How lobbydiff works

When you search a player, we pull their recent ranked matches from the Riot API and check the current rank of every opponent in each lobby. By averaging those opponent ranks across your games, we estimate where the matchmaking system thinks you belong - your hidden MMR.

How the MMR estimate works

For each of your last 20 ranked games, we look up all 7 opponents and get their rank (tier, division, and LP). We convert each rank to a numeric value and compute the lobby average. The difference between your rank and the average rank of your lobbies is your "delta."

A negative delta means your lobbies are lower-ranked than you - the matchmaking system may be placing you in easier games than your rank suggests. A positive delta means you're being matched against stronger opponents - your hidden MMR is likely higher than your displayed rank.

Confidence levels

The confidence indicator reflects how many games were used in the estimate. More games means a more reliable picture. With 15+ ranked games, the estimate is marked as high confidence. With fewer games, take the result with a grain of salt - a few unlucky or lucky lobbies can skew it.

Rank data is refreshed automatically for all Emerald+ players across all 15 regions:

Match history updates are prioritized by rank and activity - higher-ranked players see new matches faster. Leaderboard data refreshes every 10 minutes.

The Challenger leaderboard shows every Challenger player in each region, updated every 10 minutes with live LP, wins, losses, and win rates pulled directly from the Riot API.

Lobby Average

The "Lobby Avg" column shows the average rank of opponents this player has been matched against in their last 50 games. It gives you a sense of how strong their lobbies are - Challenger players with a lower lobby average are getting placed into weaker games.

Top Carries

The carry icons show the 7 most frequently played units across a player's last 50 ranked games, ordered by play count. Only units that had 3 items equipped are included - filtering out non-carry board units. The 3 items shown under each unit are the individual items most commonly equipped on that unit across those games, displayed independently. They are not a recommended 3-item build - they represent the 3 most frequently seen items on that carry, which may come from different games and compositions.

Small regions and matchmaking

In regions with fewer players like OCE, JP, or TR, Challenger players often get placed into lobbies with Master or Diamond opponents simply because there aren't enough high-ranked players queuing at the same time. This makes their lobby average look much lower than their actual skill level. A Challenger player with a "Grandmaster 200 LP" lobby average on OCE isn't necessarily worse than one with "Challenger 900 LP" lobbies on Korea - they just have a smaller matchmaking pool.

Opponent rank timing

When we compute lobby averages, we use each opponent's rank from around the time the game was played, not their current rank. Since we scan all Diamond+ and Emerald+ players every 10 minutes, most opponents have a rank snapshot within 10 minutes of the game. This means lobby averages are typically accurate to within 5-15 LP of the true value.

Some opponents show as unranked

If we can't find a rank for an opponent, they're excluded from the lobby average calculation. This can happen with brand-new accounts, players who haven't placed in ranked yet, or rarely due to API errors. The lobby average is calculated from the opponents we can resolve - if only 5 out of 7 opponents have known ranks, the average is based on those 5.

First search is slower

The first time a player is searched, we need to fetch all their match data and resolve every opponent's rank. This can take several seconds. After that, the data is stored and subsequent searches only process new games played since the last lookup - usually instant.