Summary

We need an updated report for the 4.8 KR metric measuring how quickly bad-faith edits made by temporary accounts are reverted by patrollers.

Metric

Reduce by 40% the time it takes from when a bad-faith edit is made by a temporary account to the revert of that edit by a patroller, over a 7-day window, for 75% of those reverts.

Metric definition
Bad-faith edits are defined as:

• Edits made by temporary accounts.
• Reverted by a non-bot user with extended rights.
• “Extended rights” means at least the right to reveal temporary account IP addresses. These users are assumed to be patrollers.

Baseline and target

• Baseline: 75% of reverts defined this way were reverted within 5.83 hours.
• Target: 75% of reverts defined this way are reverted within 4.1 hours.

Task

Please generate a report showing how this metric performed from April 1 to April 23. (Since the data end on 30 April, we do not know which of the edits in the last 7 days of the month will have been reverted within 7 days.)

The report should include:

• The P75 time between the temporary-account edit and the patroller revert.

Nice to have

• Please also include a weekly trend showing how this metric has evolved since January 2026, with one data point per week.

For the weekly trend, please include:

• Week start date.
• P75 revert time.