How Team ELO Works (2v2, Doubles & Team Matches)

By the TrackMyElo team · Published July 2, 2026 · 6 min read

Head-to-head ELO is simple: one winner, one loser, points change hands. Team games — foosball doubles, 2v2 basketball, 5v5 video games — add a genuinely hard question: who gets the credit? If your duo wins, was it you or your partner? This guide explains the two ways to rate team play, the math that makes them work, and which one fits your group.

Option 1 — Rate the team as a unit

The simplest approach: treat each fixed pairing or squad as a single “player” with its own rating. “Alice & Bob” is one entry on the ladder, and their rating moves exactly like a singles rating would. This is the right call when the same people always play together — a regular doubles pairing, a set five-a-side squad — because the chemistry of a specific lineup is the thing being measured.

The weakness is obvious the first time someone’s partner is out sick: a fixed-team rating tells you nothing about “Alice & Carol.” If your lineups shuffle, fixed teams multiply endlessly and none of them accumulate enough games to mean anything.

Option 2 — Rate individuals from team results

The more flexible approach: every player keeps a personal rating, and team matches update all of them. The mechanics are a small extension of normal ELO:

1. Average each team’s player ratings to get a team rating.
2. Compute the expected score from the gap between the two team ratings — exactly like a 1v1.
3. Apply the resulting points change to every player on each team: winners all gain, losers all lose.

So if a 1050-average duo beats a 950-average duo, the winners each gain what a single 1050 player would for beating a 950 player. An upset by an underdog team pays out more, a favorite grinding out an expected win pays out less — the core ELO logic carries straight over.

Why rotating partners sharpens individual ratings

At first glance, updating both teammates identically seems like it can’t separate the carrier from the passenger — and over one night, it can’t. The separation comes from variety. A strong player keeps winning whether paired with the office champion or the newest hire, so their rating climbs across all pairings; a weaker player only wins with strong partners, and their losses with weak partners pull them back down. Over a few weeks of shuffled lineups, each player’s rating converges on their own contribution. If you always play with the same partner, the system genuinely can’t tell you apart — which is exactly when fixed-team rating (Option 1) is the honest choice.

Which option fits your group?

• Same lineups every week → rate fixed teams. The pairing is the player.
• Whoever’s free plays → rate individuals from team results, and rotate partners deliberately.
• Both formats at once? Run them as separate leagues. Doubles skill and singles skill are different things, and one blended number describes neither.

Setting it up in TrackMyElo

1. Create a league and enable team matches in the settings.
2. For fixed teams, create the teams on the Teams tab; for rotating lineups, pick the players on each side when you record the match.
3. Keep the K-factor at 32 to start — team results are a slightly noisier signal than singles, so avoid cranking it higher. See how to choose a K-factor.

For a worked example of team rating in a specific game — including the offense/defense wrinkle — see how to rank foosball players.