https://www.growthbook.io/ logo
#announcements
Title
# announcements
f

fierce-belgium-15334

09/26/2022, 6:35 AM
Does rollout decrease move user from B to A? Like I have feature roll out initially with 10% users and user X fall into A and then I increased to 100% now user X moved to B group after some time I have decreased roll out to 10%. here does user X get A or B ?
f

fresh-football-47124

09/26/2022, 6:49 AM
Changing the rollout percentage may cause users not exposed to get a treatment group (or vice versa), but they will not change treatments.
so in your example, at 10% they got A, at 100% they’ll still get A, and at 10% again, also A,
As long as you don’t adjust the experiment split
f

fierce-belgium-15334

09/26/2022, 6:56 AM
When I tested at User X got false at 10% and when I increased to 100% User X got true and then when I changed to 10% User X still getting false.
is it a correct behaviour ?
f

fresh-football-47124

09/26/2022, 7:11 AM
its a bit hard to tell with binary experiments
false could be ‘unexposed’ and also could be they are in a treatment group with false as the value
you need to check the trackingCallback firing or check the evalFeature() method
f

fierce-belgium-15334

09/26/2022, 7:14 AM
does “source”: “defaultValue”, when eval feature means exposed or not ?
for only feature roll outs features does trackingCallback called ?
f

fresh-football-47124

09/26/2022, 7:14 AM
defaultValue mean they matched no override rules, and were served the default value for that feature
trackingCallback is only called for experiments
f

fierce-belgium-15334

09/26/2022, 7:15 AM
ok
12 Views