The player did not behave as expected.
The playtest did not go as expected. The player elected to resort to violence almost immediately, taking the knife from the house on the first day. They tried to cut down one of the bushes outside their house, an action that I declined, and then went on to murder the convenience store clerk. They then took the gun and used it to murder the only townsperson with the clowns. Because no one but the clowns witnessed these crimes, the player was not immediately arrested. The player then proceeded to kill both clowns and went home. They asked if they could kill themselves, an eventuality I had not anticipated. I declined their request. The police officer arrived at their home, and they murdered the police officer (whom I had neglected to provide with a weapon). As there were no more possible actions, I ended the game at this point.
The player killed every other character in the game on the first day.
The most striking lesson from this playtest was undoubtedly the tendency of players to test the limits of your system. I expected the player to behave within the bounds of normal social behavior, and expected them to resort to violence only in the later stages of the game, if ever. In retrospect this was an unbelievably foolhardy assumption. That being said, although I was taken by surprise and unsure of how the system would respond in a couple of cases (killing the only spectator in front of the clowns, for example; do the clowns call the police? I decided no), overall I was able to extrapolate a response from the rules I had laid down for the world. The fact that an unexpected narrative emerged from the system I had created feels like something of a triumph, and, in my opinion, led to this iteration being feeling much more “alive” than the last where the player’s actions were much more severely limited.
I also realized a short way into the playtest that I had not given ample consideration to the tendency of players to experiment with any interactive game object they are presented with. I expected the player to take an initially non-violent approach, but the only obviously interactive objects I provided them with were implements of violence. If I were to do a third iteration on this game, I would be tempted to hide the weapon objects from the player somewhat and provide them with interactive objects that might encourage them to explore other avenues.
It is also worth noting that the bed object did not give any indications as to its interactivity, and although this issue did not impact the playtest, it would likely have proved to be a stumbling block for any player that did attempt a nonviolent approach.