Holdout
10000Project IDs this Holdout applies to. An empty array means All Projects.
The userId of the owner (or raw owner name/email for legacy records)
Lifecycle stage of the Holdout. Use the start, start-analysis, and stop endpoints to move through the lifecycle.
draft, running, analysis-period, stopped ID of the companion experiment that stores this Holdout's analysis. Use it with the experiment results endpoints (e.g. GET /experiments/{experimentId}/results). Pass phase=0 for results across the whole Holdout period, or phase=1 for results limited to the analysis period.
When true, this Holdout is not selected automatically when creating Feature Flags or experiments in Projects assigned to the Holdout.
Proportion of traffic placed in the holdout group (e.g. 0.05 for 5%). An equally sized comparison group is sampled from the remaining traffic for measurement, so 2× this value enters the holdout experiment in total. Must be greater than 0 and at most 0.5.
x <= 0.5Attribute used to assign users to the Holdout.
Targeting condition as a JSON string.
Per-environment state, keyed by environment ID. Environments not present are disabled.
Feature Flags held out by this Holdout. Manage these through the Feature Flag endpoints.
Experiments held out by this Holdout. Manage these through the experiment endpoints.
The email address of the owner, when the owner can be resolved to a known user.
Statistics engine used to analyze this Holdout.
bayesian, frequentist When the Holdout entered its analysis period.
Automatic stage transitions for the Holdout. Dates must be consecutive: startAt before startAnalysisPeriodAt before stopAt. Send null to delete the schedule.
The next stage transition that will run automatically.

