A useful comparison starts with the conditions
Choose two notes saved in the five-shot workspace. The comparison checks shooting path, environment, shot context, jumper/version labels and recorded Human controls. If those conditions differ, the page calls out the differences rather than presenting the sessions as an isolated timing experiment.
It is normal for the timing field itself to differ: that may be the one variable you deliberately tested. A change from a standing Button test to a moving shot is different. It changes which timer may be selected and therefore changes the question being tested.
Read the counts at their actual scale
Five attempts are a small practice sample. The table displays counts of the release and tempo observations you supplied. It does not claim a green rate, game-win rate, latency measurement or verified accuracy result.
A session with more Excellent release entries can be worth repeating under the same conditions. It does not establish that a build is universally better. Notes with missing version or jumper labels remain explicitly incomplete as a record of the conditions.
Repeat the question that the notes can answer
Keep the better-documented baseline as a starting record, then repeat the same kind of attempt. If the original groups mixed fades and standing shots, collect a clean group before moving another number.
Session storage belongs to this local browser origin. To retain a note outside it, export the JSON from the recording workspace. Keep observations and conclusions together so a later timing change can be traced back to what was actually tested.