GameAnalytics

It's hard to test AB's. There's no in-built system to force the game into one cohort. Would be nice to have it or any advice how to make it

Sometimes especially when test sample is not that huge a few very engaged players can alter the picture entirely. Is it possible to filter 2-5% most engaged players to get more median results?

Add some custom progression event completion% into AB like it would be great to see what % of users finished 10 or 30 levels in the game in cohorts A and B

add number of users in the test in the Engagement tab ( i can only see it in conversion) this will help to understand if there's meaningfull amount of players

We created an A/B test variant with text values, but the graph displays numbers instead. How can I determine which number corresponds to which variant? It would be helpful if this information were displayed in the test details.

I am encountering an issue with the behavior of an A/B test configured in Remote Config. Below are the details: I created a remote value named "Split6". An A/B test was successfully set up for this value. Upon initializing the application on iOS, the value retrieved corresponds to the one set on the "Remote Configs" page (verified via Live Events). Reinstalling the application on iOS does not change the retrieved value; it remains the same. If I manually update the value in the "Remote Configs" tab, the updated value starts to appear. This behavior suggests that the application is not retrieving the value from the A/B test but directly from the Remote Config. I have verified that the version number and variable name in the A/B test are correct. The issue seems to occur only on iOS devices. Could you please help me identify why the A/B test values are not being applied? Let me know if additional details or logs are required. Thank you for your assistance!

It would be great to be able to see all the A/B test configurations after the A/B test has started.

As the title says, Ad Revenue / DAU is impacted by retention. e.g. If I have very low DAU, vs a version with higher retention, but lower impressions per DAU, version B could actually be the better result.

Comparing LTV values for each a/b variant can be very useful to decide on the results or targeting LTV increase as the core aim of a/b test. This feature will be very useful for everyone.

I would like to see "Banner Ad Views per user per day" added to the core metrics of the A/B test.