The ancient Romans played many ball games, but not all of them had specific names. Some of the games they described are difficult to reconcile with the games that we know of. A number of these games involved play around a circle or circles, for example the game we've named Pila (see the Pila webpage), and the games called sky-ball and hop-ball. The latter game, hop-ball, may be the game that we herein call Roman Ball, named in honor of the Romans and in lieu of any further evidence that this game actually existed.


      Several obscure references exist that suggest that children and adults played a game with a ball around a circle. Isidorus mentions Romans playing a ball game in which he mentions "the circle of players standing by and waiting." Greek children played a game in a circle in which they caught a ball that was thrown or bounced "into the sky." This may or may not have been the same game called "hop-ball" by the Romans. It may also be one of the ancient Egyptian games represented on the walls of Beni-Hassan.

      The un-named game played by Trimalchio is almost unanimously believed to be Trigon, yet there are some incongruencies that suggest it may have been a different game. The Spartans played a game that involved a circular field surrounded by a moat, called Platanistas, but apparently without a ball. The Greek game of Phaininda may have been played around a circle, as suggested by E. Gardiner. A version of "Monkey in the Middle" (possibly also Phaininda) was played around a circle, unlike the modern version played across two sides. Clearly the geometric perfection of the circle had some influence on playing fields of ancient games, whereas in modern games only rectangles are employed.

      The game called "ourania" (sky-ball) by the Greeks, was likely also played by the Romans. According to E. Gardiner, the ball was thrown up into the air and the object was for the players to catch it. Presumably there would be a circle within which, or on which, the players would stand. However, if the ball came down within the circle it seems impossible for anyone not to catch it. Conversely, if the ball were thrown far outside the circle it would seem impossible that anyone could catch it at all. A thrown ball seems most improbable.

      The Latin term for propelling a ball, expulsim, could apply equally well to throwing it as to bouncing it off the ground. But only bouncing the ball makes for a playable (i.e. scorable) game. It could be that there was only one circle and the thrower stood inside it to bounce the ball skywards. The next person to catch the ball would then step inside the circle. Alternatively, there could have been two circles, which gives us the game of Roman Ball as described by the rules above.

      The ball-game played by Trimalchio in The Satyricon is not named as Trigon, but has been assumed so by many historians. However, several apparent contradictions between this description and Trigon remain unexplained. We explore here the postulate that this game was very different from Trigon, and from all other games, and we call this game Roman Ball. This may be what Romans meant in general when they referred to "ball-playing" and "ball-players," even though this term could be applied to most ball games. The evidence detailed herein is more circumstantial and inferential than it is conclusive, but it does resolve a number of literary and historical loose ends.

      Taking it literally, the description of the game played by Trimalchio in the Satyricon was played around a circle, not a triangle, and therefore is completely distinct from Trigon. The ball bounces, per the Satyricon, and therefore this must be a small pila, or the same ball used in expulsim ludere (handball). The trigon was a hard ball. Balsdon states that this was a serious game with each player having his own scorer, called a pilicrepus, and his own ball-boy, and that more than one ball could be in play at a time. In the Satyricon there was only one scorer. Trigon is played by three players, presumably, but Trimalchio plays with a "bunch" or a "troupe" of boys. Furthermore, Trimalchio does not seem to be any kind of serious ballplayer -- he was wearing sandals and playing the game purely for leisure. The scoring of Trimalchio's game could be done either by counting the catches, or the misses. Our experiments, in fact, made this quite obvious. In Trigon, the ball is scored only when it is dropped, although it is possible that catches were scored up to 21 points, at which the game would likely end as a tie if no ball had been dropped.

      Roman Ball may be the game described as "hop-ball" by LaFaye and by Carcopino. Since the ball must bounce in the center circle, it could be described as "hopping." Alternatively, having experimented at length, we find that players must often jump or "hop" to catch the ball, as it tends to fly over their heads.

      The image of Roman ball-players shown on the main page does not seem to be Trigon. The players are in a crowd, instead of being in a triangular position. There is only one ball in play and it is high in the air. The players are not dressed for as serious a game as Trigon -- with togas on they can't be using both hands. Though there is no apparent circle drawn on the ground, the game of Roman Ball, as we have postulated it, ends up bears a resemblance to this picture. Being a game of pure fun, the scoring, and the clothes, may be of less importance.


  1. Draw 2 concentric circles on the ground, 5 feet and 20 feet in diameter.
  2. Players ( 3 or more) may stand or run anywhere outside the large circle.
  3. The ball must bounce in the inner circle, the 'strike zone', and pass beyond the outer circle.
  4. If the ball is not caught and hits the ground, the thrower gets a point.
  5. The player who catches or retrieves the ball throws it next..
  6. The first player to reach 21 points wins the game.

    • The player with the ball may run around the circle and try to catch his opponents out of position.
    • The player who catches or retrieves the ball may return to the circle quickly for the same reason.


    • The first throw can be made by anyone, but should be from standstill. The game then begins on the second throw.
    • If a ball fails to pass beyond the outer circle it should be replayed. Alternatively, "street rules" can allow anyone to grab it, and the thrower loses his scoring opportunity.
    • Offensive interference would have been considered very unsportsmanlike by the Romans. Americans, however, competitive as they are, might ultimately need new rules and a referee.


  • Any rubber ball works fine, but a rubber baseball works best. A baseball with a rubber core is probably ideal. The Romans seem to have used a softball-sized ball (see pictures). Tennis balls and handballs work well, but 'superballs' add some dimension to the game. Playground balls, nerf balls and volleyballs don't work well (not enough bounce). See The Ideal Roman Ball for some recommended ball specifications.
  • A grass lawn or hard dirt surface works fine, especially since you can dive for the ball. Use a stick and some string to spray-paint or chalk white circles. The Romans used ground gypsum for this.
  • Balls bounce better on concrete, and chalk works for the circles, but diving for the ball is definitely out of the question, at least without wearing pads. Gym floors are good but the walls could be too close.


      If there are two concentric circles, the ball must bounce in the inner circle and pass beyond the outer circle, otherwise the inner circle serves no purpose. A game in which the ball is thrown only through the air would require one circle, but not two and a very large one at that. Several courts could be located at the baths, or in private yards such as at Pliny's villa, which suggests moderate dimensions.

      Stephen Benko (in one of his books not yet unidentified) makes reference to condemned criminals who were sometimes made to play this game to the death in the arena (presumably the winner was not executed).

      No courts with inscribed circles have yet been found, or, if they were, have not recognized as such. The public ball courts, called sphaerista, served multiple purposes and therefore the lines for Harpastum, triangles for Trigon or the circles could be drawn in the dirt, as necessary. The palaestra was a large field used for athletics and also for ball games. Private ball courts may have been constructed with stone floors.

      We have tried to play variations of this game, and discovered that it works very well in its simplest form. Although there may have been additional rules, none seem to be necessary. Almost any number of players can be accomodated around the outer circle.

      Admittedly, these rules represent a modern invention but, having stumbled into this simple circular ball game, it is clear it could easily have been invented and played by any culture capable of making a bouncing ball and drawing circles on the ground.

    since the Ides of November
    in the 2749th year after the founding of Rome.