In mahjong, an unseen tile can become less likely without moving anywhere.
Nothing physical has happened to it. What changes is the evidence around it.
A discard appears on the table. Another player exposes a set. More tiles become public. The wall shortens. The hidden state of the game is still hidden, but the range of credible possibilities has changed.
The table is an information environment
Mahjong emerged in China in the late Qing period and spread through homes, clubs and communities across many countries. Its rules vary by region and tradition, so any mathematical analysis has to begin by defining which rules are in force.
That qualification is not administrative fussiness. It is Mathematics. A probability belongs to a specified sample space and mechanism. Change the rules and the probabilities change with them.
At the table, players usually know three kinds of things: what is in their own hand, what has become public, and what remains unknown.
The interesting work happens between the second and third categories.
A discard is more than a discarded object
Suppose several copies of a tile are already visible. Fewer copies can remain hidden. The tile may still be possible, but its practical availability has changed.
This distinction is small and fundamental:
Possible does not mean equally probable.
Students often stumble over exactly this point. Listing all possible outcomes is not enough to assign probabilities unless those outcomes are generated with equal likelihood.
Mahjong keeps forcing the question back onto the mechanism: how many copies remain, what has already appeared, how many draws remain, and what actions have changed the information available?
Actions become evidence too
A player’s choice may also alter another player’s model of the game.
Repeated discards from one suit may suggest one hypothesis. An exposed combination narrows another. None of these observations guarantees what a player holds; people can act for several reasons. But behaviour joins the evidence field.
Now counting and interpretation meet.
Combinatorics tells us what arrangements are possible. Conditional reasoning asks how the credibility of those arrangements changes after new evidence arrives.
The second task is often harder because it demands restraint. Evidence should change a belief proportionately, not magically turn a plausible story into certainty.
Uncertainty is not the same thing as ignorance
School exercises often end with one exact answer. Real decisions frequently arrive earlier than certainty does.
Probability gives us a disciplined language for that condition. We can separate known from unknown, estimate relative likelihood, update when evidence arrives and leave uncertainty visible where it remains.
That habit matters far beyond games. A student’s latest score is evidence, not destiny. One careless error is evidence, not a diagnosis. A pattern across several pieces of work may justify a stronger conclusion than one isolated incident.
The quality of the judgment depends on how faithfully confidence tracks evidence.
Mahjong is also family time, competition, ritual, memory, conversation and regional rule culture. Its social life is larger than its probability structure.
But the hidden Mathematics leaves one unusually portable idea:
Information begins to have value before it becomes certainty.
Good judgment is not waiting until nothing is unknown. It is knowing exactly how much the evidence allows you to say now.
Further reading
Oxford University Press — Mahjong: A Chinese Game and the Making of Modern American Culture
Academia Sinica — Madiao and Mahjong in Popular Culture and Elite Discourse