Skip to main content

Table 1 Description of feature categories among the benchmark features

From: Social networks for enhanced player churn prediction in mobile free-to-play games

Type

Description

Per Day

Features relating to how often per day do certain events, e.g. how many matches a player plays, wins and loses, occur on average, median, minimum and maximum for each player.

Per Match

Features relating to how often per match do certain match related events occur on average, median, minimum and maximum for each player. These events include how often a player wins or loses a point and misses a stroke.

Over all

Features relating to how often match related events occur for each player during the 14 day observation period.

Match Types

Features relating to the how many Bet-, Friend- and Tournament matches a player plays.

Playing period

Features relating to how many days players play the game during the 14 day observation period and the number of days between their first and last played days within the observation period, i.e. the length of their playing period within the observation period.

Streaks

Features relating to how many win and loses streaks a player has, the number of matches within their streak and how long a streak is. The number of matches within a winning streak is defined as the number matches won in a row and a losing streak is defined as the number matches lost in a row.

Currency

Features relating to the hard and soft currency of players.

Purchases

Features relating to the purchases made by players, with both hard and soft currency.

Friends

Features relating to matches played against friends.

Player

Features relating to the statistics of a player

Win-Loss Types

Features relating to the different types of wins/losses, e.g. a player automatically wins a match if their opponent abandons the match.