I was looking through some notes I took for my 3rd year project today and I found this quote:
One of our undergraduates played the computer game of "Doom" while wired up to several physiological sensors detecting signals that change with affective responses. We saw minor changes in several of the signals when he "found the rocket launcher" or when he "was killed." However, the biggest response we found, significantly higher than any other in the game, occurred not during a stressful life-threatening battle, but at a more surprising moment: when the software failed to work properly. None of the violent events in the game aroused the player as much as the software problem.
And underneath I had written the following note to myself: "Should assume that players believe/are invested in the game world? (Games have evolved a lot since then)"
So apparently I was thinking about this very same thing a few weeks ago, the idea just got burried in with a load of other irrelevant stuff.