Original post (with no pictures, but some links) http://3dtotalgames.com/abstract-concepts-are-entitled-to-a-...
Greetings all. The posts this week are likely to be a little short. This is because I have, in my infinite wisdom, scheduled playtests 10am to 5pm every day. Given a standard 9-5 working day that leaves me with one hour for everything else, so I won't have as much time to write this week. I'll compensate for this by being highly disciplined with my time and launching straight into the relevant topic each day, with no preamble. Starting tomorrow apparently.
This morning, I've been thinking about when the game gets to take its turn. In a lot of games there are actors represented by the game who have things to do. The wildlings in Game of Thrones need their chance to invade, the zombies in Zombies!!! crave human flesh, arguably even the city council in Monopoly take action once or twice a game. It's not quite true of every game, but it comes up a lot.
The games I'm developing follow this pattern too, Assassins has no challenge if the guards do not react to player movement, Wizard Academy would be pointless if fires didn't spread and demons didn't rampage and finally Murder TV would be somehow incomplete if a faceless director didn't sometimes hit the button marked "burninate". Deciding when these actions should take place isn't trivial.
A very neat solution is for them to occur at the end of each players turn, this is fair to all players and will happen predictably allowing players to develop deeper strategies. However it also means the game will go several times per player, worse, the number of times the game acts before a given player gets another turn will change based on how many people play. While this can be a pain to balance it can be a nice feature in a cooperative (or mostly cooperative) game, in that it can counteract the advantages obtained by having a greater number of players working together. On the other hand if the game's actions represent entities that are roughly as powerful and competent as the players then they probably shouldn't get multiple turns to each individual player turn.
Options for cutting this down each come with their own problems. Treating the game like it is a player and assigning it a place in the turn order can produce a balance issue. The player sitting to the left of the game always gets the best of any opportunities it produces (and the games actions should produce interesting choices where possible). Moving the game's position in the turn order around can cause other problems Whenever it moves it is guaranteed that either one player will get two turns in a row without the game acting, or the game will get two turns in a row without the player acting. On the other hand if this is predictable players can strategise around that feature and it might add depth to the game.
Another option is presented by the game actions sometimes occuring in response to players actions, such as in games where searching will occasionally produce a card indicating that an antagonist acts. This makes the elements controlled by the game more unpredictable, but requires little or no bookeeping on behalf of the players and overcomes turn order issues.
I don't think there is a "one true way" for a game to take its turn. Like everything else in design there are a series of trade-offs. Any solution will have some benefits and some costs, the trick is to find a solution that's benefits are in line with the core of the game and that has costs that are negligible in the wider context of the game. More and more I feel like the art of designing a game is similar to finding a good strategy while playing a game, the best moves cost a little and achieve several goals at once.