We know that when playing a game, players will have different types of behaviors and events in the game. For example:
- For RPGs, users will have behavioral events such as fighting monsters, upgrading, and buying equipment within the game.
- For PVP games, users will have behavioral events such as adding friends within the game.
- In card games, users will have behavioral events such as purchasing cards and using cards within the game.
Here are the users and events. We found that "Data Analysis" is analyzing User and Event.
When users have various types of behavioral events, the events will have information related to Who, When, What, Where, How, etc.
- Who：The user who triggered the Event.
- When：Time when the Event is triggered.
- Where：Location information such as IP, country, province, city, and district when the Event is triggered.
- How：How the user triggered the Event, such as the type of device, OS type, OS version number, device brand, device model, device resolution, version of the game app, and other information.
Where What, When, Where, and How are all property states of the Event.
In the case of a competitive match game, "users teaming up for a match game" can be viewed as an "Event" named "match game" :
- Who：Users who joined match game.
- What： Type of match game.
- When：Time of match game.
- Where：IP, country, province, city, district and other location information at the time of the match game.
- How：The user's phone model, system version, and phone resolution at the time of the match game.
This is Event model and Event property information。Every behavior to be analyzed in the game can be defined as Event and should be defined as Event, which is a prerequisite for data analysis.
You can configure Event information in TapDB's "Configuration" - "Event Management" page.
User will also have a lot of information, such as registration time, registration address, registration channel, user level, device type, gender, age and other information, which are all property states of User. By using these properties, you can quickly filter users in behavioral analysis. or example, if we want to analyze the activity of paying users, you only need to analyze the users with the value of "accumulated payment amount" greater than 0.
In the case of a competitive match game, "users teaming up for a match game" can be seen as the "who" in "event".
- Who are the users who participate in the game?
- How long have they been playing the game in total?
- What is the user's score?
- Which heroes have they played?
- What is the user's gender?
- What is the user's age?
- When did the user register?
- What is the user's registration channel?
- What is the user's cumulative payment amount?
This is the User model.
The Event model and User model are called: User-event model.
Based on the User-Event model to design buried documents and collect information, you can:
- PVP games: analyze the number of participations using different characters at different levels.
- Card games: analyze the participation of players with different VIP levels in Mid-Autumn Festival events.
- SLG: Analyze how being robbed of resources by other players affects retention in the first 7 days of entry;
- RPG: Focus on key churn causes by analyzing pass rates for key levels;
- SLG: Check the resource accumulation and consumption of different levels of players for the last 7 days in order to determine how to give out gifts. It is best to achieve visual display of net inflow and outflow;