how do i know if a pine script indicator repaints before i trust it

[endlessroad]

downloaded a free pine script that looks incredible on the chart, calls tops and bottoms almost perfectly. which is exactly why im suspicious. how do i actually verify whether it repaints or recalculates on historical bars before i build anything around it?

[Dean]

fastest practical test: add it to a chart, then switch the chart to bar replay and step forward one candle at a time. if the signals stay put exactly where they were, its fine. if arrows jump, disappear, or move to a better spot as new bars form, it repaints. the replay tool exposes it in about two minutes.

[alexturner]

bar replay is the right tool. also check the code for security calls with lookahead enabled and signals plotted on the realtime bar instead of confirmed bars. anything using future-leaning data or plotting before a bar closes will look magical historically and fail live.

[cloudyvision]

if a free indicator nailed tops and bottoms it would not be free, it would be a hedge fund. the fact that you found it on the internet for nothing is the test result. you can skip the rest.

[Chris]

the troll is blunt but the heuristic is real. that said, plenty of honest non-repainting indicators are free too, they just look boring because they only fire on closed bars and miss the exact extremes. boring and reliable beats magical and fictional every time.

[Caleb]

where is the bar replay button, ive never used it. is it on the free plan?

[derekwhite]

its in the top toolbar, looks like a little play icon, available on free with some limits on how far back you can replay. pick a starting point a few hundred bars back, hit play paused, and step forward manually. watch the indicator as each candle closes. thats the whole verification.

[Adam]

bar replay caught two repainters for me already.

[parkerlee]

just tested my 'best' indicator with replay and watched the arrows literally rearrange themselves feel stupid for trading it for a month. thanks for this thread.

[carterw]

good outcome. one more habit: even for a confirmed non-repainter, forward test it on demo before risking money. historical accuracy and live profitability are different things, repainting is just the first filter, not the last.