Steam Games Dataset : Player count history, Price history and data about games

This dataset consists of player count history, price history, and data about 2000 applications available in the Steam games platform. The 2000 applications are the applications in Steam with the highest player count within the last 24 hours on the 11th of December 2017 chosen based on the player statistics provided in SteamDB (https://steamdb.info/graph/). The player count history of each application from 14-12-2017 to 12-08-2020 is available. Player count was collected in 5-minutes intervals for the top 1000 applications (PlayerCountHistoryPart1.zip) and in 1-hour intervals for the next 1000 applications(PlayerCountHistoryPart2.zip) using the ISteamUserStats interface of the Steamworks web API. One file per each application is available containing UTC Time and Playercount columns. The price history of non Free to Play applications are available from 07-04-2019 to 12-08-2020. Price information was collected using the StorefrontAPI in USD currency and is available in daily frequency. Hence, price changes that did not last for 24 hours would not be available. One file per each application is available containing Date, Initialprice, Finalprice, and Discount columns. Data about the applications were collected using the StorefrontAPI, except for tags which was collected using SteamSpy API. The format of each file is described as follows: applicationInformation.csv : [ appid, type, name, releasedate, freetoplay ] This file consists of the application ID used in Steam, application name, the release date of the application, application type representing whether the application is a game, mod, demo, advertising or dlc. ( There are 1963 games in the dataset out of the 2000 applications. ) Whether the game is Free to play or not is represented as a binary feature. applicationDevelopers.csv: Consist of the appid and a list of the developers of the application. applicationPublishers.csv: Consist of the appid and a list of the publishers of the application. applicationSupportedlanguages.csv : Consist of the appid and a list of the languages the application supports. applicationPackages.csv: Consist of the appid and a list of the package IDs the application is included in. applicationGenres.csv: Consist of the appid and a list of the genres assigned to the application on Steam store. applicationTags.csv: Consist of the appid and the list of the tags assigned to the application on Steam store. Research Paper(s) resulting from this dataset: D. Vihanga, M. Barlow, E. Lakshika and K. Kasmarik, "Weekly Seasonal Player Population Patterns in Online Games: A Time Series Clustering Approach," 2019 IEEE Conference on Games (CoG), London, United Kingdom, 2019, pp. 1-8, doi: 10.1109/CIG.2019.8848108.

本数据集涵盖Steam游戏平台上2000款应用的玩家数历史、价格历史及相关元数据。这2000款应用是基于SteamDB（https://steamdb.info/graph/）提供的玩家统计数据，选取2017年12月11日过去24小时内玩家数排名最高的Steam应用。每款应用的玩家数历史数据覆盖时段为2017年12月14日至2020年8月12日。其中前1000款应用的玩家数以5分钟为间隔采集，数据存储于PlayerCountHistoryPart1.zip；后1000款应用的玩家数采集间隔为1小时，数据存储于PlayerCountHistoryPart2.zip。采集过程使用了Steamworks Web API的ISteamUserStats接口。每款应用对应一个独立数据文件，包含协调世界时（UTC）时间与玩家数两列字段。非免费游玩（Free to Play）应用的价格历史数据覆盖时段为2019年4月7日至2020年8月12日。价格信息通过StorefrontAPI以美元为单位采集，采集频率为每日一次，因此持续时长不足24小时的价格变动将无法被记录。每款应用对应一个独立数据文件，包含日期、初始价格、最终价格与折扣四列字段。应用相关元数据除标签外均通过StorefrontAPI采集，标签数据则通过SteamSpy API获取。各文件格式说明如下：1. "applicationInformation.csv"：字段为[appid, type, name, releasedate, freetoplay]。该文件包含Steam应用ID、应用名称、应用发布日期、应用类型（用于标识该应用为游戏、模组、演示版、广告或DLC），以及一个二进制特征，用于标识该应用是否为免费游玩游戏。本数据集2000款应用中包含1963款游戏。2. "applicationDevelopers.csv"：包含应用ID与该应用的开发商列表。3. "applicationPublishers.csv"：包含应用ID与该应用的发行商列表。4. "applicationSupportedlanguages.csv"：包含应用ID与该应用支持的语言列表。5. "applicationPackages.csv"：包含应用ID与该应用所属的捆绑包ID列表。6. "applicationGenres.csv"：包含应用ID与Steam商店为该应用分配的游戏类型列表。7. "applicationTags.csv"：包含应用ID与Steam商店为该应用分配的标签列表。使用本数据集产出的研究论文：D. Vihanga、M. Barlow、E. Lakshika与K. Kasmarik，《在线游戏的每周季节性玩家群体模式：一种时间序列聚类方法》，收录于2019年IEEE游戏会议（CoG），英国伦敦，2019年，第1-8页，DOI: 10.1109/CIG.2019.8848108。

搜集汇总

背景与挑战

背景概述

该数据集收录了Steam平台上2000个高玩家数量应用程序的玩家数量历史、价格历史及详细信息，覆盖时间从2017年12月至2020年8月。数据集特点包括玩家数量以不同时间间隔收集（前1000个应用每5分钟，后1000个应用每小时），价格数据以美元计且每日更新，并包含应用类型、开发者、语言等元数据，主要用于游戏玩家行为分析和时间序列研究。