EXP-01 / Provider reproducibility
Track Board
Network totals
Traces committed
14
14 were never re-run
Providers
1
0 ranked, 1 unrated
Network reproducibility
Unrated
No rerun has ever been performed here.
Bonded
66.29MKORTX
0 KORTX slashed to date
reproducibility.score_pct is null when the network has not accumulated enough rerun evidence to state a rate. It is never rounded to 0 or 100 to fill the field.
Ranked providers
0 of 1 providers
The score is the lower bound, so the ranking answers what the evidence supports rather than what it hopes for. Both halves are drawn: a provider sitting at 69.9% lower bound with an 87.7% point estimate has been measured cautiously, not measured badly.
Nothing recorded
No provider has accumulated enough re-run evidence to be ranked.
Committed traces do not enter the ranking until they have been re-run and compared.
Unrated providers
1 of 1 providers
Providers appear in `ranked` only once their evidence reaches the minimum effective sample size (5.0). The rest are listed under `unrated` rather than at the bottom of the ranking: unmeasured is not the same as unreliable.
| Provider | Traces | Re-run | Comparisons | Short of a rating | Interval so far | Incidents |
|---|---|---|---|---|---|---|
| 7J6JV1...aPQ1X4 | 14 | 0 0.0% of committed | 0 0.0 effective | 5.0effective | No re-run yet | 0 |
A provider is rated once its weighted evidence reaches 5 effective samples. Effective samples discount old and lower-tier comparisons, so the shortfall is stated in those units rather than as a count of re-runs still to run. A provider with no re-run at all has no interval: that is no measurement, which is a different answer from a poor one.
Indexer ok / read through slot 453,936,092 / 2026-10-06 15:02 UTC