MetricUpdated: 2026-08-28
What is sample weight?
Sample weight is the relative influence a historical record has in up/down backtest statistics.
Plain meaning
Backtests adjust influence when the intraday path is unclear. Close-only data, or a late bounce that only turns positive at the close, is usually not treated as a full-trend sample. This is not the same as time-weighted relative error on the allotment page.
Why it matters
Weighting can stop a few misleading paths from dominating hit rate and average return. Read deviation types together with the weighted result.
How to read it
Read it with sample action, deviation type, sample period and layer size. Weighting is a comparability signal, not a guarantee of correctness.
How it appears here
Backtest rows can mark trend traps, abnormal moves or sample actions, and the statistics reflect those weights.
Common mistakes
- Looking only at weighted results without sample notes.
- Treating this weight as the same number as time-weighted allotment error.
- Assuming all IPOs should always be equally comparable.