According to certain player typologies, players report preferring some game dynamics more than others [1]. However, it is unclear whether these self-reported preferences show carryover effects during actual playing. For instance, do players who report liking dynamics of killing and shooting show positive emotions during playing games that incorporate such elements? We recruited active gamers (N = 24) and divided them into two groups: those who preferred and those who disliked aggressive game dynamics. In the experiment, both groups played a first person shooter game as well as watched a video of someone else playing the game. We recorded psychophysiological responses for facial muscle activation, electrodermal activity and heartbeat to explore the participants’ emotional valence and arousal states during playing and watching. The results indicate that the groups did not differ in their emotional responses, suggesting that self-reported playing preferences may not be an accurate way to measure actual emotional responses during playing. The results also illustrate that gaming seems to affect players in a similar way regardless of their self-reported likes and preferences. Furthermore, the results showed that actual playing induces more positive and negative emotions (as indexed by smiling and frowning) than watching someone else play, and is emotionally more arousing.