Model: Datapoint

1 min

Model Parameters

Datapoint
- confidence_intervalnumber[]
  Example
```
[
  0.010399560300730457,
  0.0850821459929161
]
```
  Description
  The confidence interval measures the uncertainty around improvement. It starts out wide and shrinks as more data comes in. Significance means that the confidence interval is completely above or completely below 0. If the result is significant and positive, the confidence interval will be above 0. If the result is significant and negative, confidence interval will be below 0. If the result is inconclusive, confidence interval includes 0
- confidence_interval_scalednumber[]
  Description
  The confidence interval with bounds that are scaled by the baseline conversion rate.
- end_of_epochboolean
- is_most_conclusiveboolean
  Description
  Indicates that this is the best performing variant for this metric. Also referred to as the 'Winner'
- is_significantboolean
  Description
  Indicates if significance is above your confidence threshold
- lift_statusstring<equal | better | worse>
  Description
  Indicates whether a variation is doing better/worse than the baseline after taking the metric's winning direction into account.
  Allowed
  equalbetterworse
- significancenumber<double>
  Description
  The likelihood that the observed difference in conversion rate is not due to chance
  Format
  double
- valuenumber<double>
  Description
  The estimated improvement for this variant compared to the baseline
  Format
  double
- value_scalednumber<double>
  Description
  The estimated improvement scaled by the baseline conversion rate.
  Format
  double
- visitors_remaininginteger<int64>
  Description
  The number of estimated visitors remaining before result becomes statistically significant. A value of 9223372036854775807 means the value is not available.
  Format
  int64

A value with statistical context