Make the Compare graphic show run-to-run spread, not just the median#4879
Open
soulgalore wants to merge 1 commit into
Open
Make the Compare graphic show run-to-run spread, not just the median#4879soulgalore wants to merge 1 commit into
soulgalore wants to merge 1 commit into
Conversation
The per-metric A/B graphic drew two rails of dots with a bold median-to-median delta beside them. That over-weighted the median: a 270 to 278 ms shift read as a regression even when the runs overlapped so heavily the statistical test called it noise, and the effect size (the overlap-aware measure of how big the change really is) was hidden on every non-significant row. Now each series draws its middle-80% band (p10 to p90) behind the dots, and a grey stripe marks where the two bands overlap: a wide grey band means the shift sits inside run-to-run noise, a clean gap means it is real. Cliff's delta rides on every chip, significant or not, so the honest "how big" is always visible. The median percentage delta is shown only when the change is significant, and the median line is dimmed otherwise, so a couple of noise-milliseconds stop competing with the verdict. The graphic stays pure server-rendered HTML dots (no SVG, no client JS) so it keeps working on file://, and the row anchors are unchanged so existing deep links keep landing. Co-Authored-By: Claude Opus 4.8 (1M context) noreply@anthropic.com
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
The per-metric A/B graphic drew two rails of dots with a bold median-to-median delta beside them. That over-weighted the median: a 270 to 278 ms shift read as a regression even when the runs overlapped so heavily the statistical test called it noise, and the effect size (the overlap-aware measure of how big the change really is) was hidden on every non-significant row.
Now each series draws its middle-80% band (p10 to p90) behind the dots, and a grey stripe marks where the two bands overlap: a wide grey band means the shift sits inside run-to-run noise, a clean gap means it is real. Cliff's delta rides on every chip, significant or not, so the honest "how big" is always visible. The median percentage delta is shown only when the change is significant, and the median line is dimmed otherwise, so a couple of noise-milliseconds stop competing with the verdict.
The graphic stays pure server-rendered HTML dots (no SVG, no client JS) so it keeps working on file://, and the row anchors are unchanged so existing deep links keep landing.
Co-Authored-By: Claude Opus 4.8 (1M context) noreply@anthropic.com