Team Splitter

Paste a list of names and instantly split everyone into balanced random teams. Pick how many teams you want or how many players per team, then reshuffle until it feels fair.

How to use

  1. Paste your list of participant names.
  2. Choose how many teams you want, or how many players per team.
  3. Shuffle to get random teams — reshuffle as many times as you like until it feels fair.

When the head count does not divide evenly, the leftover people are handed out one at a time from the first team onward. That keeps the gap between the largest and smallest team at one person.

In depth

Random and balanced are different goals

Splitting a group serves one of two purposes.

Random assignment — so nobody can complain. Fairness is what matters.

Balanced assignment — so the game is close. Enjoyment is what matters.

They conflict. Pure randomness can stack the strong players on one side, and balancing requires somebody to judge ability.

Pick by occasion: random for social play, balanced for anything competitive.

When the numbers don't divide

Eleven people into two teams is 6:5. This tool distributes as evenly as possible.

In actual games, the spare person commonly takes on refereeing, scoring or rotating in.

In sports where numbers translate directly into advantage, putting a stronger player on the smaller side compensates.

Changing the number of teams often reads better than tolerating an odd one out — eleven people split 4:4:3 across three teams feels less awkward than 6:5 across two.

When some assignments have to be fixed

Pure randomness is often impractical.

Some people shouldn't be together: those who don't get on, or who have already been paired repeatedly.

Others must be together, like someone needing an interpreter and someone who can interpret, or a beginner and a guide.

Then there is role balance: one goalkeeper per side, or a required position in each team.

With constraints like these, fix those placements first and randomise only the remainder.

Repeated splits produce repeated pairings

Over several rounds the same combinations recur. It is expected from randomness, but people experience it as "that team again?".

Smaller groups make it worse: splitting eight people into two teams has only 35 distinct arrangements, so repetition arrives within a few rounds.

The fix is a round-robin schedule — plan the combinations in advance so everyone mixes evenly.

Re-randomising each round never solves it. If you expect several rounds, plan them all before the first.

How many teams

The sport decides it. Football is eleven a side, basketball five, volleyball six — the group has to be a multiple.

Discussion groups work best at four to six. Larger and someone stops contributing; smaller and there are too few perspectives.

Even numbers of teams simplify games, since an odd count means one team sitting out or receiving a bye each round.

When the count is awkward, changing the number of teams usually helps more than forcing the split.

Computed on your device

Participant names and assignments are not transmitted; everything happens in the browser.

The roster persists in browser storage so a recurring group does not need re-entering.

Clear browser data if you used a shared device.

FAQ

How are teams assigned?

The name list is shuffled randomly, then split according to the number of teams (or players per team) you chose. No one gets an unfair advantage.

What happens if the group doesn't divide evenly?

Any leftover players are spread out evenly, so team sizes stay as close as possible.

Can I reshuffle if I don't like the result?

Yes — reshuffle as many times as you want until you get a split you're happy with.

Method and sources

What this tool bases its numbers on, and how far those numbers go.

What this tool is
A random split into near-equal groups. It balances head count and nothing else — it knows no skill, position, history or preference.
Formula
Participants are shuffled with a Fisher-Yates pass and dealt round-robin into the requested number of teams. That makes team sizes differ by at most one, and every assignment consistent with those sizes equally likely.
Worked example
Nine people into two teams gives 5 and 4: the shuffle decides who is where, and the remainder lands on the earlier team, so sizes never differ by more than one.
Limitations
  • Balanced by count is not balanced by strength. A random split will happily put every strong player on one side, and does so as often as chance dictates.
  • Math.random is not cryptographic. The ECMAScript specification promises only approximate uniformity, so a determined participant could in principle predict or reproduce a split.
  • Re-rolling until the teams look fair is not randomness any more. Whatever you choose to accept is the real selection rule.
  • It cannot honour constraints — keeping two people apart, fixing a captain per side, respecting positions. Those need a different procedure.
Effective date
Last reviewed

Related tools

  • Marble Race Roulette — Marbles roll down the board to pick a winner.
  • Ghost Leg (Ladder) — Randomly match names to outcomes with a ladder.
  • Plate Balance Game — Move the pole with your mouse and keep the plate up. Longest survival wins — settle a bet with it.
  • Darts Game — Throw at a real dartboard and play 01, Cricket or Count-Up against friends.