GAMBIT, the Singapore MIT Game Lab, have developed a series of games based on "research questions" from game academics. The inspiration behind Pierre: Insanity Inspired is this question from Jesper Juul: "How does [sic] different ways of communicating failure influence the player’s experience and performance?"
In Pierre, you control a little critter on a rotating circle, divided into three segments with different illustrations in each segment. Thingies fall from the sky, and if a thingie with the same illustration as a segment happens to be over that segment at the moment, and you move through it, part of the illustration lights up. You complete the level by lighting up all three illustrations. You can move clockwise and counterclockwise about the circle, and can jump; spiky balls also fall from the sky, cutting you off from thingies. You can either wait from them to go away or jump over them. Later on other enemies, such as spiky balls that whoosh through space or shmup-like attacks of lines of spikey balls appear.
If this was all to the game, I probably wouldn't be writing about it; it's mildly entertaining at best, not terribly interesting as a game qua game, in other words.
What's more interesting is its engagement with Juul's question; the game imparts 'failure' in a variety of ways, from the mild to the extremely rude. On the mild end, when you run into a spikey ball, your character blinks rapidly, Mario-like; conversely, when you trigger a thingie at an inappropriate moment, an image of Pierre appears from one screen corner, tongue stuck out, and says something like "You're the worst player ever!". And if you fail a level, you get a screen like the image above, with a prominent "F", while sardonic laughter plays and Pierre says "Loooooserrrr!"
What's missing here, I think, are the signifiers of success, which exist but are far less prominent; a completed symbol glows, Pierre occasional shows up with a smile on his face, and you get a grade for the level of, say, B. The negatives are far more impactful than the positives, the reverse of most games -- I'm thinking of Ash Ketchum. Yes! I got the Volcano Badge! Tada!
Before you begin to play, you're warned that information is sent back to the server to record your experience of play for research purposes. I imagine the most important datum is where in the game you stop playing, and what event triggered that. But the actual research value of this is debatable, since "stopping playing" is more probably triggered by annoyance over time than by a single event. Still, it's an interesting approach, and the rudeness of the feedback is in its own way amusing.
Seems mildly sociopathic, to
Seems mildly sociopathic, to me? And it'd seem mildly sociopathic to engage it, since that'd involve turning off any human connection to him?
Philosopher Gamer Blog
Value scales.
A little something that could be seen as a little problem for the scientific exactitude of this is how we don't all use the same way to mark a student's work. In some country, that letter system is used, but in mine, we're marked on twenty points most of the time, so a letter has no real meaning to me, even if I perfectly well know what it does mean.
It's like that thing I read in a book ; if you live in a tropical country and read a poem about the coldness of Norway, well, you might know what cold is, for sure, but you've never experienced it in any real, natural way.
Praise and Statistics
The interesting thing you may not have noticed is that for about 10 seconds after you complete one of the shapes, getting hit by an object elicits a positive response, like "You're doing alright", and "Don't give up". Dying within this time frame leads to a positive, supportive death scene, rather than a terrible humiliating one.
You could argue that that time needs to be longer, since I only noticed it entirely by accident, and only on the last level which took me about 4 tries to complete. It's not much, but it could be enough to keep someone playing the game for that one last try.
As to people leaving because of time rather than annoyance, you would expect the time data to have a fairly well defined distribution, perhaps exponential with peaks around the 5 or 10 minute mark. Since the game can be completed in about 10 minutes, the effect of time should be fairly easy to control for. They would effectively "subtract out" the people who obviously quit for time, giving them more control over the data. This is a standard statistical practice.