Game Analytics:
Data Collection & Visualization
CS 526 Mobile Games
Copyright By PowCoder代写 加微信 powcoder
Graphing analytics shows how people are playing your game
There is not just one way to track how people play your game
Your game will be unique, so be attentive what to track as it develops.
The first step is to construct a solid pipeline of getting analytics data from Unity and putting it into a graph form that your team can read.
No dates on X axis (It doesn’t help as activity spike will be on mid-term)
Getting Started
What makes a good visualization?
When Data visualization makes analytics theme easier to understand and contributes to improving the game.
Simple is better
Make it easier to see trends and predict outcomes.
Indicate player’s opinions
Data Collection
You will need to collect analytics data from users playing your game through the online WebGL build during midterm
Analytics Data Storage
Data Collection
Requirements:
Must work on WebGL build (important to test this: many off the shelf analytics solutions are incompatible)
Handle ~100 users playing your game simultaneously during midterms
Ability to export the data for plotting graphs
We are asking teams early on to designate members who will be responsible for analytics (PMs) and implement analytics data collection / graphing
Possible solutions: Firebase, Google Forms (see methodologies PDFs)
Note: Unity’s built-in analytics solution does not currently support WebGL
Set up analytics data collection
Export the data (e.g. as CSV or JSON)
Create graphs from the collected data
Python/Visualization libs(eg. Matplotlib)
Sheets/Excel
Example: Python + charting lib
This example will use matplotlib and pandas:
Import CSV/Json file
Traverse the data and store the data you need
Create a dataframe using pandas
Plot the graph using matplotlib
Example: Excel
Import Json
Get Data → From File → From Json
Transform Json data using Power Query
Convert the nodes into columns
Transform into tables
Expand the contents into columns
For each objects column→ Expand to new rows.
Repeat two steps above to get the table you need.
Load the data into the Excel worksheet
Filter the data
Create a pivot table
Create a graph
Previous Team’s Methods
There is a folder named ‘Analytics Methodologies’ on Blackboard that has various examples from previous teams how they created graphs from analytics data – with written tutorials. Go through them to choose one or create your own.
Analytics people on team: Your task now
Research different ways to collect and graph data from Unity
Unity forums/other forums
Methodologies Examples (On blackboard)
Choose a system and prove it works.
Create graphs based on some test analytics data.
Start integration of analytics hooks as your team’s game develops.
You should have at least three working analytics graphs showing gameplay feedback when your team delivers their greybox prototype (See Syllabus).
Needs to work in a WebGL build
程序代写 CS代考 加微信: powcoder QQ: 1823890830 Email: powcoder@163.com