Not sure if this is possible with Unity Analytics, but wanted to check in before I go off and look for something else.
I’ve got a level-based game, and every level, the player can earn either 1, 2, or 3 stars. I’d like to measure level difficulty by examining how many stars are earned on each level, so I set up a custom event with the following parameters:
Event: Beat Level
- Parameter 1: INT Level [returns 1-90]
- Parameter 2: INT Stars [returns 1,2,3]
What I would like to do is view all the levels & their average star counts at the same time. This would allow me to quickly identify levels that are too easy/too difficult. To achieve this, it makes sense to assign the levels parameter to my x-axis, and the stars parameter to my y-axis.
Currently, this is not possible, as time is always displayed along the x-axis. Defining a date range would be enough control for me (I don’t need to see a per-day breakdown and find it frustrating that this is always along the x-axis).
One possible workaround is to create a custom event for each level such as BeatLevel01, BeatLevel02, etc etc, and then from there view parameter stars. This is extremely problematic because it requires viewing only one level at a time, not to mention setting up 90 events. And what if I’d like to set up an event LoseLevel? Through the workaround method, that would double the number of events from 90 to 180.
To make things even more complicated, there is a filter I’d like to add to view this data accurately. Since our game has a high percentage of replay-ability, it’s important to distinguish repeat-level plays from first-time level plays. So I created a boolean parameter called firstTime. Ideally, I’d like to only look at the average star counts on a per level basis for first-time level players.
There are a few more events I’d like to be tracking and a few more parameters that I did not list for the simplicity of this post, but you get the idea. Being able to compare parameters for an event along x and y axis would be extremely beneficial to understanding the big picture in terms of level design.
You may argue that funnels could achieve a similar purpose by measuring player drop-off rates on an aggregated per-level basis. The problem here is that funnels have no way to distinguish between a level that is easy and a level that is fun but challenging, because it only knows when players turn off their app. Being able to flexibly measure star count, and also, for example, most used game characters on a per level basis would be a significant step towards understanding how to improve our game, and currently Unity Analytics limits our designers from seeing the big picture.
I’m interested to hear any workarounds to this issue, if they weren’t already addressed above, and if it isn’t possible, I’d love to hear what other data measuring solutions are currently out there that might be a good fit given our requirements. Do many of you feel the same way about Unity Analytics, or am I just an edge case? Would appreciate your thoughts if you’ve got this far! Thanks.