Perhaps I mis-interpret how unity analytics is doing the custom events calculations or there is a very big issue with the analytics.
I am sending a custom event with an int value.
I have created a segment with 1 rule:
“Segment test 01”:
“Event”, “MyEventName1”, “EventParam1”, “is at most”, 2048.
In the data explorer, I am displaying a custom event graph as following:
“MyEventName1”, “Segment test 01”, “EventParam1”, “average”.
The results are numbers around 8K to 9.33K (8000-9000+).
How would averaging numbers that never exceed 2048 give me an average bigger then 2048? (2050 is as weird as 8000 for that matter).
Have I missed the idea of how this works?
(please note these are numbers from active users, counted by the thousands, and while I can’t seem to make sense with the results shown in Unity analytics, Fabric’s Answers do seem to bring up logical values for roughly the same events).
@angeloferro - perhaps you could assist with this?
Is there a way to average custom events, whose number parameter X is between range A to B? I do not need to average the whole events for users that meet this definition but to average only the certain events who meet this definition.
What I believe I am now getting is the average of all the events of all the users who had the definition met at least once.
Hey @GuyTidhar … sorry, we’re all just back in from the holiday break.
Yeah, that looks odd to me too. Let me get someone investigating that. Could you PM me your project ID?
[EDIT]Note, though, that segments are by user, not by event. So I guess that one explanation is that if your event allows the user into the segment, but you also allow events into that segment that throw off your averages, that result might be entirely expected. When we see the specific setup, it might be easier to make that determination.[/EDIT]