[REVIEW] Open Setup Reviews

This forum is for discussion of individual Open Setups, including theoretical balance.
Forum rules
User avatar
mith
mith
Godfather
User avatar
User avatar
mith
Godfather
Godfather
Posts: 9053
Joined: March 27, 2002
Location: McKinney, TX
Contact:

Post Post #659 (isolation #0) » Fri Apr 13, 2018 12:28 pm

Post by mith »

In post 654, callforjudgement wrote:Add an Innocent Child who cannot be killed normally until mylo to a regular setup with no killing/protective roles and an odd player count, and unless it's a strong player, you'll actually have very little balance impact whatsoever. It's nice in terms of helping the dayplay but it doesn't change any of the nightplay dynamics.

This isn't exactly the same, but the general principles are similar.
4:12:1 Double Day + Bulletproof IC has an EV of 41.76%, way better than 4:13 Double Day (26.40%). So, how early they can be killed is kinda important. A 1-shot BP IC won't add much, and the biggest benefit is on whether they are alive at mylo, but it's not inconceivable that a partially BP IC with some other abilities could get the EV up to a reasonable number.

(This is not an endorsement of this particular setup, which has other problems. Just some drive by maths. :))
User avatar
mith
mith
Godfather
User avatar
User avatar
mith
Godfather
Godfather
Posts: 9053
Joined: March 27, 2002
Location: McKinney, TX
Contact:

Post Post #661 (isolation #1) » Sat Apr 14, 2018 3:47 pm

Post by mith »

(I actually just picked Double Day because of your first post, and because I had just gotten it working in my script. :))

I think your comparison is a poor one - yes, if you add an IC, it won't provide any benefit if you are also going from odd to even. No one designs games like that, though (and in fact the setup in question has an odd number of players).

If you added a fully BP IC, on the other hand, town would just no lynch down to 2:6:1, and now have a clear advantage (EV goes up to 45.31%). Granted, much of that advantage is at lylo (if the IC is BP only until it gets to 1:2:1, EV drops to 35.42%, and if the IC loses BP as soon as a Mafia is lynched it's 33.57%.)

I suspect the correct way to play the Sheriff would be to protect every night, and in fact protect the same player every night for a while, choosing the towniest looking player N1. If they're wrong and pick scum, they take the scum out (which is obviously a big EV boost) but risk town picking scum as Sheriff and autolosing (less good). If they pick town, though, that player is protected until targeted, and then WIFOM the next night as far as using the sniper or not. I'd guess there's a pretty good chance of a Sheriff being around at endgame, really, and the Sheriff would be even more powerful at a small count than the BP IC - for example, at 1:2:1 town EV is 1 unless the scum still have sniper since at 1:1:1 Night Sheriff can protect someone and either take out the scum or keep the town alive and shoot the scum next night.

Which isn't to say that the Sheriff can get the EV all the way up to 50%, but they might get close.
User avatar
mith
mith
Godfather
User avatar
User avatar
mith
Godfather
Godfather
Posts: 9053
Joined: March 27, 2002
Location: McKinney, TX
Contact:

Post Post #666 (isolation #2) » Wed May 09, 2018 7:38 am

Post by mith »

Just because a game
can
be run in the Open Queue doesn't mean it
should
be run in the Open Queue (or any queue). Making the first lynch flipless doesn't add anything positive to the game.

(Disagree that it's unbalanced. EV is the same whether the first lynch is flipped or not, and 3:10 White Flag is ~48%. According to the wiki, Town has won 2 out of 8, 25%, but that is not a statistically significant sample size. Also, play better town.)
User avatar
mith
mith
Godfather
User avatar
User avatar
mith
Godfather
Godfather
Posts: 9053
Joined: March 27, 2002
Location: McKinney, TX
Contact:

Post Post #668 (isolation #3) » Wed May 09, 2018 12:02 pm

Post by mith »

Your numbers are off. Blank-Town-Town could be 3:4 or 2:5; it would take three Town flips to determine the D1 lynch. Regardless, once it's potentially LyLo, town should probably just act as though they lynched Town D1 and evaluate from there (if they lynched scum D1 and it screws up their analysis, they get another chance D5).

I also disagree that it eliminates the gambit. Scum can absolutely gambit D2-D4 after a D1 scum lynch, it's just that town won't know whether something is a potential gambit or not until D5. It doesn't particularly matter, anyway - all the gambit is is an extreme form of bussing, town has to account for that as a possibility anyway. (Also, as the originator of the gambit - people get way too hung up on the gambit. It's strong if done right, but it works in those situations because towns fall so hard for it. Just knowing that the gambit is a possibility should make towns wary of it occurring, and provoke better analysis of the potential gambit players during the rest of the game, rather than just falling totally for a pairing argument. Play better, town.)

All making D1 flipless does is give the town less information D2 (and D3/D4 unless they have flipped a scum). IMO, that is not a positive addition. YMMV.
Post Reply