Levels of Analysis

Some weeks ago Filip Couliuanos did a piece for Gamasutra about gameplay analysis. Its really good, read it. In very general terms, he looks at two AAA action games (Batman: Arkham Assylum and X-men Origins: Wolverine) and basically looks at what the player is doing and how much time he spends doing it. This system simply adds up the minutes the player spends doing a certain action (be it combat, watching a cinematic, solving a puzzle or moving through the game world) and then plots them in a color coded graph.

With the whole game charted you're then able to analyze its overall pacing. In the example above (the pacing graph for X-Men Origins), we can clearly see that boss battles are all piled up at the end of the game, presenting a large spike in pacing. This method is really good at presenting an overall view of a game's pacing; specially for AAA games, where there are several things to do which can be divided in clearly differentiated chunks.

However, if the chunks are not clearly differentiated (for example, if during a puzzle you also engage in combat) this turns into a problem. Couliuanos tries to solve this using the catch-all term "Special Gameplay" which would include any section where the mechanics deviate from the game's core. Although this does not reflect the true pacing of the section, it does help find where a potential spike or valley in pacing could occur.

A bigger challenge comes from games where there is not a lot of variety in what player's do through out the game. Retro, indie and older titles are a great example of this. Beat 'em ups like Streets of Rage, shooters like Contra or vehicle combat games like Icebreakers (shameless self-promotion, I know) don't have a lot of variety in player actions; however this does not mean they have a flat pacing. How then do we describe it. This will be the subject of our next entry.