Splitting people into teams sounds easy until someone says "the same people always end up together" or "I always get picked last". Whether it's a pickup soccer game, group work in class or a team-building activity at the office, the way you make teams sets the mood for everything that follows. Here we compare the most common methods, show how to keep teams balanced and explain how to prove the split was random.

First: what does "balanced" mean?

When people ask for balanced teams, they can mean two different things:

Balanced in size: every team has the same number of people or, when the numbers don't divide evenly, sizes differ by one person at most. A good split always solves this.
Balanced in skill: the strongest players don't all end up on the same side. This depends on knowing the people, and chance alone doesn't guarantee it.

It helps to separate the two from the start. Randomness is what keeps anyone from claiming favoritism; skill balance, if it matters, is handled before the draw (more on that below).

Four ways to make teams (and their problems)

Captains picking one at a time

The schoolyard classic: two captains take turns choosing. Teams often end up balanced in skill, but there are well-known downsides:

Someone is always picked last, and that's no fun for anyone.
Captains tend to pick their friends.
With a big group, it gets slow.

Counting off "one, two, one, two"

Fast and needs nothing. The catch is that it depends on where people stand: two friends only need to leave one person between them to land on the same team. It looks random, but it isn't.

Names in a hat

It works and everyone gets it. But writing the names takes time, slips get lost and, once they're drawn, there's no record of how the split happened. If someone doubts it, there's nothing to show.

An online team generator

Paste the list, choose how many teams you want, and you have the split in a second. Done well, it has three advantages: it's fast with any group size, it keeps teams balanced in size and, if it uses a public seed, anyone can check afterwards that nobody hand-picked the teams.

Step by step with Cuki's team generator

The Cuki team generator is free and needs no sign-up. Here's how it works:

Paste the list, one name per line. You can copy it from a group chat or a spreadsheet.
Choose how to split: by number of teams (for example, 3 teams) or by people per team (for example, 5 each). Before you start, you see how many teams you'll get and their size.
Tap "Make teams". The list is shuffled with a public seed and dealt one by one: the first person to team 1, the second to team 2, and so on. That's why team sizes never differ by more than one person.
Copy the teams in one tap and paste them in the group chat.
If anyone doubts it, tap "Verify on the server": the calculation is repeated with the same seed and the same list, and it has to produce exactly the same teams.

It takes up to 500 names at a time, so it works for a single game or a whole school year group.

How many teams, and of what size? Examples

When the numbers don't divide evenly, what matters is keeping teams as even as possible. Some typical cases:

5-a-side soccer with 10 people: 2 teams of 5. The ideal case.
5-a-side soccer with 12 people: 2 teams of 6 (rotating subs) or, if you choose "5 per team", 3 teams of 4 that take turns.
A class of 25 students in groups of 4: 7 groups of 3 or 4.
A trivia night with 18 people in 4 teams: 2 teams of 5 and 2 teams of 4.

When you choose "people per team", the tool makes as many teams as needed so none goes over that number, and keeps them balanced. If you'd rather have bigger teams, choose the number of teams directly.

What if skill levels are very different?

Randomness doesn't look at who plays better. If your group has a big skill gap, there's a simple technique that keeps the draw and balances skill:

Split the list in two: the strongest players and everyone else.
Make teams from each list separately, with the same number of teams for both.
Pair the teams in reverse order: team 1 from the first list with the last team from the second, team 2 with the second to last, and so on. Pairing them in the same order (1 with 1) can stack each list's extra players on the same team.

Each team gets a similar share of each group, and nobody decides by hand who plays with whom. Another option is to choose the captains first (one per team) and draw everyone else.

How to prove it was random

Distrust shows up when a split leaves no trace. That's why every split on Cuki comes from a public seed: a long number generated on the spot that, together with the list exactly as you typed it, determines the result. With those two pieces of data, anyone can repeat the calculation and get the same teams. If you want the details, they're on the transparency page.

A practical tip: share the seed together with the teams. If someone later says it was rigged, the answer is already in the message.

Common mistakes when making teams

Duplicate names: if there are two "Johns", tell them apart (John P. and John G.). The tool warns you when it finds identical names.
Changing the list after the draw: if someone joins late, make the teams again with the new list instead of adding them to a team by hand.
Reshuffling until you like it: if you're going to draw, agree beforehand that the first result counts. Reshuffling until it suits you defeats the purpose.
Not explaining the rules: say upfront whether teams are made purely at random or with the two-list technique. Clear rules prevent arguments.

In short

To make teams without arguments you need three things: a random split, teams balanced in size and a way to check it. Try the team generator with your next list and, if instead of teams you need to pick just one person, use the name wheel.

How to Make Random Teams: Fair Methods Without Arguments