Saturday, 8 December 2012

Visualizing Gamer Achievement Profiles using R

In this post, I'll describe how to go about visualising and interpreting gamer achievement data using R, the open source tool for statistical computing. Specifically, I'll show how you can create gamer achievement profiles based on publicly available achievement records from the Steam community API.

The visualisations and data interpretation will hopefully be of interest to a general audience, but for the more technically inclined reader I've included the steps required to create the visualisations. If you're mainly interested in the analysis and interpretation, you might want to skip ahead to the Achievement Rate Distributions section.

If you're not a coder, don't be put off - R really is straight forward. The following histogram, for example, can be created from a data set using just two lines of code:


This histogram shows global achievement rates (in percentage points) for all Steam achievements - more on this below.

Achievement Data

So what gamer data are we talking about?

The Steam community API provides both individual and global achievement records. For individual gamers, you can retrieve the lists of achievements they hold on a game-by-game basis. For the community as a whole, the API provides access to the global achievement rate - that is, the percentage of players who hold that particular achievement.

Using the approach described in a previous blog post, it's relatively easy to obtain these data sets, though a little time consuming when it comes to reading the global achievement rates for all games.

The global achievement data set that I created looks like this:


The data is simply one line per game achievement, with three whitespace-delimited columns corresponding to the Game ID, the Achievement ID, and the global achievement rate. You'll notice that the achievement IDs are quoted using the pipe character, which is necessary because some achievement IDs include spaces or quote characters.

The achievement data for specific gamers is quite similar:


Here, each line corresponds to one achievement held by a gamer, whose identity is indicated in the first column. I also chose a different quote character here because some gamer IDs happened to include the pipe character.

R Basics

R is popular tool among data miners because, among other things, it provides an easy way to generate "publication ready" charts such as histograms and scatter plots.

Getting up and running with R is simple. You can download an installation image via the R project homepage. Once installed and started, R provides console for issuing commands, as shown below:



To load a data set, you can use read.table:

achrates <- read.table("ach-rates-full.txt", header=T, quote="|")

The above reads the contents of a data file (ach-rates-full.txt) in table format into memory, accessible via the variable name achrates in this case. The parameters indicate that the file includes a header line, and that column values are quoted using the pipe character.

To view the data, simply type the name of the variable followed by carriage return and R will print out the contents. Use dim to obtain the dimensions of the data, e.g.:

> dim(achrates)
[1] 30081     3

I also found the subset function to be handy. You can use it to create a new dataset, based on some criteria such as user name or game ID. For example to obtain all global achievement rates for Half Life 2, you can type:

ar.hl2 <- subset(achrates, Game == 220)

That's all you need in order to read a data set, view the contents, and to select a subset. But let's move onto something more interesting, and generate a few histograms...

Achievement Rate Distributions

To generate a histogram of values from your data, use the hist function. The histogram shown at the start of this post (and repeated just below) was generated from the global achievement data as follows:

hist(achrates$Rate)

This generates a simple, no-frills histogram of the global achievement rates for every achievement in Steam.


How to interpret the data? Fundamentally, the data appears to show that the vast majority of achievements in Steam are held by only a small percentage of players for each game. This isn't so surprising, given that many games on Steam are for casual gamers. Also many games can be bought in bundles, which can lead to many games either being left unplayed, or played just once or twice - certainly that's the case for the games in my Steam account. It's also worth noting that a few achievements seem to have been created for test purposes, so will naturally only be held by a tiny proportion of gamers (i.e. the game developers, more than likely).

Digging into the data a little deeper provides further insight into the playing habits of Steam gamers. The following lines generate a histogram for a particular user, based on individual achievement data:

gamerdata <- read.table("user-ach-rates.txt", header=T, quote="~")
gd.user <- subset(gamerdata, User == "SomeUser")
hist(gd.user$Rate)

Two of the gamers in my social circle (let's call them Mario and Luigi) have quite distinct profiles of the type of achievements they tend to get.

Mario has over a thousand achievements, coming from a total of 23 games. The histogram of global rates for his achievements looks similar to the overall distribution:


So Mario holds many achievements that are not typically held by other gamers for the games he plays. Luigi on the other hand has about 450 achievements, coming from 35 games. His histogram looks like this:



The difference is quite apparent: the achievements that Luigi gets tend to be those held by a good proportion of other gamers, and he has fewer of the hard to get achievements.


Interpretation

Broadly speaking, the above profiles describe two quite distinct types of gamer.

The first - Mario - has a few games that he plays all the time. Mario gets most or all of the achievements, clocks up lots of game hours, and perhaps tends to the e-sports end of the gaming spectrum - playing multi-player games with friends or adversaries over the net.

The second - Luigi - has more games, and tends to dip in and out them. This type of gamer is perhaps more interested in the game experience or story, rather than obtaining every achievement or exploring every area of a game. A Luigi gamer fits more in the category of casual gamer.

Of course these are my interpretations of the data from a few simple data plots, and would need to be backed up with further data capture and analysis to hold any serious weight.

But hopefully they hint at what might be possible with such data. One can imagine, for example, building classification systems that are able to categorise gamers based on their achievement profile. Such categorisations could be used to generate recommendations or targetted adverts, friend suggestions etc. There may also be other rich sources of related data available to further enhance the gaming ecosystem.

Note on data quality

In a previous blog post, I drew attention to a few issues present in data retrieved from the Steam community API, and some of these cropped up again while I was creating the visualisations here. As such, the set of global achievement rates may not be complete or may have some spurious entries from test achievements which may slightly increase the skew towards low achievement rates. 

Friday, 16 November 2012

Anomalies in Steam Community data

In a recent post I introduced the Steam Community API, and showed how to retrieve gamer data and perform a few simple but fun analyses.

While writing the posting, I came across several problems associated with the data that's returned. If you're thinking about using Steam Community data, it's worth bearing these anomalies in mind because of the impact they'll have on downstream processing and further analysis.

Frustratingly, the quality of the data available through the Steam Community API is quite variable - in particular there are many discrepancies between global achievement data compared to achievement data for individual players. I also came across several global achievement rates that were clearly invalid, and in some cases found that global achievement records for games were totally missing.

The net result: it's hard to trust that the data that's returned. It is still possible to analyze returned data, but you're going to need strong validation, normalization (e.g. of player achievements against a 'gold standard'), and potentially multiple attempts to retrieve equivalent data to ensure what you have is accurate.

Below, you'll find a (non-exhaustive) list of data quality issues I came across, along with some examples, and a little discussion about the problems they introduced and workarounds I used.

Disclaimer: the issues described here reflect my experience while using the Steam API to retrieve data in bulk, to allow me to analyse data for a large number of gamers. Your experience may differ - if so, please let me know.

"Test" achievements

One of the most obvious problems you might find are spurious achievements associated with games. These include a few that can easily be filtered ("null" or empty names) as well as others that are more problematic - such as many 'test' achievements. For example:

TEST_ACHIEVEMENT
TestAchievement
Achievement_Test

Those readers familiar with the Valve games Portal and Portal 2 will realize why these can't be easily filtered - many valid achievement names include some variation of the word "test", e.g. portal_escape_testchambers.

I only spotted these when querying global achievement statistics, so it's possible that problem may just be a filtering issue on that particular API endpoint.

Another example can be seen when compare my (woeful) achievements for AaAaAA!!! - A Reckless Disregard for Gravity with the global achievement data. You should see one extra achievement testo2 which a vanishingly small number of people have achieved - more than likely because it's a leftover artifact from when the game was integrated into Steam.

Out of date achievement lists

Another closely related issue is that personal achievement lists can get out of step with global achievement lists, casting doubt on the reliability of any comparisons made between player achievements.

For example while processing the global record for The Legend of Grimrock, I noticed two additional achievements in comparison to my record:

FIND_ALL_TREASURES, complete_game_normal

It's worth noting that the FIND_ALL_TREASURES achievement appears in lower case in my record, but the complete_game_normal entry was missing completely. As a result, it's necessary to normalize all achievement records for players before making any comparisons, which unfortunately means making assumptions about why entries are missing (e.g. that the game hasn't been played recently) and how to fix the data.

Interestingly, since last viewing the global stats for this game, the data has become one of the ...

Missing global records

A more severe, though in some ways easier to handle issue is that global stats for some games are just missing - though this seems to be an intermittent issue.

The aforementioned Legend of Grimrock is currently one of them, as was Civilization V a couple of weeks ago. This seems to be an API specific issue, because the equivalent website for The Legend of Grimrock shows many achievements as I write this.

It seems that obtaining global stats for games is a bit of a hit-and-miss affair, so be careful with any apparently empty lists of achievements you may see and don't assume that such responses are correct when considering further processing.

Achievements with huge percentages

The final issue I've come across is that some of the global stats for achievements are simply incorrect. For example RAGE by iD Software has several achievements held by over 730,000% of players.

Thankfully this particular issue is easily detected, and offending achievements can be filtered easily.

Photo credit: wilhei55 / Foter / CC BY

Sunday, 4 November 2012

Harvesting Data from the Steam Community API

Introduction

The Steam community API is a web service that provides public access to information about Steam users, their games, achievements, and other related information. In this blog posting I'll describe some of the interesting data you can access, as well as how to model, retrieve, and process that data. I'll also show you how to generate a few fun, simple rankings and statistics for a group of steam gamers.
 
This is primarily a technical article, but it concludes with the results of a simple analysis performed over a small number of friends and aquaintances on Steam, which may be of interest to the non-technically inclined.

The examples shown here can be reproduced using the sample code found in this GitHub repository. It's a work in progress, but hopefully provides enough insight so you can either repeat the results or build your own equivalent.

Accessing the API

The first thing to know is that Steam community data is accessed using a RESTful web service, through a number of related endpoints. Many of the endpoints don't require authentication, but some require you to register for a key which you then provide as a parameter when interacting with the API.

You'll find links to the API documentation below - see the first link for details on how to get a key:

Steam Web API Documentation (high level)
Steam Web API Reference
Steam Web API Self documenting API endpoint
How to access Community Data 

The "Web API" supports both XML and JSON formats, while the closely related "Community Data" endpoints only support XML - it appears the latter are just public pages with an additional parameter of xml=1. In the rest of this posting, I provide XML examples for consistency, but the JSON resources seem to be equivalent. All of the URIs described below can be accessed using the HTTP GET verb, and in all cases appear to be browser-friendly (try clicking the examples).

There are also one or two client libraries available for different languages, notably steam condenser which is available for Java, PHP, and Ruby. Unfortunately I hit a bug caused (I believe) by changes to the behaviour of the steam API, and ultimately decided to using directly HTTP given that the API is relatively straightforward.

Available data

What kind of data can be accessed via the API? Some of the most interesting types of data are user profiles and user game lists, along with user achievements, which many users choose to make public. It's also possible to retrieve global achievement lists for games, which include percentages showing the proportion of players with the game who have a given achievement.
There's actually lots more information available, such as friend lists, statistics on play time, etc. But for now, let's focus on the above.

To make it easier to work with the data, it helps to establish a core domain model. That is, a set of concepts and relationships describing the problem domain. This helps with understanding the data, reasoning about how to process it, and further down the line how describe the data in code.

Given that we're interested in users, games, and achievements, the domain model is fairly simple:


Simple UML domain model for player data. Diagram courtest ObjectAid.

The above diagram was generated from core domain classes in the sample project, using a view-only UML modeling tool called ObjectAid. Aside from the three main concepts, you'll see relationships representing the fact that users have games, that games have achievements associated with them, and that users have acheivements either held or yet to be achieved. You'll also see a few attributes for key data such as steam ID, game name, etc.

Data retrieval

Before retrieving data from the various API endpoints, you'll need to find one or more Steam IDs. There are a few different ways of referring to Steam users, these include personas (nicknames), login account names, identifiers reported by game servers that start STEAM_, and 64 bit community IDs.

We're interested in the 64 bit steam community variants, which unfortunately require a little effort to obtain. A good starting point is your profile page which can be accessed via an "id" or via the unique profile ID - the behaviour of the "id" endpoint is ambiguous, but it appears to attempt to resolve users by their registered nicknames. I ended up viewing the page source on my friend list page or on specific profiles in order to obtain community IDs. It's also worth noting that the API endpoint for retrieving friend lists may be the most reliable method.

If you have a Steam ID in one of the other formats, you might look into one of the sites dedicated to converting between the various ID formats (example). However none of the sites I found generated 64 bit community IDs so friend lists may be the best option.

Once you have an ID or two, you can start to pull down some data. Below, you'll find an outline of key API interactions needed to get player data, game lists, and achievement information.

User profiles
First up, user profile data. You can get individual player summaries using a simple variant on the user profile link - just add xml=1 and you'll receive a computer readable version (example). Alternatively, use the Steam Web API to retrieve player summaries in batch, as follows:

URI:
http://api.steampowered.com/
ISteamUser/GetPlayerSummaries/v0002/?
key=[YOUR KEY HERE]&steamids=[STEAM 64 IDS]&format=[xml OR json]



For the sample stats shown at the end of this posting, I just needed the user's persona name which I retrieved using the second method shown above.

User game lists
A user's game list can also be retrieved using a simple variation on a web page URI. In this case, add /games?xml=1 to the end of a profile page URI (example) and you should have a complete list of games owned by the player.

The key data need you'll need to pull out of the response is the appID - that is, the unique identifier for the game. The response also includes other player data you might find useful, including game names, play time, etc.

User achievement lists
Once you've retrieved a list of unique game IDs for a particular player, it's possible to start retrieving something a little more interesting - individual achievements for those games.

Again, you can obtain a player's achievements for a game in two ways: using a profile page URI, and using the Web API. Helpfully, you'll find links to a player's game achievements in the game list response. Adapt these by adding xml=1 and you're away (example - warning: possible game spoilers for XCOM: Enemy Unknown). I actually used the alternative provided by the Web API, as follows:

URI:
http://api.steampowered.com/
ISteamUserStats/GetPlayerAchievements/v0001/?
appid=[GAME ID]&steamid=[STEAM 64 ID]&key=[YOUR KEY HERE]&format=[xml OR json]

The key information you'll need here are the unique identifiers for the achievements (apiname) and the flag indicating whether or not the player holds that achievement (achieved, values 1 or 0).

Global achievement data
The final endpoint that's worth reviewing retrieves global achievement lists and percentages for games. This data only appears to be available via an unauthenticated endpoint through the Web API (example), as follows:

URI:
http://api.steampowered.com/
ISteamUserStats/GetGlobalAchievementPercentagesForApp/v0002/?
gameid=[GAME ID]&format=[xml OR json]
  
Example response

This achievement data is useful both for cross-referencing with user achievement data, and for comparing individual achievements with global levels. The code in the sample project pulls down this data and uses it to validate and normalize user achievement lists.

Putting it all together

So by now, it's hopefully clear what kind of data you can access via the Steam Community API, and how to retrieve it using the various HTTP endpoints. But that's not quite enough to be able to start working with the data.

Below, you'll find an outline of the steps required to put together a cohesive data model that can then be analysed, persisted, and processed further:

For one or more Steam 64 identifiers:
- Retrieve user profile data, create a user record.
- Read the list of user games (capturing game IDs), associate them with the user.
- Read user achievements per game (capturing game ID, plus achievement ID and status) associate with user.

The end result should be a collection of users, each with an associated set of games, and for each user/game pair, a set of achievements both held and yet to be achieved. In the sample project, I use Java classes to hold user, game, and achievement entities, along with a few Collection objects to record game lists and achievement outcomes. The sample code also retrieves global achievement data for validation and normalization.

Popular games, Biggest Achievers

Using the above approach, I harvested data for eight Steam friends and aquaintances, then generated a list of the top games for the social group along with a ranking of the most accomplished players.

Image courtesy of smarnad / FreeDigitalPhotos.net
First, the top 10 games, according to popularity:

Pop.GameAch'ments
8Portal15
7Portal 250
6Amnesia: The Dark Descent0
6Counter-Strike: Global Offensive193
6Counter-Strike source148
6Counter-Strike source: Beta154
6Half-Life 2: Deathmatch0
6Half-Life 2: Lost Coast0
6Left 4 Dead 269
6Super Meat Boy49

No surprises there then. Portal and Portal 2 being highly popular, followed by a few of the top indie games and several stock Valve games. The only slightly puzzling thing being that Half-Life 3 sorry I mean 2 appears at position 11 (not shown), and is only shared by five players, while six have a copy of HL2: Deathmatch. The likely reason being that one person owns the Source Multiplayer pack which includes HL2: Deathmatch.

Next, who are the biggest achievers. Names have been anonymized to protect the innocent:

PlayerAchievmentsRatio
Chas1000/394925%
Olly at home DOTT789/655012%
Shreddies574/324518%
gryffindorpotential433/369712%
cryptoGoat193/198910%
HappyKittens40/6576%
unpronouncable29/11183%
Cuppa11/21741%

Congratulations Chas, the runaway winner with one thousand achievements, and the highest proportion of possible achievements held. Coming in close behind are Olly and Shreddies, Olly holding the higher number of achievements but Shreddies having achieved a higher relative proportion. Bringing up the rear, the wooden spoon award goes to Cuppa has both the lowest number of achievements overall, and the lowest overall proportion.

These particular stats are just for fun and are shouldn't be taken too seriously, but they do hint at some more compelling uses of the data. For example, it would be interesting to go beyond pure rankings and further analyse the achievements held by gamers from a particular social group. But that's the topic of a future post.

A final note on data reliability

One final thing to mention is that my experience with the quality of data exposed by the Steam Web API has been mixed. The data exposed by the by the community pages (i.e. the public pages, with xml=1 added) seems more reliable than the data provided Steam Web API. One reason might be that the Web API provides less filtering, while the profile page is designed to be human readable and thus may be subject to greater filtering or curation. 

Next time, I'll be discussing those issues in more detail as well as discussing the importance of, and problems associated with, obtaining reliable data.

Sunday, 28 October 2012

Fresh Pickings

Welcome to The Variable Tree, a blog all about Programming and Software Engineering, with a leaning towards articles about data mining and analytics.

This blog covers range of topics, but I (that is, me, James Siddle - see Bio below) have a particular interest in data mining and related topics. That includes things like data extraction pipelines, Natural Language Processing, classification and prediction using Machine Learning, data storage techniques, and more. That said, you might also find topics about programming in general, perhaps something about new programming languages, domain modelling, observations on development processes, or the odd article about interesting applications.

First up, I'll be writing a few articles about how to retrieve, process, and analyze data from the Steam community API, showing how to generate a few interesting statistics from online gaming communities.

Hopefully you'll find something of interest, enjoy reading :)

Image courtesy of anankkml / FreeDigitalPhotos.net

James Siddle
I'm an all round software engineer, and I've been programming since an early age - I started with BASIC on an Acorn Electron, and have touched every more or less every platform and language you can imagine since then. I work for Digital Science, an arm of Macmillan publishing, we build information systems and services to help scientists. That includes plenty of data extraction, Natural Language Processing, analysis, storage, and ultimately the creation of services to allow customers access to the data we gather and the information we extract. This blog is not affiliated with Digital Science, and all views expressed here are my own.