Live Action Mafia

Game 2?
Page 1 of 1

Author:  jasonye [ Sun Dec 31, 2017 10:31 pm ]
Post subject:  Game 2?

Does anyone else want to play more of this over the rest of winter break/IAP? I would definitely do so.

Author:  laberca [ Sun Dec 31, 2017 10:32 pm ]
Post subject:  Re: Game 2?

I'm up for it!

Author:  ksedlar [ Sun Dec 31, 2017 11:03 pm ]
Post subject:  Re: Game 2?

I would do this. As someone who got very frustrated last game, I found this game pretty fun and not frustrating.

Author:  Daniel Grazian [ Sun Dec 31, 2017 11:05 pm ]
Post subject:  Re: Game 2?

I would 100% play again!

Author:  jamb [ Mon Jan 01, 2018 1:40 am ]
Post subject:  Re: Game 2?

Would do over IAP... I have kinda dropped the ball on all my other winter break commitments and should catch up on those now, so not winter break.

Author:  achester [ Mon Jan 01, 2018 1:42 am ]
Post subject:  Re: Game 2?

Would do over IAP... I have kinda dropped the ball on all my other winter break commitments and should catch up on those now, but would be temptable into playing over winter break anyway.

Author:  brunnerj [ Mon Jan 01, 2018 3:35 am ]
Post subject:  Re: Game 2?

I would play.

I would like to shift the game more toward mafia on the mafia-codenames spectrum: I found that alignments came together a little to fast, and would be more fun with less alignment certainty.

Ways to do this are changing total names/fraction of names that are true names (less total names makes the game more about finding players alignments, higher fraction of true names also does this.)
and changing witness/espionage (reducing/removing witness, or making the witness/espionage parity check not work makes the game more mafia and less codenames)

Also, we considered but didn't decide in time about making team's daily hp public/each player of a team knowing a random target's hp.

Author:  ksedlar [ Mon Jan 01, 2018 6:07 am ]
Post subject:  Re: Game 2?

I think the main factors causing this game to be less mafia-esque than last game were:
#1: Shots counted more in the beginning. I think the number of names you want to shoot at the beginning is pretty constant; like, it doesn't seem like initial clue size is actually that dependent on number of players (or at least it isn't in principle). I'm pretty sure that last game, with 9 shots, green only had like 6 things we wanted to shoot at day 1, so it was no trouble to spend 2 shots on implanting deep spies in red/blue. This game, we definitely had at least 6 things we wanted to shoot day 1. So actually spending a shot to shoot for another team was just too costly. Even shooting at a word that had some chance of being clued by python, some chance of being clued by Jason (i.e. CHINA) was too costly.
#2: People don't know when they're being witnessed, and can be witnessed without their consent. This is a big mechanic change from last game, which makes it hard to infiltrate effectively, because any time you might be getting benefit from lying about your shot, you might also be outing yourself WITHOUT EVEN KNOWING. You also might just get surprise witnessed, without knowing ahead of time.
#3: Hard mechanical evidence came in way earlier, due to witness/espionage parity checks. If you only have 3 people claiming to have a witness or espionage, then modulo inactivity, you get 3 proven innocents, and 4 people forced into a hard beef. And this can start from the very beginning of game! Last game, proven innocence only started once people started getting hit a lot in a row / people started dying. Also, it could only prove people innocent, not get people into beefs.
#4: Game was shorter / smaller, so people needed to shoot for real earlier.

These are definitely ordered, btw. I also think that while fixing #2 and #3 would make faking easier early on, they wouldn't actually incentivize making shots for another team. IMO, that's the thing that makes deep infiltration possible: when there's a real chance that making a shot for another team might be worth it. And idk if there's a good way to do that other than "make each individual shot less valuable, AKA increase the total number of shots."

I think that without incentive for faking shots, alignments are always going to come together pretty early, just from the perspective of "look at who is shooting at you." Relevantly, it turned out that "just look at who is shooting at you" was a better algorithm than "try hard to infiltrate another team and figure out what claims they're getting," from the corwind vs cmcclena perspective in this game.

Author:  brunnerj [ Tue Jan 02, 2018 3:10 pm ]
Post subject:  Re: Game 2?

I agree that the fewer early game shots did make it much less worthwhile to shoot for another team.

I think #3 is the most important thing to change, but #2 is also important. I don't think it is necessary to be making shots for other teams to infiltrate if the rules are changed some, and you can at least mess with alignment certainty without making shots for them.

Looking at who is shooting you does not distinguish your team from the team you are shooting at, so it only halfway gives you alignments.

Author:  achester [ Tue Jan 02, 2018 4:23 pm ]
Post subject:  Re: Game 2?

0. Less total names and more True Names: this seems to make the game more mafia only by making it less Codenames, which I'dn't like.

1. Shots counting more in the beginning: sure, adding a few more shots (with the same total clue throughput) seems good, for letting teams spend some of them on plots.

3. Removing the parity check: sure. A minimal change would be to say that each person gets witness or espionage independently with probability 1/2, or even that subject to that each team gets the same total number of witnessings. For a bigger change, see below.

4. Let's make game bigger again by increasing the total number of names. :)

5. Effectiveness of observing who's shooting at you: we could fix this and the color-cycle asymmetry problem simultaneously by just giving each person words of all three colors in addition to bystanders, so people of all three teams could plausibly want to shoot at you; you'd still care _what_ words they're shooting at you for, but that's at least Codenamishly interesting. (Or we could just get rid of the third team; I don't think we need them now that we have the True Names thing.)

2. Changing witness/espionage: sure. Removing witness and reworking espionage also goes with a proposal I have to make the Codenames clue numbers more meaningful (which also reduces the incentive to use that as a side information channel):

It's sad that there's nothing really making spymasters state their clue number accurately; the line between deciding boundary cases and just using it as an extra channel independent of the clue is too thin. What if we say that each spymaster actually records a list of the words they mean to clue, and that has some mechanical effect? For instance:
-(Replacing espionage): when you shoot at a word, you're told a(ll) clue(s) that clued it.
-(Incentive to not include too-borderline words): when a team shoots at the last word for a clue, they get a prize, like learning that they've done so, or being told who shot at each of the clued words.

Author:  ksedlar [ Wed Jan 03, 2018 1:30 am ]
Post subject:  Re: Game 2?

0: I agree that removing the portion of fake names only seems to decrease the Code Names aspect. In general I think that random shooting shouldn't be very productive; if it is, then again, this makes little incentive to shoot for another team in an infiltration effort.

3: I like the idea of probabilistically determining the espionage/witness ratio, but then making it the same for all teams.

I think that if "when you shoot at a word, you're told a(ll) clue(s) that clued it," actually disincentivizes honest clue numbers. Specifically, the first clue will probably be of size n, where n is the number of true names you are supposed to shoot at. This would make if so that you can always tell whether you shot a true name or not, which is particularly valuable in the case where you randomly shoot at a true name without the spymaster intending it.

The problem with incentivizing prizes for shooting the last clue seems to be either that a good prize disincentivizes big clues (which are fun), not so good prize isn't much incentive at all. I'm currently thinking that the proposed prizes are "not much incentive at all."

On the other hand, I also really enjoy asking spymasters about information via parity checks. :) But I agree that there should at least be a minor check to make sure that spymasters have *some* explanation of how a clue relates to a word, like writing the relationships down. That's why I only asked python to move the number up or down by 1 at most; I figured there's always some sketchy relation that could be optionally added or removed, while keeping true to the principle that the number should *actually* clue the number of true names related to it.

Page 1 of 1 All times are UTC - 5 hours
Powered by phpBB® Forum Software © phpBB Group