There's no such thing as an overeager vig

[Kelly Chen]

In general I'd agree that vigs should kill. Just as in general I'd agree that the town should lynch.

(Side note: The whole "random deaths are good for the town" thing is actually a good argument for suicide vigging, too - it shows that gaining a "Free" lynch is, in fact, always a good thing.)

Is the vig's special knowledge about his own alignment unimportant?

Even if so, I don't think this argument does support that. The town has to weigh the possibility that the vig really is protown, in which case he can kill off a lot more random people than just himself.

[Kelly Chen]

if your odds are better then average (say, you have a strong hunch person X is scum

I don't think strong hunches make an average player's odds better than average. According to the statistics, people in C9 games lynch very slightly

worse

than random. They could improve their game by using dice tags. Apparently, most of the things the average player thinks are scum tells, simply aren't. I don't think people here have assimilated this insight yet.

I guess you're referring to D1 lynches in newbie games. I don't know that I would generalize that to average players in any situation.

Not sure what all this means for vigs. I've never been a vig, but I'd probably be quite aggressive.

Are you an average player? Roll a die. If you get yourself, roll again. Vig that person, and as many other people as the mod allows. You're not vigging yourself, so your odds are better than the town's lynches.

[Kelly Chen]

If you're making the kills totally randomly (i.e. it's not even a vig doing it), I don't understand how the values of C and D make any difference.

How do you find that with random kills, the town should win 75% of the time with a 3:1 situation? It seems clear to me that this is still 50/50.

[Kelly Chen]

EBWODP

Looking at the math, I think C and D are a distraction. If you know A and B and take the setup to be in balance, you can determine whether another situation is similarly balanced just by the new ratio of A to B.

Say we start with 9 town and 3 scum. To preserve this same balance we need to end up with three times as many town as scum.

(9 / 12)((9 - 1) / 3) + (3 / 12)(9 / (3 - 1))
(3 / 4)(8 / 3) + (1 / 4)(9 / 2)
2 + 1.125
3.125 new ratio

Seems to work, although I don't understand why.

3 townies and 2 scum (ratio is 1.5):
new ratio seems to be 1.8.

2 townies and 1 scum (ratio is 2):
division by zero error.

This doesn't seem to be related to the obvious fact that the town has to lynch (even randomly) if it wants to win, since even if we found that random lynches generally worsen the town's situation, the town would still have no choice but to try.