[werewolf] Not that it's actually helpful information...

[Adam Luter <adam.luter@gmail.com>]

I was bored so I ran a werewolf simulation.

Setup:

1) The seer is known after the first night and he gives all his info.
2) Illusionist is not known but can foil a seer's read.
3) The seer picks randomly from anyone but himself, and the
illusionist picks randomly from anyone.
4) The angel protects himself on the first night, and if he is not
discovered, he protects the seer any remaining nights.
5) Wolves kill a villager at random and succeed (except if they try to
kill who the angel protects).
6) Villagers kill randomly if they have no leads, otherwise they pick
randomly from what they have from the seer (even if it ends up being
false sometimes).

(In case it isn't clear, (6) means that the werewolves vote with the
villagers, even if it means they kill one of their own).

This yeilds a 44% win rate for villagers, with the current (game 200)
setup (3 wolves to 10 villagers).  You oddly enough get 43% ties and
only 13% losses.  (A tie is if it's day and the number of wolves plus
the illusionist is exactly equal to the number of villagers minus the
illusionist).

I know there are several holes in the strategy, as I know I may have
seer's and illusionist's abilities muddled (I don't know how they are
supposed to work exactly).

One obvious problem is that I haven't got it so if the angel protects
himself, the wolves attack the seer the second night.  Because it's
actually a random chance, they would have to pick between the angel
and the seer, not knowing who the angel would pick too (and so it may
be in their best interests to just pick randomly, and let the seer
live).

The tie rate is so high because the code was written with villager
strategy in mind, so I didn't come up with any smarts for the wolves. 
Though it would seem that if they did anything -but- vote randomly
with the villagers, you would have a small push up for the villagers'
win percent (since it would be easier to detect them).  However, the
wolves would also win more often.  I do not think the wolves' wins
would (significantly) affect the 44% win rate the villagers already
have.

-Adam

p.s. no i will not setup the simulation with different strategies, but
I can change it so that the number of villagers and wolves are
different, if you'd like -- i don't want to waste any more time on it
otherwise :) .