The circle method, which is the whole trick
A round robin schedule has to satisfy one hard constraint: within any round, no entry appears twice. Assigning pairs at random and fixing the collisions afterwards is how most people try it, and it degenerates fast past six teams.
The circle method solves it structurally. Fix one entry in place. Arrange the rest in a ring around it. Pair them off across the ring, then rotate the ring by one position and pair again. After n minus 1 rotations every pair has met exactly once and no round contains a repeat. That is what this page runs, and it is why the fixture list comes out valid without any collision checking.
Odd entry counts get a phantom entry added to the ring. Whoever is drawn against the phantom in a given round takes the bye. Because the phantom rotates like everything else, the byes distribute evenly rather than landing on the same team twice.
Game counts, and why they hurt
| Entries | Round robin games | Rounds | Knockout games |
|---|---|---|---|
| 4 | 6 | 3 | 3 |
| 6 | 15 | 5 | 5 |
| 8 | 28 | 7 | 7 |
| 10 | 45 | 9 | 9 |
| 16 | 120 | 15 | 15 |
| 24 | 276 | 23 | 23 |
| 32 | 496 | 31 | 31 |
Round robin games grow with the square of the entry count while knockout games grow linearly. At four entries the two formats cost six games against three; at 32 they cost 496 against 31. That single fact drives almost every real tournament structure. Anything above about eight entries with a fixed time slot ends up as group stages followed by a knockout, because the groups keep the fairness where it is cheap and the bracket handles the elimination where it is expensive.
A workable middle ground for 16 entries: four groups of four, six games per group, 24 games total, then the group winners into a four-team bracket for three more. Twenty-seven games instead of 120, and everybody still gets three matches guaranteed.
Seeding and where byes go
A bracket of 13 entries is padded to 16, and the three empty slots become byes. Convention gives them to the top three seeds, which is what happens here when you leave the order as typed. The pairing itself is the standard one: first seed against last, second against second-from-last, and so on. This puts seeds one and two in opposite halves so they can only meet in the final, and it means an upset early on does not clear the path for another strong entry.
If your event draws lots rather than seeding, choose Shuffle. The entry order is randomised with a Fisher-Yates pass before the bracket is built, so the byes fall randomly too. Press the button again for a fresh draw; nothing is remembered between presses.
Choosing a format
Round robin is fair and slow. Every entry plays the same opponents, one bad match does not end anyone's day, and the final table reflects the whole event rather than one afternoon. It also means half the fixtures late in the schedule are dead rubbers between teams with nothing left to play for.
Knockout is fast and brutal. Half the field is finished after one match. For a works sports day or a charity event where people came to play, that is usually the wrong shape, and adding a plate competition for first-round losers or a third-place playoff costs almost nothing.
Double round robin, the third option here, is the league format: everyone plays everyone twice, once at each venue. It doubles the fixture count and removes the complaint that the draw handed someone all their hard matches away from home. It is the right choice for a season and the wrong one for a single day.
Ranking a round robin table
Once the games are played you need a tie-break order, and the time to agree it is before the first whistle rather than after the last. The near-universal starting point is three points for a win and one for a draw, which was adopted specifically to make winning worth more than two draws. From there the common sequence is head-to-head result between the tied teams, then goal or point difference, then points scored, then a coin toss or a playoff.
Head-to-head first and difference first give different champions often enough to matter, and both are defensible. Pick one, write it on the sheet, and post the sheet where everyone can see it.
Related tools here
This page builds fixtures between entries that already exist. To form the entries themselves from a list of people, the team divider shuffles a name list into evenly sized groups. For a single random choice, a captain or an order of play, use the random picker, and seat assigner handles placing people at numbered positions. If you need to record who actually turned up, the attendance sheet maker prints the grid.
Questions people ask
How is this different from the team divider?
They do opposite halves of the same job. The team divider takes a list of people and splits it into groups — fifteen names into four teams of 4, 4, 4 and 3. This page takes teams that already exist and works out who plays whom, in what order, across how many rounds. A typical event uses both: divide the people into teams first, then paste the team names here to get the fixture list.
Can it do double elimination or a losers bracket?
No, single elimination only. Double elimination runs to 2n minus 2 games, or 2n minus 1 if the losers-bracket winner takes the grand final to a reset, and the bracket structure needs to track where each loser drops to, which is a different shape of problem from what this generates. For eight entries or fewer, drawing it by hand on paper is genuinely quicker than configuring software for it. Above that, use a dedicated bracket application.
How do I split this into groups?
Run each group separately. Paste group A's names, take the fixture list, then paste group B's and take that one. For sixteen entries in four groups of four you get six games per group, 24 in total, then paste the four group winners back in as a knockout for three more. Compare that against 120 games for a single round robin of sixteen and the reason group stages exist becomes obvious.
Does the schedule balance home and away?
The single round robin does not attempt to, because with one venue there is nothing to balance. Choose double round robin and the second half mirrors the first with the fixtures reversed, so every pair meets once at each venue. What the mirrored format does not do is prevent long runs of consecutive home or away games for a particular team; if that matters for your competition, take the pairings as generated and reorder the rounds by hand, since the set of matches per round stays valid however you sequence the rounds themselves.
Why does the bye go to the top seed?
Because a bye is an advantage — a free pass to round two with no risk and no fatigue — and giving it to the strongest entries is the convention that keeps the bracket consistent with seeding. If byes went to the weakest, the seeding would be actively working against itself. If your event awards byes some other way, or draws them randomly, choose Shuffle so the entry order is randomised before the bracket is built.