Archive for the ‘Visualization’ Category

FIFA 10 came out on Oct 20th and it has already stolen my heart. Not because of the game, I haven’t played, nor is it soccer (or football), which I do not follow in any capacity. The reason I’ve fallen for this game is because of this website: FIFA Earth.

In the growing world of social networks and system integration, game companies have been jumping on the band-wagon. Uncharted 2 has Twitter integration (though they seem to have problems with it), Xbox Live with have Facebook/Twitter capabilities, and Giant Bomb links player’s achievements together. Integrate and aggregation is the talk of the town.

Now what? What do we do now that we have achieved this integration. Well, visualization and analytics of course. Sure Uncharted 2 can tweet when you finish a game chapter but how many other players have completed that chapter too. When do those players play, did they finish the chapter in the same amount of time as you? While integrating achievement systems, Facebook and Twitter makes communication easier it does not make the information presented easier to understand. Now that we have these swiss-army knife systems let’s do something with them.

FIFAEarth

Enter FIFA Earth. There are other visualization/analytic game systems available; Valve shows Steam statistics on their website for example. However, it is rare to find one that is informative, interactive and aesthetically pleasing. FIFA Earth is just such a system. It does not have an extensive list of information to sift through but what it has is style.

FIFA Earth aggregates every match that is played in FIFA 10 from around the world, while any player is connected to the PlayStation Network. From what I have been told soccer is a very important sport worldwide so it only makes sense that a video game simulating real world soccer is just as popular; which means a lot of recorded games. 39 million to be exact. Each day’s total games played is available and is similar to other visualizations like Steam or Noby Noby Stats.

Each time a player completes a game in FIFA 10 it is uploaded online and goes into two different categories. One is for that player’s country and another is for the player’s chosen club. Categories based on country and club are ranked against each other by their ‘win percentage,’ which is calculated by their wins, losses, and draws from all the matches in that category (across all players). There is also an option to view their rank changes across time. I see that players from Ecuador have a winning percentage of 53 today and have been above 40% since October 7th.

FIFAStat

FIFA Earth shows historical win percentage data for categories based on country and club.

All of this information is displayed using a three dimensional image of earth as viewed from space. Which makes sense given the data presented is all geo-located. This feature is also combined with FIFA Earth’s own Twitter feed, which displays tweets about FIFA 10. Each tweet is marked on the globe and they stretch across the screen in a list next to the Earth’s image, it’s quite elegant.

FIFATwitter

I only noticed tweets with the word ‘FIFA’ in them so I do not know if other keywords are searched for in FIFA Earth’s Twitter feed.

When it comes down to it FIFA Earth has three things: relevant Twitter feed, records the total games played and records the winning conditions of each game (separated by categories). Said like that it doesn’t seem like a lot of information but the presentation is excellent. I hope we will see more websites and game features like this in the future.

New toys and new data.

Giant Bomb, a game news/wiki website that is focused on user generated content, released their API awhile ago allowing users to tap into their unique game data. Also, I recently got my hands on a copy of the data visualization software program Tableau. Add one tablespoon of PHP and you have some nice graphics.

I’ve only been working on this today but I was able to grab all of the user review data that Giant Bomb offers. There are almost 8000 reviews total which was good because I don’t think my laptop could take much more data. The Giant Bomb API is one of the easiest I have ever worked with, you just ping URLs which is great. Their user reviews consist of mainly a time, score, text review, name of reviewer and the game reviewed.

Now I ran into a problem with Tableau, it has nothing to compare large amounts of text data. I wanted to do word counts and searches on the user review text but Tableau doesn’t have that functionality. It was even truncating the reviews and not showing them at full length. I’ll have to look into getting around that maybe I’ll find my python parser to do some string manipulations.

Anyways, I was able to run the numbers and get some nice looking graphs. In the future, once I download more of their data, I can combine different tables and make more elaborate visualization. For now I focused on user review scores in respect to length of review, date published, the most frequent reviewers and the games with the most reviews. I’ll step through each of these after the fold.

gbur-size-_descriptopm

Read the rest of this entry »

Just some information visualization sites that I have found quite entertaining recently.

Google Insight – See what the world is searching for.

Olympic Medal Count – USA kicked ass in 1904.

Movie Box Office – Let the movie waves crash over you with their millions of dollars.

Why in the growing field of game and media studies do we rely on old formats and procedures for submitting conference papers? Having common formats for printing is one thing but coming from a design perspective this seems restrictive.

I was in a project meeting today where my team members and I were discussing a paper we are writing for a conference. For my section I’m reviewing a list of games and their mechanics. Each game is rated as to how best they fit a certain player type that we are presenting in the paper. As a note to myself, and to my other team members, in the title header for each game I place a key showing which player types the game provided for the most. We have been producing a lot of documents for this project and I have been trying to work in more visually oriented note-taking because I think it helps make sense of large written documents. For this latest document another team member liked the ranking notes I made so we are going to try to add a small visual element in each game section to show how each game relates to our player types.

This paper is for a new game conference so we will have an easier time working with the formatting. But I could see other conferences going nuts because we are using images without subtext or as part of the main header. That’s what I don’t understand why is the academic conference world still text heavy? Is it because we have other venues like poster or demo sessions that can be more visually oriented? (Though CS posters might as well be papers) Perhaps people feel they must rely on text to get more of their ideas across? Or are we just stuck using systems that were created for the hard sciences and us soft sciences just have to follow along if we want to be treated with respect?

It is ironic that the project I am working on is looking at the different ways that people learn and tries to alter games in order to fit those needs. Yet the conferences where this project will be published rely on text-based means for getting information across to others, instead of a more visual one that I, personally, would prefer.

Saw this over at the information aesthetics blog: wordle. It’s a website that takes a text document or a users del.icio.us tags and creates a tag cloud. These clouds are customizable: you can remove common words, choose different fonts (~20 available), change the color scheme, and set the layout of the words (the Any Which Way options is sweet). What is really neat is that another word can fill any open space within another word. I made my picture from my del.icio.us tags, which I have not used in a long time but apparently I was very into social networks (the SN tag), Web 2.0, and conferences. The creator Jonathan Feinberg cannot release the code because IBM owns it, which kinda stinks. One thing I would like to see is an RSS option so I could take my blog and output a picture of my most common words.

This also reminded me of Chris Harrison’s work on word visualization. He has a lot of great projects over at his website.