Archaeologists may need to rewrite one of the oldest chapters in the history of games.
A new study published in American Antiquity argues that Indigenous peoples living on the western Great Plains were making and using dice at least 12,000 years ago, near the end of the last Ice Age. If the identification is correct, these objects predate the earliest securely recognized dice from Mesopotamia, the Indus civilization and the Caucasus by more than 6,000 years.
That does not prove with absolute certainty that dice were invented in North America before anywhere else. Archaeology can only compare evidence that has survived and been discovered. The stronger conclusion is that these are currently the earliest known convincing examples of dice, dramatically extending the archaeological history of games of chance.
The research also suggests that gaming may have served purposes far more important than entertainment. Dice could have helped ancient communities negotiate exchanges, interact with outsiders and build social relationships across large territories.
Read the original Good News Network story
The Oldest Strong Evidence Comes From the Folsom Period
Archaeologist Robert J. Madden of Colorado State University traced the strongest early evidence to Folsom archaeological deposits in what are now Wyoming, Colorado and New Mexico.
The Folsom cultural complex dates to roughly 12,845 to 12,255 years before present. Madden identified 20 diagnostic or probable dice from Folsom contexts at Agate Basin in Wyoming, Lindenmeier in Colorado and Blackwater Draw in New Mexico. Nineteen were made from bone, while one probable example was fashioned from steatite, or soapstone.
These pieces bear little resemblance to the familiar six-sided cubes used in modern board games and casinos.
Instead, they were generally two-sided objects, sometimes flat and sometimes slightly curved. Their surfaces could be distinguished through carved lines, edge markings or pigment. Tossing several pieces would generate varying combinations depending on which surfaces landed facing upward.
In practical terms, they functioned more like multiple coin tosses than modern numbered dice.
One Possible Example Could Be Even Older
The archaeological trail may extend farther back.
Madden identified a small polished and incised bone fragment from the Clovis-era Murray Springs site in Arizona as a probable die. Clovis archaeological contexts date to roughly 13,050 to 12,750 years ago.
However, the evidence is less certain because researchers cannot establish whether the two surfaces of the object were intentionally differentiated. The study therefore classifies it only as a probable example rather than using it as the main evidence for the claim.
That distinction is important. Headlines describing “13,000-year-old dice” are possible, but the more robust evidence comes from the slightly later Folsom sites.
The study’s central conclusion consequently remains conservative: Indigenous North Americans were intentionally producing dice no later than about 12,000 years ago.
Archaeologists Had Already Found the Dice—They Just Didn’t Know What They Were
One of the most interesting aspects of the discovery is that many of the objects were not newly excavated.
Some had been sitting in archaeological collections for decades.
The problem was classification. Small pieces of shaped or decorated bone can have many possible functions, and archaeologists have traditionally been cautious about identifying an object as a gaming piece without compelling evidence.
Madden attempted to overcome that problem by creating an objective morphological test.
The starting point was ethnologist Stewart Culin’s monumental 1907 work Games of the North American Indians. Culin documented hundreds of Indigenous games and recorded 293 sets of Native American dice from communities across the continent.
Madden compared characteristics shared by those historically documented examples and then searched the prehistoric archaeological record for objects with matching features.
The original Culin volume remains preserved by the Smithsonian and provides an extraordinary historical catalogue of Indigenous North American gaming traditions.
Explore Stewart Culin’s Games of the North American Indians at the Smithsonian
The Search Identified Hundreds of Ancient Dice
The resulting dataset was much larger than a handful of Ice Age artifacts.
Madden identified 565 diagnostic dice from 45 prehistoric archaeological sites, plus 94 probable examples. Together, the study catalogued 659 diagnostic and probable dice from 57 sites.
The artifacts span thousands of years and numerous cultural traditions.
They appear among mobile hunter-gatherer societies, semisedentary communities and later agricultural populations. The sites were associated with cultural complexes including Folsom, Agate Basin, Fremont, Basketmaker and Pueblo traditions, among others.
That long distribution matters almost as much as the age of the earliest pieces.
It suggests that games of chance were not a brief experiment that disappeared. Instead, dice appear to represent an extraordinarily persistent cultural tradition extending across roughly 12 millennia.
Read the full open-access American Antiquity research paper
These Dice Appear Thousands of Years Before Old World Examples
Before this study, securely identified early dice were generally associated with complex Bronze Age societies.
Researchers have dated the earliest confidently recognized Old World examples to around 3500 BCE, or approximately 5,500 years ago, in areas including Mesopotamia, the Indus civilization and the Maikop culture of the western Caucasus.
The Folsom examples are more than twice as old.
That shifts the geographical narrative as well as the timeline. Gaming technology associated with intentionally producing random outcomes may not first appear in the archaeological record of early cities and agricultural civilizations. Strong evidence now places it among highly mobile hunter-gatherers in North America thousands of years earlier.
This is why the research is about more than the history of recreation.
Dice Represent an Early Attempt to Harness Randomness
Throwing dice creates something unusual: a deliberately produced random event.
Humans cannot know exactly which side will land upward on an individual throw, yet repeated throws create recognizable patterns. Madden argues that deliberately manufacturing and repeatedly using such objects implies at least a practical understanding of chance, fairness and randomness.
There is no evidence that Ice Age players calculated mathematical probabilities in the modern sense. Formal probability theory did not emerge until thousands of years later.
Yet using dice successfully in games requires participants to recognize that outcomes are uncertain and that competing players should have comparable opportunities to win.
The study therefore places Indigenous North Americans surprisingly early in humanity’s practical engagement with ideas that would eventually become central to probability, statistics, insurance, economics and science.
Gaming May Have Connected People Who Rarely Met
The social implications may be even more revealing.
Previous research on Native American gambling suggests games often operated between people who were neither close relatives nor complete strangers. Shared gaming rules could create a neutral setting in which people from different groups interacted.
Madden argues that gambling could therefore have worked as a kind of social technology. People could wager objects, negotiate their value and exchange goods immediately without needing an existing system of reciprocal obligations.
Such games may have helped circulate goods, information and social connections across large areas.
That possibility is particularly intriguing for highly mobile Folsom hunter-gatherers. Archaeologists have long debated whether normally dispersed bands periodically gathered in larger groups.
Multiple dice were found at sites already considered possible aggregation locations. Fifteen diagnostic or probable dice came from Lindenmeier and four from Agate Basin, leading Madden to suggest gaming artifacts might eventually help archaeologists identify prehistoric meeting places.
The interpretation remains a hypothesis, but it presents gaming as something much larger than spare-time amusement.
Women May Have Played a Major Role in the Tradition
Another intriguing possibility concerns gender.
Historical and ethnographic accounts indicate that many Native American dice games were associated particularly strongly with women. One analysis cited in the new research examined 131 cases in which players’ gender was recorded. In 81 percent, the game was played exclusively by women; 12 percent involved both sexes and only 7 percent were exclusively male.
It cannot simply be assumed that the same pattern existed 12,000 years ago.
Still, Madden argues that the evidence creates an important question for future archaeology: whether women may have played a leading role in developing, transmitting or maintaining these gaming traditions.
That possibility could alter assumptions about where important innovations in prehistoric social and intellectual life originated.
The Discovery Changes More Than the History of Games
The headline attraction is obvious: ancient Native Americans appear to have been rolling or tossing dice thousands of years before the earliest currently known examples elsewhere.
But the larger story is about what those small pieces of bone represent.
They suggest that Late Pleistocene communities were creating structured games, generating controlled randomness, agreeing on rules of fairness and possibly using gambling to connect normally separated groups. The tradition then appears repeatedly across the archaeological record for thousands of years.
Future discoveries could push the chronology still farther back or reveal comparably old gaming objects elsewhere in the world. Archaeological claims about “the first” should always remain open to new evidence.
For now, however, the strongest available evidence places the world’s earliest securely recognizable dice in Indigenous North America.
The objects may be tiny, but their significance is enormous: people living on the Great Plains near the end of the Ice Age were not simply surviving. They were playing, socializing, exchanging and experimenting with chance.
