Skip to main content
Scatter of merged PRs in the selected period, plotting cycle time (x, log scale) against lines changed (y, log scale). Each dot is one PR, colored by size bucket. Limited to 2,000 most recently merged PRs for performance. Click any dot to open the PR on GitHub.

​
Why it matters

PR size is the single biggest predictor of cycle time. This chart makes that relationship concrete for your team and surfaces outliers that don’t follow it.

​
What to look for

  • Top-right cluster (big + slow) — confirms the “ship smaller PRs” argument with data.
  • Far-right outliers (any size, multi-day cycle) — PRs that got stuck. Check whether in review, waiting on CI, or idle.
  • Bottom-right (small PR, long cycle) — process friction unrelated to size: bad CI, review drought, or forgotten PRs.
  • Top-left (big PR, fast cycle) — either a rubber-stamped review or well-prepared structural change.

​
See Also

PR Size Distribution