Get the latest Education e-news
  • Affordances of Elliptical Learning in Arcade Video Games

    - Sébastien Hock-koon

  • Method

    As a part of my research, I trained on the Alien Vs. Predator arcade game and became a supergamer by performing a one-credit run. This was featured on the 45th episode of a TV show called Superplay Ultimate on the Nolife channel in France (Pilot, 2011). Superplay Ultimate features the performance of an expert player, or superplayer in French, on one video game. The performance is commented on by the player and the presenter. I did not use the default settings for my run, but rather increased the number of extends (i.e. lives you can win by earning points) and increased the difficulty from 4 to 8.

    Before I started this training, I had been playing this game for years and had already finished it dozens of times. I thought I knew this game but I was wrong. I trained for six months; it took me 200 hours, which included 150 hours of actual play. I kept a thorough report of each training session which included the following information:

    • The length of each run and the settings
    • Where and why I lost each life
    • What I learnt during the session about the game (new strategies, new mechanisms...)
    • My physical and mental shape (which had an influence on the way I played)

    Table 2: Evolution of runs' duration throughout training.

    A run's length is directly related to the level I reached. Finishing the game took me at least 65 minutes. This table only includes sessions where I actually tried to finish the game as opposed to sessions where I was just training. I also set aside runs where I stopped playing before losing because it was obvious I was not in a good physical and or mental shape to play properly. Even in these conditions, some runs ended very early even at the end of my training. Becoming able to succeed does not prevent a supergamer from failing. Furthermore, I performed a one-credit run only three times out of almost two hundreds. At my best, my success rate hardly exceeded fifty percent.

    However the most interesting part is that of the strategies I used to pass each section of the game. From the very beginning, I regularly passed the first ten-minute section of the game, yet my strategy for these first ten minutes has deeply changed throughout my training. The strategy reports allowed me to follow my evolution through several layers of understanding. At the beginning of my training, I found some strategies to beat the early portions of the games. These strategies appeared desperately risky and inefficient with more experience; in order to progress, I had to find and use more efficient affordances. But theses affordances only reduced the probability to lose during the first sections of the game, I was never sure to beat them.

    As I had to stop using non-optimal affordances, improving myself on this arcade game implied unlearning as much as learning. My guess is that it would require around six more months of training to finish the game with one life and at least six additional months to get close to the world record. I will explain here how I learnt and unlearnt the way shooting works in Alien Vs. Predator and how the game properties afford to players the specific learning of great games.


    Alien Vs. Predator is a 2D arcade beat'em-up game created by Capcom in 1994 on CP System II. In this game, up to three players may choose between four characters-two predators and two cyborgs. They have to stop an alien invasion in the city of San Drad. The game uses an 8-direction stick for moves, one button to jump, one button to hit and one button to fire your gun. It is possible to combine one button and one direction or move on the stick in order to do special moves. Pressing two buttons at the same time launches a powerful attack which makes one's character lose some hit points.

    There are mainly two ways to kill your enemies: you may hit them or shoot them. We will focus on the shooting mechanism. Linn Kurosawa is the character with whom I chose to perform the one-credit run. She is less resilient and harder to master than the other characters, but she is my favorite character and has the most interesting gun mechanism in terms of learning. There are several levels of understanding of how Linn's pistol works. Here are the four levels I went through.

    The first level of understanding is explicitly described in the game instructions, "Button A: Fire". Some details are given about the gun gauge:

    • When Gun Gauge is green, the gun can be fired.
    • When Gun Gauge is red, the gun can't be used. Watch out!

    This information is true for every character. Once you have played the game, you implicitly reach the second level of understanding: when you fire, the gauge empties; when you do not, it fills up. When the gauge is red, you cannot fire, which means that you are more exposed to your enemies. You also realize that the four guns do not have the same effect. Human guns fire bursts of bullets while predator guns fire one single shot that explodes on impact. The explosion may knock back several enemies at the same time. At the second level, predator guns seem to be more efficient.

    The third level of understanding highlights a strong difference between Linn's gun and the others' guns. There are two things that are not said about her gun:

    • When Gun Gauge is red, it is impossible to move.
    • Linn's Gun Gauge can only be refilled after being completely emptied.

    With the other characters, when the gauge is empty, you may still fight, and if you fire only once, the gauge will fill up automatically. Linn may only reload her gun when it is out of ammunition, and when she does, she cannot move at all, making her even more vulnerable. When I started my training, I was at this third level. I thought Linn's gun was less efficient than the others. It could not touch several enemies with the same bullet, it did not automatically reload, and it prevented Linn from moving when she had to reload it. I changed my mind when I reached the fourth level of understanding.

    This fourth level is hidden but can be found. Indeed, Linn's gun has more ammunition and does more damage than the other characters' guns. With one single gauge, Linn's pistol may kill a few enemies, when the other guns cannot even kill one. The other's guns are just emergency weapons to be used as backup while Linn's gun is a real fighting weapon. There is also a possibility to move (and fight) while reloading. To do so, the player has to jump, shoot the last bullet in the air, land and try to shoot once. It is then possible to move freely. It is tricky or even difficult to do, since you not only have to be able to do it, you have to integrate this chain into the way you fight.

    Without this tip, Linn's gun is powerful enough to take care of at least four enemies at the same time. Shooting knocks the targets back, giving a skilled player enough time to reload the gun. Moving while reloading only becomes useful against more enemies than that. But once you master it, Linn's gun is truly the most powerful weapon of the game. As far as I am concerned, I would tend to think that it is possible to finish the game just with this gun, while it is impossible to do so with the other guns.

    Affordances on learning

    What I have described may appear as a self-experiment that only highlights a personal and not generalizable experience. This is where the differences between Gibson's and Norman's affordances really become helpful. Gee (2003) and Becker (2008) use their own experience of gameplay in their research, but they focus on the learning occurring between the beginning and the end of a video game. Conversely, in an arcade video game, most of learning happens after the player finishes the game for the first time. Similarly, David Sudnow (2000) described his pursuit of the perfect game on Breakout (Atari Inc., 1978). According to the author, computers, and computer games, are the union of three older tools: television, typewriter and piano (ibid., p 23):

    Of all things exterior to the body, in its every detail [piano] most enables our digital capacities to sequence delicate actions. [...] At this genetically predestined instrument we thoroughly encircle ourselves within the finest capabilities of the organ.

    The author describes a tool whose "every detail" calls for the acquisition of the "finest capabilities of [our hands]" in order to use it properly. In other words, the possibilities of action require specific properties of the object as well as specific capacities of the subject. Thus, Sudnow deals with affordances. As we have seen, Gibson's affordances are Boolean; they exist or do not exist. If anybody is able to do something in a video game, it means that the video game and the player respectively have the properties and the capacities affording this action. If the affordance was considered impossible, it may imply a misperception of the object's properties and then open the field of possible affordances. It is also true for my own experience with Alien Vs. Predator; however it is still not generalizable. But this article deals with possibilities rather than actualities.

    On the one hand, the game itself is the result of an affordance offered by the properties of digital technologies and the capacities of game developers; it means that digital technologies may be used to create such mechanisms. Since there are many great games, there are also many developers able to create them. In addition to that, a player just cannot disobey a program's code, no more than there is "a choice about obeying gravity" (Lessig, 2006, p 110). Consequently, a video game's properties will be the same for every player and any property found in a video game could be found, or at least reproduced, in another one.

    On the other hand, every player does not have the same capacities. As a game designer, game design teacher and video game researcher I have more practice and knowledge than many players. Maybe the way I learnt is not afforded to average players. However, it is possible to reverse this problem by considering affordances of "non-learning" instead of affordances of learning. If someone as experienced as me could have spent years playing this game and not understanding it, the same affordance should be available for most of players.


comments powered by Disqus