
The Two Most Common Mistakes
- Treating a single high score as progress — best scores get reset, the average line does not lie. Read the slope of the trend, not the highest point on the line.
- Using data to make excuses for yourself — "my accuracy is low today because I changed difficulty". Record the conditions alongside the score and only compare under identical ones.
Open your training data and look at the trend line first, not your best score.
Open your training data