The Story
Where it came from, and why it turned out this way.
DesignWilliam Julius Champion Jr. · United States, 1940s
It is not a traditional African game
Many people assume Kalah is an ancient African mancala. It is not. It is a commercial game designed and patented by the American William Julius Champion Jr. in the 1940s, and heavily promoted as an educational aid in American schools through the 1950s and 60s. The mancala family genuinely is ancient, but Kalah is a young and deliberately simplified member of it — the real traditional competitive mancala is the game next door, Oware.
Two rules, both in service of speed
Champion's changes reduce to two: add a store, and grant another turn when your last stone lands in it; and capture across when your last stone lands in an empty pit on your own side. The first supplies the satisfaction of chained turns, the second the payoff of a big one-move haul. Traditional mancala often takes several laps to resolve; Kalah compresses the rhythm into a handful of moves. This is a mancala designed for classrooms and toy shelves.
The price is that it has been solved
Simplification bought solvability. Around 2000, work by Geoffrey Irving and colleagues fully solved several Kalah configurations using retrograde and search techniques, including the common six-pit game with four stones per pit: a first-player win. That leaves Kalah with no strict competitive suspense, and its value shifted to teaching — it remains one of the most-used introductory examples for explaining game trees, minimax, and alpha-beta pruning.
Why the chain is the whole game
Understanding Kalah takes one piece of arithmetic: a pit n places from your store containing exactly n stones drops its last stone into the store, and you move again. What strong players do is string several such pits together, taking three or four turns in a row while redistributing the board into the next chain. The opening pit sitting four places from the store with exactly four stones is the canonical first link.
- Mancala is not one game but the name of a family, spread across Africa, the Middle East, South Asia, and Southeast Asia, with hundreds of distinct rule sets.
- Captures work only from an empty pit on your own side — the rule most often confused between games, since Oware's harvesting mechanism is entirely different.
- At scoring, each player sweeps the stones left on their own side into their own store, so emptying your side early is not automatically good; check what remains on your opponent's side first.
Complete Rules
Start with what a turn allows and how the game is won.
- Board
- 6 pits and 1 store per player
- Setup
- 4 stones per pit, 48 total
- Winning conditionthere are 48 stones in play and every one of them eventually ends up in one of the two stores. Whoever holds more in their own store wins. Sowing, extra turns, and captures all exist for the single purpose of moving more stones into your store.
- Choose one of your non-empty pits, lift out all of its stones, and drop one stone into each following pit going counter-clockwise.
- Drop a stone into your own store when you pass it, but always skip your opponent's store. A complete circuit is therefore thirteen landing spots for you, not fourteen.
- If the last stone lands in your own store, you immediately take another turn. This can trigger repeatedly, so a single turn may contain several sowings.
- If the last stone lands in a pit on your own side that was empty — so it now holds exactly 1 stone — you capture that stone together with every stone in the pit directly opposite, and put them all in your store. If the opposite pit is empty no capture happens and the single stone simply stays there.
- Captures work only on empty pits on your own side. A last stone landing in a non-empty pit of yours, or in any pit on your opponent's side, never triggers a capture.
- The game ends the moment either player's six pits are all empty at once. Both players then sweep the stones remaining on their own side into their own store, and the larger store wins; 24 to 24 is a draw.
Your First Game
Use this order of thought to get through a complete game quickly.
- Count the stones in your hand first, then count pit by pit along the sowing direction to find where the last one lands. Only the last stone's position determines whether you get an extra turn or a capture — the intermediate drops only shape the position afterward.
- Finding an extra turn follows a fixed formula: a pit n places from your store containing exactly n stones will drop its last stone into the store. At the start of the game, the pit four places from your store holds exactly four stones, which is the standard opening extra-turn move.
- The key to captures is your own empty pits: an empty pit is a hook, and if you can land your last stone in it while the pit directly opposite is well stocked, one move harvests a large haul.
- Before finishing your move, run the same count over your opponent's six pits to see whether they can take an extra turn next, or capture one of your fat pits.
- The facing pairs never change: they are the two pits vertically aligned in the interface. Since the two rows sow in opposite directions, your pit farthest from your store faces your opponent's pit nearest to theirs, and your nearest pit faces their farthest. Internalizing this mirror saves an enormous amount of counting.
Practical Strategy
Move from knowing the rules to playing with a plan.
- The pits nearest your store are your tempo bank: use them to set up an extra turn, then use the new distribution after that turn to set up the next link in the chain. Three or four sowings in one turn is not unusual.
- When a lot of stones pile up opposite one of your empty pits, your opponent will do everything they can to sow that pit empty — and conversely, you should avoid letting stones accumulate opposite one of your own empty pits.
- Sometimes declining a small capture pays better: leave the pit empty as a hook, wait for your opponent's side to fatten up, and harvest later. One such capture can be worth several moves of small gains.
- The pit farthest from your store is the engine for large-scale sowings; hoarding it lets you sweep across the entire row at a critical moment, rewriting the distribution on both sides at once.
- At scoring time your opponent sweeps their own remaining stones into their store, so deliberately emptying your side only pays when your lead exceeds the total left on their side. When you are behind, do the opposite and stall to avoid triggering the end early.
Winning and Special Rules
Details that are easy to miss but can change the result.
- One turn may contain several sowings through repeated extra turns. The interface animates each sowing in sequence, and the move counter counts every sowing as a move.
- The end condition is either side's six pits being empty, including the case where your own chain of extra turns empties your own side. At that instant the game ends even if you still hold the right to move again.
- All forty-eight stones eventually reach the two stores, so one side always passes the halfway mark unless the result is an exact 24 to 24 tie.
- Sowing direction, skipping the opponent's store, extra-turn detection, and capture detection are all implemented once and shared by the interface, the AI search, and the master hint.
Common Mistakes
The traps new players fall into most often.
- Forgetting to skip your opponent's store, which shifts every landing spot by one and invalidates your entire extra-turn and capture calculation.
- Assuming any empty pit allows a capture. Only empty pits on your own side count; a last stone landing in an empty pit on your opponent's side does nothing at all.
- Focusing only on how many stones this move earns you, without checking whether it incidentally fills a pit into a perfect extra-turn starter for your opponent.
- Importing Oware rules and expecting a last stone to harvest by making an opponent's pit total 2 or 3. Kalah has no such rule, and no feeding obligation either.