Recording the input was simultaneously a straightforward and tedious venture. I say tedious mostly in reference to my struggles finding software that would accomodate my initial plan for this project, which was to play The Legend of Zelda: Wind Waker HD on the WiiU while recording the GamePad input on my laptop or tablet. Unfortunately, despite the seeming simplicity of the task and my long and stubborn efforts, the amount of software support for something of such niche relevance as archiving strings of GamePad input was simply not great enough. I'd need to code a program myself- and I am no programmer. It worked out well enough in the end with Bastion, though. The prime reason I was so set on Zelda was the iconic and historically consistent button mapping, such as the sword being perpetually mapped to B, but as I reflected on alternatives I realized that action and rpg gaming already had a significant set of iconic actions which I could capitalize on. Weapon swinging, potion drinking, dodge rolling- all of these actions tether a massive proportion of games on the market today, and Bastion suited my purposes wonderfully.
The recording itself was straightforward enough: I simply fired up my input recording software, the cheesy yet effective Key Recorder Pro, and allowed it to amass my desired archive of information as I happily played the game. Not only did it record keyboard input and mouse clicks, but my mouse movements and the length of the delays separating each event as well. It was a piece of software that was primarily intentended for recording mouse movements in order to effectively run a replay of those movements and automatically complete tedious tasks without necessitating much of the user's input. I had intended to use the recorded mouse trajectory as a tool to effectively draw a map my clicks and click locations during gameplay by opening a full screen in paint or somesuch software, but there was a unfortunate bug in the software where, tragically unbeknownst to me, every attempt I made to save a mouse recording had silently failed. I was still able to retrieve the button input easily enough, but unfortunately this otherwise anticipated aspect of my project will have to wait for a later date.
Given the massive number of events recorded by the software throughout the course of the gameplay and the necessity of manually process all of the data I wanted to use with Voyant, I ended up choosing two gameplay hours worth of recorded input to process and use for my analysis for the site. The two hours of gameplay was daunting enough to format properly, as I needed to do some fancy maneuvering with Microsoft Word and its "replace all x with y" function in order to mold the raw data into something that could provide me with statistics and communicate properly with Voyant. Using Voyant, I was able to take the huge sample of recorded input and show the comparative frequency with which any of the buttons were pressed through wordmaps and graphs. These stand as one of the primary tools in my textual analysis.