Mafia Is a Mathematical Game
Author: Vitaly - mr. Koteo (Brisbane Mafia Club)
Why Mafia Is a Mathematical Game
Sports Mafia is often described as a game of psychology, intuition, or “reading people”. Those elements exist — but they are not what decides games at a competitive level.
At its core, Sports Mafia is a game of uncertainty management played under strict structural constraints. Victory does not go to the loudest voice or the strongest emotional read. It goes to the team that reduces uncertainty before the opposing team can hide, distort, or delay information long enough to reach a winning state.
This chapter explains why Mafia can be analysed mathematically. It does not yet teach full probability theory — it prepares you to see the game through a mathematical lens.
Mafia as Information Theory
At the start of a standard Sports Mafia game:
10 players are seated
3 of them are Dark
7 of them are Red (including the Sheriff)
From the Red team’s perspective, any placement of the 3 Dark players among the 10 seats is initially possible.
Mathematically, the number of possible Dark-team configurations is:
C(10,3) = 120Think of picking darks sequentially:
First dark: 3/10 (three out of 10)
Second dark: 2/9 (two out of 9)
Third dark: 1/8 (one out of 8)
Multiply: 
This means there are 120 different possible worlds consistent with the rules.
Important clarification:
These are not strategies
These are not suspicions
These are not reads
They are simply hypotheses about reality.
At the beginning of the game, Reds cannot distinguish between these worlds. Any early certainty is therefore not knowledge — it is confidence without sufficient information.
This is the fundamental information asymmetry of Mafia:
Dark players know exactly which world they are in
Red players must discover it gradually
Every meaningful action in the game either:
removes some of these worlds as impossible, or
fails to do so
How Game State Changes Drive Mathematical Progress
In Sports Mafia, mathematical change happens whenever the game state changes.
A game state change occurs when:
a player is killed at night, or
a player is voted out during the day.
Both events reduce the number of players alive and therefore reduce the space of possible Dark-team configurations.
However, these two types of change have different mathematical characteristics.
Night Eliminations: Strategic but Expected State Changes
Night eliminations are the result of a Dark team decision.
They are:
intentional,
strategic,
and chosen by the Dark team.
From a rules perspective, the Dark team may choose not to kill. In practice, this is rare and usually limited to specific late-game situations. Any missed kill — intentional or accidental — generally benefits the Red team by giving it more time.
From a mathematical perspective, a night kill:
reduces the number of players,
shrinks the solution space,
and advances the game toward a terminal state,
even if the Red team learns very little from the outcome itself.
Night eliminations therefore represent strategic but predictable pressure on the game state.
Day Voting: Decision-Driven State Changes
Day eliminations are:
the result of collective player decisions,
based on available information and reasoning,
and therefore reflect the quality of the Red team’s logic.
Mathematically, voting is more powerful than night kills:
it can remove either a Red or a Dark player,
it can accelerate convergence if correct,
or delay it if incorrect.
A correct vote-out removes Dark players from the solution space directly. An incorrect vote-out leaves the number of Dark players unchanged and increases the relative impact of future errors.
This is why voting quality matters so much at later stages of the game.
Why Both Matter
From a mathematical point of view:
nights guarantee movement,
days determine direction.
Together, they define how fast uncertainty collapses and whether it collapses toward the correct answer.
Understanding this distinction is essential for reading the game correctly and for appreciating why later chapters focus heavily on vote mathematics and critical decision points.
How Uncertainty Shrinks in Real Game Flow
Let:
N= number of players alive after a nightD= number of Dark players alive (unknown to Reds, but fixed in reality)
The number of possible Dark-team configurations among living players is:
Start of the game (before Night 1)
After Night 1 (first kill)
One player is killed.
In theory, the killed player could be Dark — but in real Sports Mafia, killing a Dark player on the first kill is mostly strategically irrational and therefore highly unlikely.
So the realistic mathematical state is:
This is the first meaningful reduction of uncertainty:
Already, 36 impossible worlds have effectively been removed.
After Day 1 + Night 2
During the day, one player is voted out. That player can be Red or Dark.
Then Night 2 occurs, and another player is killed — again, almost certainly Red.
So after Night 2:
Two players have left the table
At most one Dark could have been eliminated so far
The realistic possibilities are therefore:
Corresponding possible worlds:
So after just two informative nights, uncertainty has reduced from:
The exact branch is not yet known — but the space has collapsed dramatically.
Why Logic Beats Emotion
From a mathematical point of view, a useful action is one that eliminates possible worlds.
Most emotional reactions do not.
Common emotional traps include:
revenge voting,
voting someone “because they are strong”,
reacting to confidence, tone, or pressure,
trying to force certainty too early
These actions often fail a simple test:
Does this action eliminate any possible worlds?
If the answer is “no”, then no information has been gained — even if the action felt decisive.
Emotion increases confidence without increasing accuracy. Logic increases accuracy, even when confidence feels uncomfortable.
This does not mean emotions must be suppressed. It means emotions must be constrained by structure.
Logic, Intuition, and Probability
Players often describe good decisions as “intuitive”.
In Sports Mafia, intuition is best understood as pattern recognition built from experience. It has value — but it is unreliable under pressure and difficult to verify.
For this reason, intuition must be bounded.
You may hear this expressed as a figurative rule:
“90% logic, 10% intuition.”
The numbers are not literal. They express a principle: logic must dominate.
As the number of players decreases:
each decision carries more weight
errors become harder to recover from
probability matters more than instinct
Voting Patterns as Mathematical Signals
Votes are the most important observable data produced during the day.
In real Sports Mafia play:
votes happen quickly,
and players must mentally track voting patterns in real time.
This cognitive limitation is part of the game.
From a mathematical perspective, votes are powerful because they are:
public,
binary,
and forced.
Each vote creates a commitment that cannot be undone, even if it is later explained away verbally.
Although votes are only formally written down after the game for analysis, during the game they function as constraints in the players’ mental models.
Players who can:
remember who voted with whom,
notice repeated alignments,
and integrate this information over multiple rounds
are effectively performing real-time data aggregation under pressure.
This is why voting patterns carry far more mathematical weight than individual statements.
Formation of Two Distinct Teams and Calculability
Early in the game, uncertainty is maximal:
This space is too large to analyse directly.
As the game progresses, repeated voting behaviour often causes players to organise into two distinct voting groups. This does not mean the groups are correct — only that they are stable.
The formation of two distinct teams does not guarantee truth. It guarantees calculability.
Once players are consistently voting in the same directions, the nature of the problem changes.
Instead of analysing individual players, the game becomes an exercise in distribution:
How many Dark players can exist inside each group?
This reframing reduces the effective solution space and makes probability-based reasoning possible.
Importantly, this is a structural effect — not a psychological one. Even an incorrect grouping is easier to analyse than a chaotic table with no stable alignment.
The ability to recognise and reason about these emerging teams is a key step in transitioning from intuitive play to mathematical play.
Chaos, Structure, and Winning Conditions
Chaos is not neutral.
When:
votes are scattered,
directions constantly change,
no stable coalitions form,
uncertainty collapses much more slowly.
This benefits the Dark team, whose win condition depends on time and parity, not certainty.
Red team wins by convergence. Dark team wins by prolonging ambiguity.
Structural Thinking in Sports Mafia
The mathematical way to play Mafia is not to ask:
“Who is Dark?”
It is to ask:
“Which worlds are still possible?”
A structure is any constraint that removes large sets of impossible worlds:
voting coalitions,
avoidance patterns,
Sheriff checks,
forced commitments
Winning Red play steadily reduces the solution space until only a few worlds remain. Winning Dark play delays that reduction until parity is reached.
Victory is not about early certainty. It is about reaching high probability before the game ends.
What This Chapter Sets Up
This chapter explains why Sports Mafia is a mathematical game.
The next chapters explain how:
From this point on, Mafia should feel less mysterious and more structural, analytical, and calculable.
Suggested next step
I recommend moving directly into Probability Essentials, starting with:
priors,
conditional probability,
and how those
{35, 21}branches get weighted.
Last updated