Our impact estimates, scored against reality.
Whenever PulseWatch publishes an estimated user-impact range during an incident, we record it. When the vendor later publishes an official postmortem, we add their number and our delta. Every row below is auditable.
Estimates verified
0
Vendor-confirmed and indexed
Hit rate
0%
0 of 0 fell inside our published range
Mean absolute error
0 pp
Average gap (percentage points) vs vendor
No vendor-confirmed estimates yet
Once any vendor publishes an official postmortem with an impact percentage, we'll index it here alongside the estimate we published at the time.
How this works. While an incident is active, PulseWatch publishes a ranged estimate of likely user impact, derived from the vendor's market share, the incident's severity, synthetic-probe coverage, and peak user-report volume. Units are always user-facing actions (checkout attempts, API calls, sessions) — never financial figures. After the vendor publishes their official postmortem, an admin records the confirmed number here and the row's delta is computed automatically. Methodology is versioned, so old rows remain attributable to the math that produced them. Read the full methodology →