How to make pickup soccer teams fair (and end the argument)

Captains picking alternately produces decent balance and terrible vibes — someone is always picked last, and the captains' friendships leak into the draft. Random splits are socially painless but competitively chaotic; you'll get lopsided games two weeks out of four. "Organizer decides" gives the best single-game balance and the worst long-term position: every loss becomes evidence that you stacked the teams.

Every method above fails on the same input problem — nobody agrees how good anyone is, and self-assessment is comedy. The fix is peer rating: the people who actually play with you, rating you across real dimensions (pace, shooting, passing, dribbling, defending, physical), with the results applied anonymously and in aggregate so no single opinion — and no grudge — moves a number.

In FC Pickup, only teammates who checked in and played can rate, ratings apply once there's a quorum, and they move a player card the whole group can see. The data gets more honest every week, because it's built from games that actually happened.

With real ratings, team-splitting stops being politics and becomes arithmetic. Lay both squads out on a pitch view where everyone can see who's where, and if the split looks off, shuffle and re-deal. When someone grumbles about stacked teams, the answer is a screen, not a defense of your judgment.

One habit worth adopting regardless of tooling: balance across positions, not just total skill. Five decent defenders will lose to a mixed side every time. Spread the keepers and the runners first, then even out the rest.

Groups that record results get better balance over time, because the W-D-L record exposes systematic lopsidedness. If Blue wins four straight weeks, your splitting method has a bias; fix the method, not the memory of who won. Locking results shortly after each game (24 hours in FC Pickup) keeps the record from being litigated later.