Should Honest Men, reveal their roles to Jailor?

[ZZorange]

Guys the thread is titled should Honest Men reveal their roles. Literally everyone who has posted here is a known liar and therefore the thread doesn't apply to them. Hence the line at the end

"Now of course if you're a regular player of Mafia who is able to lie, this thread does not apply to you. As you can claim town as evil."

However this does let me know who read the full post so,

Props to LagAttack for reading the whole post!

However LagAttack obviously mathmatically claiming will result in less failorage but it'll be a constant. With the not claim strategy first off it's the best strategy against kidnapper scum, which is what you want as a townie, because fuck the mafia. "Eat a dick kidnapper" is the best response. Also the not claiming strategy has the potential to become better overtime, as opposed to claiming which as stated stays constant.

I don't have to be a rocket surgeon to say that Mafia does not have many new players, and the few casual players it has will learn overtime to not jail a man who is Town 94% of his games.

Keep in mind during all of this we are ignoring any spite jailors in the equation because spite jailors will execute regardless of role as most spite jailors are long timers who know that they are jailing someone who is 94% town but they execute anyway because they hate said player.

Overtime given enough samples the pool of failors will gradually shrink with the no claim strategy and become the optimal strategy. Which I think is something you missed in your initial evaluation LagAttack the future prospects of the strategy.

[yzb25]

Guys the thread is titled should Honest Men reveal their roles. Literally everyone who has posted here is a known liar and therefore the thread doesn't apply to them. Hence the line at the end

"Now of course if you're a regular player of Mafia who is able to lie, this thread does not apply to you. As you can claim town as evil."

However this does let me know who read the full post so,

Props to LagAttack for reading the whole post!

However LagAttack obviously mathmatically claiming will result in less failorage but it'll be a constant. With the not claim strategy first off it's the best strategy against kidnapper scum, which is what you want as a townie, because fuck the mafia. "Eat a dick kidnapper" is the best response. Also the not claiming strategy has the potential to become better overtime, as opposed to claiming which as stated stays constant.

I don't have to be a rocket surgeon to say that Mafia does not have many new players, and the few casual players it has will learn overtime to not jail a man who is Town 94% of his games.

Keep in mind during all of this we are ignoring any spite jailors in the equation because spite jailors will execute regardless of role as most spite jailors are long timers who know that they are jailing someone who is 94% town but they execute anyway because they hate said player.

Overtime given enough samples the pool of failors will gradually shrink with the no claim strategy and become the optimal strategy. Which I think is something you missed in your initial evaluation LagAttack the future prospects of the strategy.

hmmm

consider the highly theoretical possibility certain strategies net you more spite jailors. This can make a strategy bad even if on paper perfectly informed players never choose to imprison you unless they're kidnapper or have information / evidence that specifically suggests you're scum.

also, I think lag's math not only implies more failorage, but that the increased failorage actually offsets the gain in avoided kidnapper executions.

have you considered having a brief copy-paste explanation of your strategy ready for everyone who jails you? I am certain that'd make everyone totally content with your strategy! An honest man values transparency, after all.

[yzb25]

have you considered having a brief copy-paste explanation of your strategy ready for everyone who jails you? I am certain that'd make everyone totally content with your strategy! An honest man values transparency, after all.

or you could just copy-paste a link to this thread. If they are truly worthy of you they'll be able to read it in <30 seconds, including lag's math.

[Oberon]

or you could just copy-paste a link to this thread. If they are truly worthy of you they'll be able to read it in <30 seconds, including lag's math.

Yes

[Lumi]

Guys the thread is titled should Honest Men reveal their roles. Literally everyone who has posted here is a known liar and therefore the thread doesn't apply to them. Hence the line at the end

"Now of course if you're a regular player of Mafia who is able to lie, this thread does not apply to you. As you can claim town as evil."

However this does let me know who read the full post so,

Props to LagAttack for reading the whole post!

However LagAttack obviously mathmatically claiming will result in less failorage but it'll be a constant. With the not claim strategy first off it's the best strategy against kidnapper scum, which is what you want as a townie, because fuck the mafia. "Eat a dick kidnapper" is the best response. Also the not claiming strategy has the potential to become better overtime, as opposed to claiming which as stated stays constant.

I don't have to be a rocket surgeon to say that Mafia does not have many new players, and the few casual players it has will learn overtime to not jail a man who is Town 94% of his games.

Keep in mind during all of this we are ignoring any spite jailors in the equation because spite jailors will execute regardless of role as most spite jailors are long timers who know that they are jailing someone who is 94% town but they execute anyway because they hate said player.

Overtime given enough samples the pool of failors will gradually shrink with the no claim strategy and become the optimal strategy. Which I think is something you missed in your initial evaluation LagAttack the future prospects of the strategy.

This is ignored because if they are going to execute you either way, then it has no impact on the inequality determining which strategy is better - the terms cancel out since your actions can't change the outcome. But I will include it in the following analysis as it does become relevant when personal emotions are taken into consideration

What you are describing here are the reasons I suspected you value this strategy: spite and personal emotions

A spiteful strategy rarely performs better than a non-spiteful strategy (within the timeframe of a single game - more on that later) so my analysis here will explain why someone playing spitefully might view the "Silent" strategy as better than it really is.



Now the reason why this term was originally omitted is because delta should be zero - your actions will not sway the vengeful Jailor one way or the other, therefore there is no utility to be gained. But when taking into account the personal emotive utility lost from claiming a role to a town jailor just for them to execute you anyway - delta gains a positive value, making the silent strategy more appealing.

On top of that:

• Beta measures the utility gained from effectively telling a regular player that is a kidnapper to "eat a dick" - while this has arguably no in-game utility, it does have positive emotional utility as it feels good to tell a kidnapper off. So when taking into account personal emotions, the silent strategy gains a larger beta value.
• Gamma measures the utility gained from staying silent with a new player that is a kidnapper - again, this value likely increases (although not by nearly as much since calling out a kidnapper is more likely to make them think you're town rather than scum staying silent - so less chance they spare you) the gamma value - again pushing the balance in silent strategy's favor
• Alpha measures the utility gained from claiming your role to a new town jailor versus staying silent - and again when spite is taken into account, the value of alpha is lower as if they execute you, you get the satisfaction of being right and punishing the Jailor for jailing you as you made the Jailor into a Failor.

So when personal emotions are taken into account, all of the variables are shifted in the silent strategy's favor. Shifting the strategy from NOT APPROVED to VIABLE

But that is operating under the assumption that personal emotions and personal utility are just as valuable as actual tangible in-game advantages. Applying some discount factor to personal emotional utility could easily make the strategy switch back to NOT APPROVED from a third party's perspective.

This is a good point. Let's pull up the original utility inequality again for reference:



What you are describing is a long term decrease of N and a long term increase of R. Which as seen in the second inequality, obviously makes the strategy more and more viable as the R/N coefficient becomes larger and larger.

This strategy can therefore be re-labeled as: META-DEPENDENT, NOT APPROVED IN CURRENT META IF that were the end of the story, but it's not...

Another example of a meta-dependent but currently ill-advised strategy is what I call the Town-Aligned Survivor.
The idea is that you claim survivor early on but play in a very pro-town fashion as survivor, you scum hunt well, vote with town, help with vca, etc

In the current meta, this strategy is likely to get you killed anyway despite your helping town, since as soon as town hits the smallest road-block they'll just lynch you, despite you being effective a NI townie.

But by picking the same name every game, preferring survivor, and then always playing very pro-town as survivor - the meta can be shifted in order to make this strategy viable.

The key difference though between these two strategies can be summarized in one word: externalities

In the survivor strategy, the factions that have their win-rate damaged by the off-meta strategy are the survivor themself, and scum. But the survivor has no obligation to help scum, and as they are a solo team, they also have even little obligation to help themselves - free to play however they wish. So the strategy has no externalities.

Whereas the Silent, but Honest Prisoner strategy has the major externality of severely hurting town every time a town Jailor is forced to failor. The failor made the game theory optimal decision to execute - and it only turned out to be wrong due to their imperfect information - not knowing your meta.

This significant negative externality degrades the quality of the game for other innocent town players. So that alone is cause to push the strategy back into the NOT APPROVED category.

But even if it were a zero-externality strategy, it's long term viability would still be dubious - there are thousands of players that play this game, the rate at which players forget information is almost certainly larger than the rate at which you can teach them your meta - therefore getting R/N to be large enough to flip the balance on this strategy is likely impossible as information is lost to the entropy of our forgetful and impermanent minds.

[theoneceko]

What you are describing here are the reasons I suspected you value this strategy: spite and personal emotions

hewwo !

[Lumi]

hewwo !

My comment taken out of context sounds more critical that it really is.

There are a number of types of problems in game theory in which a spiteful strategy is actually an optimal strategy. The classical example of this is the iterated prisoner's dilemma. The Honest Prisoner's dilemma (no relation to the prisoner's dilemma) has a number of parallels to the iterated prisoner's dilemma - some of which ZZorange pointed out and which I touched on in my response post.

Because of these parallels a spiteful strategy cannot be ruled as suboptimal without additional analysis. (Although in this case my additional analysis still points to it being sub-optimal)


Additionally how personal emotions affect a player's utility function in a game like mafia is a very interesting topic - I'll probably make a thread about it at some point.

[theoneceko]

My comment taken out of context sounds more critical that it really is.

There are a number of types of problems in game theory in which a spiteful strategy is actually an optimal strategy. The classical example of this is the iterated prisoner's dilemma. The Honest Prisoner's dilemma (no relation to the prisoner's dilemma) has a number of parallels to the iterated prisoner's dilemma - some of which ZZorange pointed out and which I touched on in my response post.

Because of these parallels a spiteful strategy cannot be ruled as suboptimal without additional analysis. (Although in this case my additional analysis still points to it being sub-optimal)


Additionally how personal emotions affect a player's utility function in a game like mafia is a very interesting topic - I'll probably make a thread about it at some point.

uwu
hewwo

my bren fcking broke wen i saw "24t + s " ;W;

[theoneceko]

i'd lov to see how "prsonal emtions" ffect how i ply ufu

[Oberon]

My comment taken out of context sounds more critical that it really is.

There are a number of types of problems in game theory in which a spiteful strategy is actually an optimal strategy. The classical example of this is the iterated prisoner's dilemma. The Honest Prisoner's dilemma (no relation to the prisoner's dilemma) has a number of parallels to the iterated prisoner's dilemma - some of which ZZorange pointed out and which I touched on in my response post.

Because of these parallels a spiteful strategy cannot be ruled as suboptimal without additional analysis. (Although in this case my additional analysis still points to it being sub-optimal)


Additionally how personal emotions affect a player's utility function in a game like mafia is a very interesting topic - I'll probably make a thread about it at some point.

I believe that player's emotions cloud their judgment. It either turns your utility function into something else entirely (as the player's goal is no longer to win the game), or it gives a massive penalty. I imagine that it gives results that are worse than if you'd randomly picked your night action.