Game Variant Analysis
Ajeet chaudhary
January 5, 2020
Problem statement A Bangalore based online gaming company launched 3 variants of a game for Android users. We have to find which variant is optimum balance of economy,engagement and retention.
Terms Used in Data Experiment Name =‘content’ Experiment Version = ‘2’ variant_group =variant (1, 2, 3) Cli=client (Android - 0) DAU: Daily Active Users in the Game users: Players performing a specific activity. Subset of DAU Economy Terms: Spend refers to the player using coins to buy hints Purchase (Buy) refers to the player buying coins (IAP) Earn refers to the player earning coins due to his progress in the game. Engagement Actions: view: viewing a puzzle clear: solving a puzzle fail: failing on a puzzle use_hint: spending coins to reveal hints on a puzzle. Retention: dn: Percentage of people playing on Date=D+n out of people who installed on Date=D Most of the other terms used are self-explanatory
OVERALL ECONOMY
On individual user,Lets see how much they earn,purchase and spend on average. 
- For variant 2,players earn maximum coins which mean players qualify maximum levels in this variant,purchase maximum coins mean they like to spend money to play this engaging variant and spend maximum coins to know hints.
This makes variant 2 most balanced variant economically.
Now we will see total coin flow:- 
- For variant 3 players earn maximum coins and spend maximum coins to buy hints also. Purchase numbers are nearly equal for all the variants.
- Daily users for variant 3 are higher as compare to other two variants,which increases total cash numbers, So total coins number can be misleading.
OVERALL ENGAGEMENT
From above bar chart we can clearly see variant 2 engage maximum number of users.
- clear and view are most dominant actions for all the variants.
Use_hint is least dominant action.
OVERALL RETENTION
- Here variant 2 is clear winner in the race of retention.
- This graph shows datewise installations. Variant 3 has maximum installs in the time span of one and half month (mid aug and sept data available).
CONCLUSION: Variant 2 is best balance of economy,engagement and retention.
Thank You
.