Practice analytics
See the difference between a fast single and repeatable improvement. SupaKewber combines rolling averages, session trends and smart-cube CFOP splits in one solve-review workflow.
Direct answer
Start with a group of comparable solves, not one personal best. Compare current and best averages, consistency and the shape of the session graph. With a connected smart cube, add stage percentages for Cross, F2L, OLL and PLL, then choose the stage that is repeatedly slow rather than reacting to one scramble.
Complete guide
A fastest single shows what was possible on one scramble. Ao5 responds quickly to form, while Ao12 and longer windows are more useful for deciding whether a change has become stable.
Standard deviation and the spread of the graph reveal consistency. Two sessions can have the same mean while one is tightly grouped and the other alternates between very fast and very slow attempts.
First inspect the overall trend, then open individual solves that sit far above or below it. Check the scramble, penalty, move count and neighboring results before deciding that one solve represents a skill problem.
Imported csTimer sessions can join locally recorded practice in the Solve Lab, while Manual and Smart Cube sources remain labeled so unlike timing methods are not treated as identical.
Smart-cube timestamps are converted into separate stage durations. Looking at the average share of total time helps reveal whether Cross planning, F2L lookahead or last-layer recognition is the more persistent limit.
About half of total execution is a practical F2L reference for many CFOP solvers, but the useful comparison is your own trend across multiple sessions rather than a universal percentage.
Choose one targeted change: plan one more Cross move, reduce F2L rotations, drill a weak OLL recognition angle or execute a PLL from a fixed grip. Measure the next session under similar conditions.
Training becomes easier to evaluate when the intervention is narrow. If several habits change at once, improved or worse results cannot be traced to a specific cause.
Quick answers
Ao5 is responsive, Ao12 is steadier and Ao50 or Ao100 is useful for longer-term consistency when enough solves exist.
No. Automatic stage detection requires the move stream from a supported connected smart cube.
Roughly 45–55% is a practical diagnostic range for many CFOP solvers, not a fixed rule.
Yes. Solve review includes result context and, when available, smart-cube stage data.
Practice with feedback
SupaKewber connects timing, smart-cube telemetry, training and online competition in one browser workflow.
Review solve statistics