You are viewing the Wiki. To play the game, visit the forum.

Balancing Possibilities Versus Probabilities

From MafiaWiki
Jump to: navigation, search


Original Publication: March 15, 2017 by Mastina.

Original Thread.


This has always been a long-standing belief of mine, that the game of mafia revolves largely around a simple issue: what is likely, versus what is unlikely.

Basic Thesis:

Mafia thrive on possibility; town thrive on probability.

Everything in the game traces back to this simple belief.

A mafia player thrives on making players think of the unlikely: "What if this player who was investigated by the cop is actually a Godfather?" "What if the cop claim after the lynch of a Godfather is actually scum fakeclaiming?" These are things that are of course theoretically possible, and which vastly favor the scum because they are probably not actually true.

A town player thrives on making reasonably probable conclusions: No, the innocent is not on a Godfather; no, the cop claiming after a Godfather lynch is not scum fakeclaiming. There remains the possibility of being wrong in deductions like this...but sheer statistical numbers are on your side to hold these beliefs, as 90+% of the time you will be right to have held these stances.

The Art of Misleading:

The method a mafia player uses to convince town to follow them onto something which they know is wrong is essentially exaggeration--they run through a logical process, and play up the percentages of their conclusion, emphasizing how it is so much more likely than it realistically should be. This is why so many people advise against town players lying about the strength behind their reads; it is quite literally a scum tactic.

The mafiate will explain how they reached their conclusion, pointing out the links and chaining them together as to how that conclusion could be formed. The strength of these chains is often inherently weaker than what it should be, because the mafia player already knows what they are talking about is wrong. They will devalue things which are inconvenient to their narrative and play up things which loosely support their narrative, even if these things are a stretch. (Which is, again, why as town you are recommended to never use something you do not honestly believe to be true.)

This can be used for anything: misapplication of mechanics, misapplication of roles, justifying a bad townread on a scumbuddy, justifying a bad scumread on a player they know is town. All players, no matter their playstyle, do this as scum. Wallposters, spamposters, logical masterminds, gut players. They will point out things which are theoretically something which could be thought true, and rely on people not pointing out how unlikely those things are to be true.

The Art of Balance:

The way I view the game, my reads are in a neverending state of constant flux: I doubt everything, because I think of every possibility. This is convenient for me as scum; it allows me to prepare against any town action and also lets me fabricate whatever reads I so desire. (As seen above.) As town, if I let these possibilities go unchecked and control me, I would never be able to reliably hunt (yet alone, catch) scum.

As a result, I channel the possibilities with a simple, yet versatile, question: "Which is more likely?" Which is the most likely to be the case? This is closely tied to Occam's Razor, in that often the most likely possibility is the simplest one requiring the least number of assumptions.

It is okay to have a brain which is hardwired to doubt everything--this is actually a good thing! You should constantly be challenging your assumptions, you should constantly be reevaluating, you should consistently ask "Am I actually right?"

...However, you shouldn't be afraid to answer that question with a "Yes". "Yes, I actually think I am right." You can, and should, pick and choose what to focus on, what to believe in, which assumptions are valid, and which assumptions can be safely discarded as incredibly unlikely.

Nothing Is Ever Certain!

You will never have a scumteam which perfectly fits all the data on hand. There will always be something which makes the team not be perfect. There will always be a valid point which keeps a scumteam from being flawless.

Town players (even skilled ones) will say scummy shit. Scum players will say shit which sounds town. This is a basic truth, universal to any game you will play. Which is stronger depends on the skill of the player in question as an alignment, not to mention the circumstances behind what they are saying. (A statement that is incredibly scummy in one game might be the beacon of towniness in another; similarly, a statement scummy from one player might be a huge towntell coming from a different player.)

However, regardless of how crappy the scum are, or how elite the scum are, or how good the town are, or how shitty the town are: both sides will always contain at least some of both. As a result, when weighing evidence, every player will have a possibility of being either alignment.

The job of a town player isn't to see every possibility--the job of a town player is to figure out which possibility is the most probable.

Application of Knowledge:

It is one thing to know how to weigh possibilities versus probabilities in terms of honestly evaluating the strength of your reads and your assumptions in the game. This is something most players are actually already good at, to at least some degree. Most of what I have said above isn't anything revolutionary; it's simply something we don't often see said.

What is a bit harder to understand is how to properly use it effectively.

Set Guidelines

It helps to establish some basic boundaries. For instance, it is often not hard to guess the approximate number of scum in the game. A good, solid guideline I have is to set no more than double the number of scum in the game as scumreads--ideally, no more than 1.5 times the number of scum. (So in a 2-scum game, 4 max, ideally 3. 4-scum, 8 max, ideally 6.)

Use These Terms

You don't have to explain using the exact language that I use here. But it helps to use the conceptual language here, and lots of us do this already. When we see someone point out something which we find to basically be impossible, we'll call them out on their bullshit and clearly tell them why we feel their theory, while possible, is incredibly improbable.

Be Honest

Explain your process and how you came to what you have. This is helpful for two reasons. One, it can help you reevaluate your own read. By typing out why someone is null, you might gain a read on them. Or by typing out the reasons for someone being town/scum, you might realize how much of a stretch that is and instantly take it back to the drawing board. The second advantage is peer review--explaining why a null player is null to you is opening up for great feedback from players who may have an opinion, and they'll be more than happy to build off of your established reasoning and add onto it with their own and why they have a read there.

You don't have to write novels on all your conclusions. A one-liner is often surprisingly effective at conveying thoughts. But you should still be explaining why you think that something is, or isn't, likely to be true.

Know Timing

...That being said, sometimes you don't want people to know your thought process. Selectively holding back on timing is an artform, because it can help gather useful reactions. Knowing when to not share is just as important as knowing TO share. You should always have in mind an end time for your silence, though. Maintaining it indefinitely is anti-town.

Be Realistic

In all aspects of your play. You should have in mind at all times what is a realistic scenario and what is unrealistic. But you should also know what you can realistically do, and what you cannot realistically do. For instance, if you're suffering a bad V/LA, you probably shouldn't promise to write a novella on every player in the game because that is almost certainly beyond your realistic capabilities given your circumstances.

Apply Occam's Razor

Use it often. Point out how unlikely combinations are. Point out how many assumptions (especially flawed ones) a theory relies on. Keeping things simple is a guideline I cannot stress the importance of enough in terms of possibilities versus probabilities. Most remote possibilities flagrantly violate Occam's Razor; most probabilities use Occam's Razor.

In Summary:

As town, focus on what is probable rather than what is possible, and look for those who are failing to do so.