Measure Error Bars Instead of Picking Best Trials
Whenafter measuring any repeated test or trial
Record every test run — including awkward outliers — and estimate how big your instrument or timing slop actually is. Never polish away messy variance just to flatter a pet hypothesis.
Why it works
At the Royal Mint and in Opticks, Newton pushed practical precision by quantifying angle margins and coin weight tolerances across repeats rather than single lucky runs.
Start tiny
Write down the worst measurement beside your best one before averaging.
What you’ll notice
Saves you from mistaking random noise for a breakthrough.
When it slips
Missed three noisy trials earlier? Start logging raw runs cleanly from the next test onward.
Isaac Newton, Opticks (1730), Book I Part II angle measurements; Royal Mint assay records (1696–1727).
Growth: Supporting habits that improve the conditions for useful repetitions: preparation, consistency or necessary recovery.