Search Unity

DateExplore for examining CustomEvent

Discussion in 'Unity Analytics' started by SaladbowlDev, Aug 8, 2017.

  1. SaladbowlDev

    SaladbowlDev

    Joined:
    Sep 30, 2016
    Posts:
    6
    Hello, all.

    I would like to that I am doing right with UA(Unity Analytics) for following case.

    The game I am working in has User status named Tier Level that indicates how skillful users are.
    And I would like to the distribution of Tier Levels by Life Cycle standard segments.

    So, I did as following image

    SegmentForTierLevelsAtDataExplore.png

    But I am afraid that this way would not give exact distribution status of Tier Level, because of following.

    1. User A has Tier Level 2 and User B & C both have Tier Level 1. All Use A, B, and C are in Life Cycle segment 4-7 Days.
    2. User A triggered StartMatching 2 times and Each User B & C triggered StartMatching 1 time.

    Does this scenario results the distribution of Tier Level as followoing?
    Tier Level 1: Count 2
    Tier Level 2: Count 2

    If so,this way to get exact distrutions for Life Cycles must be not precise.

    How can I achieve to get precise Tier Level distribution for Life Cycle segements?


    Thank you in advance

    Danny
     
  2. JeffDUnity3D

    JeffDUnity3D

    Joined:
    May 2, 2017
    Posts:
    14,446
    So you know, segments are not mutual exclusive. Once a user enters a segment, they will continue to be a member of that segment, even if the next day they meet the requirements for a new segment. Also, I'm not entirely clear on your requirements. But it sounds like you might be asking if custom segments are by session or by user. They are by user, so if one person triggers an event five times, the segment will still only have one user. One solution for you, if you have the Pro version, would be take advantage of Raw Data Export, described here: https://docs.unity3d.com/Manual/UnityAnalyticsRawDataExport.html
     
  3. SaladbowlDev

    SaladbowlDev

    Joined:
    Sep 30, 2016
    Posts:
    6
    Hi, Jeff


    So, my requirement is to see the exact distribution of Tier Levels for all users by passing days.

    For example, Let's assume all users are 10 and no more users will install the game.
    The distribution of Tier Levels for Life Cycle 1-3 could be following. Tier Level 1: 8, Tier Level 2: 2 = total 10 users
    For Life Cycle 4-7, it could be following. Tier Level 1: 5, Tier Level 2: 2, Tier Level 3 = total 10 users
    For Life Cycle 8-15, it could be following. Tier Level 1: 2, Tier Level 2: 4, Tier Level 3: 3, Tier Level 4: 1 = total 10 users
    and goes on.

    I heard standard segments are mutual exclusion and that's why I tried the way in previous post's image to get the result like the example. I am filtered by using Life Cycle which is the standard segment, but exams CustomEvent.

    Let me ask 2 questions to clear my understanding.

    1. Assume that User A is in the standard segment Life Cycle 1-3. Then, if User A triggered StartMatching with Tier Level 1, 2 TIMES, What number I will see in the Data Explore for StartMatching.tier for LifeCycle 1-3? 1 or 2 for Tier Level 1?

    2. Assume that User A is in the standard segment Life Cycle 1-3. Then, if User A triggered StartMatching, but this time 1 TIME with Tier Level 1 and the other 1 TIME with Tier Level 2. What number I will see? 1 for Tier Level 1 and 1 for Tier Level 2?

    If you answer these questions, I think I can understand better for using the Data Explorer.


    Thank you in advance.
     
  4. JeffDUnity3D

    JeffDUnity3D

    Joined:
    May 2, 2017
    Posts:
    14,446
    It will be 1 each time. Is this what you are seeing?
     
  5. SaladbowlDev

    SaladbowlDev

    Joined:
    Sep 30, 2016
    Posts:
    6
    Hm, so for case 2, you mean it will be 1 for Tier Level 1 and another 1 for Tier Level 2?
    This mean that User A is recorded for StartMatching(CustomEvent with the param named tier) with Tier Level 1 and StartMatching with TIer Level 2. And it's not exactly I want to, but I guess it's characteristic of Unity Analytics.

    Ok, then. Thank you for clearing my understanding, Jeff.
     
  6. JeffDUnity3D

    JeffDUnity3D

    Joined:
    May 2, 2017
    Posts:
    14,446
    That is correct. Perhaps the following example might help. Assume you have various segments, for example segment A of those users who spent more than $1, then another segment B for users that have spent more than $5, and another segment C for users that spent more than $10. Then you have a single user who has spent $50 each day for several days. Each segment would each contain 1 member, this single user. He did indeed spend more than $1 so this is true, and he also spent more than $5 so he is also in this set, and so on. One note, if you are a Pro subscriber, you have access to Raw Data Export. If so then this suggestion could work for you https://support.unity3d.com/hc/en-us/articles/115004052703
     
  7. SaladbowlDev

    SaladbowlDev

    Joined:
    Sep 30, 2016
    Posts:
    6
    Thank you Jeff to clear things.

    I immediately started playing with RDE as you advised.