At least 1,300-plus years old on hard evidence: securely dated boards from the Aksumite region of the Horn of Africa run from roughly the 2nd to 7th centuries CE. Older claims (Egypt at 1400 BCE, Neolithic Jordan) rest on undatable or ambiguous pit rows. The true age is likely greater; the provable age is not.

Where did mancala originate?

The best-supported answer is Africa, with the earliest accepted archaeological finds in the Horn of Africa (present-day Eritrea and Ethiopia). From there, trade routes spread sowing games across the continent, into the Middle East and Asia, and the Atlantic slave trade later carried West African versions to the Caribbean.

Was mancala played by Egyptian pharaohs?

There’s no solid evidence for it. The pit rows at Egyptian temple sites can’t be dated or confirmed as game boards, and Egypt’s documented ancient games, like senet, are unrelated. The pharaoh story is repeated because it’s evocative, not because anything substantiates it. Egypt does have a rich later mancala tradition.

What does the word mancala mean?

It derives from the Arabic naqala, meaning “to move,” and names the whole family of sowing games rather than any single one. The word reached European languages through Arabic. Regional names are older and more specific: oware, bao, sungka, congkak, pallanguzhi, omweso, and hundreds more.

Who invented the mancala sold in stores?

American versions are Kalah, designed around 1940 by William Julius Champion Jr., who sold sets from 1944 and founded the Kalah Game Company in 1958. He built it from his reading of traditional African and Asian games. The family is ancient; the boxed ruleset is younger than nylon.

The History of Mancala

Mancala takes its name from the Arabic naqala, “to move.” The earliest secure archaeological evidence comes from the Horn of Africa around the 2nd to 7th centuries CE, in the region of the Aksumite kingdom. Claims that mancala is the world’s oldest board game are widespread, appealing, and unproven.

Nearly every mancala page on the internet opens with “one of the oldest games in the world, played by ancient Egyptian pharaohs 7,000 years ago.” It’s a great line. The evidence for it is somewhere between thin and imaginary, and the real story, which runs through Aksumite ruins, Indian Ocean traders, the Atlantic slave trade, and a Yale graduate with a patent, is better anyway.

The name

“Mancala” comes from the Arabic naqala, to move, and entered European languages through Arab culture, where the word covered the whole class of sowing games. That’s the right way to hold the term: mancala is a family name for hundreds of related games, not one game. (How to say it, and what it’s called around the world, gets its own page: mancala pronunciation.) The academic literature likes “count and capture” games or “sowing games,” which describe the shared mechanics: pits, identical seeds, and the one-per-pit sowing motion.

What the evidence actually shows

The oldest finds that scholars broadly accept as mancala boards come from the Horn of Africa, in territory of the Aksumite kingdom, dated roughly between the 2nd and 7th centuries CE. At Matara (in present-day Eritrea), excavations produced a fragmentary earthenware board assigned to about the 6th-7th centuries CE, along with limestone two-row boards from the same era; related finds come from Yeha in Ethiopia and near Asmara. That’s the floor: mancala-type games were being played in northeast Africa well over 1,300 years ago, and the boards already look like the game we know.

Everything older is contested. Rows of small pits cut into stone at Egyptian sites, including temple sites often cited with dates like 1400 BCE, can’t be securely dated (a hole in a rock doesn’t come with a timestamp) and can’t be securely identified as game boards at all rather than counting aids, offering trays, or something else. A pitted limestone slab from Neolithic ’Ain Ghazal in Jordan, sometimes billed as a 7,800-year-old mancala board, carries the same double uncertainty, and specialists have pushed back on the game interpretation. In 2024, archaeologists reported a whole “arcade” of mancala-style pit rows cut into rock in Kenya’s highlands: genuinely evocative, and genuinely undatable so far. The pattern repeats across the literature: spectacular ages, shaky attribution.

So is mancala the oldest board game?

Probably not, if we’re being honest with the word “evidence.” Senet boards from Egypt date to around 3100 BCE, and the Royal Game of Ur is attested from roughly 2600-2400 BCE, both with actual physical boards, in datable contexts, that are unambiguously games. Mancala’s securely dated record starts millennia later.

What mancala can claim is different and arguably more impressive: unbroken living tradition at continental scale. Senet and Ur are dead games reconstructed by scholars. Mancala never stopped. The sowing games played today in Ghana, Zanzibar, Manila, and Chennai are direct continuations of whatever those Aksumite players were doing, making mancala a strong candidate for the oldest game family still in continuous play. That’s the defensible version of the boast, and it doesn’t need invented pharaohs.

How the family spread

From its African origins the game moved the way people and goods moved. Arab traders carried sowing games along trans-Saharan and Indian Ocean routes, which is the standard explanation for the family’s presence from West Africa across the Middle East to South and Southeast Asia; Tamil Nadu’s pallanguzhi, Malaysia’s congkak, and the Philippines’ sungka sit along exactly those maritime corridors. Within Africa the family radiated into remarkable diversity, from the two-row games of the west, oware and ayo, to the four-row heavyweights of the east and the Great Lakes, bao and omweso.

The Atlantic slave trade forced the game across another ocean. Enslaved West Africans carried oware’s lineage to the Caribbean, where warri took root so thoroughly that it became Antigua and Barbuda’s national game. A game that traveled in memory, requiring nothing but holes in the ground and a handful of seeds, was among the cultural possessions no ship’s manifest could confiscate.

The American chapter: Kalah

The version on American shelves has a birth certificate. William Julius Champion Jr. read about mancala games as early as 1905, designed his own streamlined variant around 1940, sold sets from 1944, patented the design in 1952 and the rules in 1955, and founded the Kalah Game Company in Holbrook, Massachusetts in 1958. He marketed Kalah as an educational game, schools adopted it, and unbranded copies eventually shipped under the family name “mancala,” which is why Americans learned a twentieth-century ruleset believing it ancient.

Kalah earned its own historical footnote in 2000, when computer analysis by Geoffrey Irving, Jeroen Donkers, and Jos Uiterwijk solved the game: the first player wins with perfect play. An old family produced the first of its members to be mathematically finished, and people keep playing it anyway, which says something durable about why games exist. (What the solving result means for actual play is covered in the strategy guide.)

Living history

The family’s story is still being written: oware championship circuits, omweso revival tournaments in Kampala, pallanguzhi heritage workshops in India, sungka boards in diaspora groceries, and a few hundred million iMessage games of Avalanche. If reading about these games beats playing them, something has gone wrong; the full variants guide maps the family, and the game on this site is the fastest way to join its newest chapter.