How to Make Fair Teams: Five Methods Compared

By Undercover Nerd Project · Published November 12, 2026 · 4 min read

Random draws, captains, snake drafts, rating-based balancing and algorithms compared, with what each is good at and where it goes wrong.

Everyone has been picked last at some point, and everyone has played in a game that was over in the first five minutes because one team was simply better. Fair teams make the game more fun for everybody, including the winners. Here are five ways to split a group, what each one is good at, and where each one goes wrong.

1. Random draw

Put names in a hat, or shuffle a list, and deal them out. It is quick, it needs no information, and nobody can argue that it was rigged.

The catch: random is not the same as balanced. Suppose 12 people play, 4 of them are strong, and you split them into two teams of six. The chance that the strong players divide 2 and 2 is only about 45 percent. About 48 percent of the time the split is 3 and 1, and about 6 percent of the time all four strong players land on one team. Random is a good choice when skill does not matter much, such as a casual game or an icebreaker, and a poor one when you want a close match.

2. Captains pick

Two captains take turns choosing players. It is familiar, it feels fair because the captains compete, and it makes the captains responsible for their teams.

The catch: it is socially costly. The last players chosen are always the same people, and a public ranking in front of the group can sting. Captains also tend to pick friends, which can leave the teams uneven. Alternating picks has a built-in advantage for the team that picks first, as the next method shows.

3. Snake draft

In a snake draft the pick order reverses each round: team A, team B, team B, team A, team A, team B, and so on. It corrects the advantage of picking first. Take eight players with skill ratings, from strongest to weakest, of 5, 4, 4, 3, 3, 2, 2 and 1. The total is 24, so a perfect split gives 12 each.

The catch: it only works with accurate ratings, it cannot handle rules such as keeping two players apart, and it gives you one result rather than a choice of equally good ones.

4. Rating-based balancing

Give every player a skill rating, then split so that the team totals (or averages) are as close as possible. This is the idea behind the snake draft, done on purpose. A simple scale such as 1 to 5 is enough. See how to rate skill for tips on doing it honestly.

The catch: totals can hide an imbalance. A team with one 5 and four 1s and a team with five 2s have the same total, but they will not play like equal teams. A good split balances the mix of skill levels as well as the total, so that each team gets its share of strong players.

5. Algorithmic balancing

A program can score every possible arrangement on several things at once, such as team size, the mix of skill levels, the gender balance and your own rules, and then search for the arrangement with the lowest imbalance score. Even 16 people have 12,870 ways to choose one team of eight, and the number explodes with more people and more teams. Nobody can check them all by hand, but a computer can search a huge number of them in a fraction of a second.

The catch: an algorithm is only as good as its inputs. Poor ratings give poor teams. It also finds many equally fair splits, so a good generator lets you press Shuffle to see another one instead of treating the first as the only answer.

Which method should you use?

SituationBest choice
Casual game, skill does not matterRandom draw
Fun, short, social, with a ritualCaptains pick (consider secret or random order)
Two teams, ratings knownSnake draft
Several teams, or you want gender or skill balancedAlgorithmic balancing
The same group every weekAlgorithmic balancing, reshuffled weekly

Practical tips for any method

More guides