How to Practice a Game So It Actually Sticks

Most people improve at a game the same way: they play, they lose, they play again. It works, eventually. It is also the slowest method available, and the reason has nothing to do with talent or effort.

Playing a game gives you almost no usable feedback.

Think about what a hand of American Mahjong actually tells you. You passed three tiles in the Charleston and lost the hand forty minutes later. Which of those two facts caused the other? You will never know. The tiles you gave away might have won it. The hand might have been unwinnable from the deal. Nothing at the table separates the two.

This is what learning researchers call a low-feedback environment, and games are full of them. You get an outcome, delayed and noisy, for a decision you made long ago among several dozen others.

What actually builds skill

Three things, and they are the same three whether the game is Mahjong, Canasta or Bridge.

1. Isolate the decision

Instead of playing a whole hand, look at one situation and make one choice.

Which three tiles do you pass? Do you take this pile or freeze it? Do you bid, or pass and wait? A single decision, with the rest of the game held still around it.

This is unnatural, which is why almost nobody does it on their own. It is also why a player who does it improves several times faster than one grinding out hands.

2. Get told why, not just whether

Being told you were wrong teaches you one thing about one position. Being told why teaches you a rule you can carry to positions you have never seen.

“That was the wrong pass” is nearly useless. “That pass broke your only pair, and the card has no line that takes a joker there” is a principle. Next time the shape repeats, you will see it.

This is the part a table cannot give you. Your opponents will not explain their reasoning, and your partner may not know theirs.

3. Meet the same idea again, later

Learn something once and you forget most of it within a week. Meet it again just as it is fading and it sticks much longer. Meet it a third time later still and it becomes permanent.

That is spaced repetition: reviewing on a widening schedule timed to when you are about to forget, rather than cramming or reviewing at random. It is well established in language learning and medical training, and it applies cleanly to games because games are made of recurring patterns.

The scheduling is the whole trick, and it is why doing this by hand rarely works. You cannot tell which ideas are fading.

Why “just play more” underperforms

Here is the gap, concretely.

Repetitions are rare. In an evening of Mahjong you make maybe a dozen Charleston decisions. A drill session gives you fifty in ten minutes.
You practice what you already do. Playing rehearses your habits, including the bad ones. Nothing at the table forces you into the situations you are worst at.
The feedback is delayed and mixed. Winning a hand you played badly teaches you the wrong lesson, and it happens constantly.
Nobody at the table is teaching. Your group wants a game, not a lesson, and stopping play to analyze a decision makes you the person nobody invites.

None of this means stop playing. Playing is the point. It is just a poor way to get better at the specific decisions that decide games.

What to drill, by game

The general rule: drill the decisions that are frequent, consequential, and give you no feedback at the table.

American Mahjong. The Charleston, first, because it is every hand and you never learn whether you passed well. Then reading the card fast enough to spot which lines your tiles support. Then joker legality, which is a rule you either apply instantly or lose tempo to.

Canasta. One decision dominates: take the pile, or freeze it. It recurs every few turns, the cost is large, and the outcome is buried in the rest of the hand.

Bridge. The auction. Every bid is a constrained message, the vocabulary is finite, and the feedback at the table is confounded by whatever your partner did next.

How Stoa Games does it

This is the whole reason the app exists, and it is why we describe it as teaching rather than playing.

Lessons in a dependency order, so nothing arrives before the thing it assumes.
Drills built from real positions, one decision at a time, dozens per session.
An explanation on every answer, covering why the strong move is strong and what the plausible alternative gives up.
Spaced-repetition scheduling, so ideas come back as they fade instead of on a whim.
A card companion for American Mahjong, which works with the card you own. We never reproduce the League’s hands; you enter your own.

American Mahjong is the game we teach in full today. Canasta and Bridge follow the same shape when they arrive.

Where to start

If you are learning a game from scratch, read its guide first. Drilling before you know the rules is just guessing.

Then practice the decisions, not the games.

## Frequently asked questions ### Does spaced repetition work for games? It works for anything built from recurring patterns you need to recognize quickly, which describes card and tile games well. It is the same method used in language learning and medical training. ### How long should a practice session be? Short and frequent beats long and occasional. Ten focused minutes several times a week does more than one long session, because the spacing is doing much of the work. ### Will drilling make me better at the table? At the decisions you drill, yes, and those are the decisions that recur. It will not substitute for table experience in reading people or handling a group's house rules. ### I have played for years. Is this for beginners? The rules are for beginners. The decisions are not. Experienced players usually have one or two habits that have never been tested, and drilling is how you find them.