How to make balanced football teams

Why doesn't picking captains work?

It's the method everyone grew up with: two captains take turns picking. It's quick, but it has problems. The best players tend to be captains or the first picked, so the teams look very similar every week. Each captain tends to pick friends before whoever would balance the side. And someone is always picked last, which is the worst way to start a game.

What about a random draw?

A draw is fair in one sense: nobody decides. But it ignores how each person plays. With few players, three strong ones landing together is enough for the game to be over in ten minutes. It's fine for playing with strangers; in a group that knows each other, it wastes everything you already know about each player. We compare them in detail in random vs balanced teams.

What should you rate in each player?

A single number ("they're a 7") mixes very different things. A fast player who never defends and a slow one who wins every tackle can have the same "level" and bring opposite things. It works better to rate four skills, from 1 to 10:

• Physical: pace, stamina, strength in the challenge.
• Defence: positioning, winning the ball, tracking back.
• Attack: dribbling, shooting, getting into the box.
• Vision: passing, reading the game, bringing others in.

With four axes you can check that both teams are even on each one, not just in total. One team with all the pace and another with all the vision can add up to the same and still make a very uneven game.

How should you spread positions?

Even in a casual game, people drift to where they feel comfortable. If one team gets everyone who defends, the other will only attack. Note each player's usual position (goalkeeper, defender, midfielder or attacker; many are two at once) and check that each team has a similar mix.

What do you do with goalkeepers?

If there are two goalkeepers, one per team, no exceptions. And since the keeper weighs heavily on the result, compensate: the stronger goalkeeper goes to the team that would otherwise be weaker. If there's only one keeper or none, people usually take turns in goal; in that case leave goalkeeping out of the balance.

How do you avoid the arguments?

Most arguments aren't about the split itself, but about who made it. Three ideas that help:

• Private ratings. When each organiser scores without anyone seeing, they score more honestly. Averaging several people smooths out individual biases.
• Several options. Offering two or three valid splits instead of one changes the conversation: it's no longer "your team", it's choosing between options.
• Let the group vote. If the group votes and the organiser confirms, the result belongs to everyone.

By hand or with an app?

If you'd rather do it by hand, this method is fair and works:

1. Rate each player on the four skills and note their position.
2. Put one goalkeeper on each team.
3. Sort everyone else by their average and deal them out in a zigzag: A, B, B, A, A, B, B, A…
4. Check skill by skill and position by position, and swap similar players until the gaps are small.
5. Keep together the people who need to be (for example, because they arrive in the same car).

It works, but with 14 players it takes a while, and you have to redo it every week with whoever turns up.

Teamix does exactly this, automatically. Each admin rates in private and Teamix uses the average; it spreads all four skills and the positions, places the goalkeepers and keeps together whoever you mark. It suggests several different options, the group votes (including people without an account, through the group's public link) and you confirm. More in the football team generator.