Data updated
The Network
Two players are linked if they played a map on the same team. Since 2021 that makes one web of 1,140 players and 4,691 teammate pairs. See who sits at its centre, or trace the chain between any two players.
Era
Connect two players
From
To
Closest to everyone
How the network is built
Two players are linked if they played the same map on the same team. Sharing an org isn't enough: two players at the same team five years apart never actually played together, and counting them as linked would reward passing through a storied roster.
- Hops. The average number of links from a player to everyone else in the network, which is the Bacon number idea averaged rather than measured to one target. The scene averages 5.24; Demon1 is the lowest at 3.60. The average number of links from a player to everyone else in the network, which is the Bacon number idea averaged rather than measured to one target. The scene averages 5.00; Demon1 is the lowest at 3.33.
- Breadth beats volume. The players with the most teammates are rarely the most central, because a long career in one region builds a big but tightly packed circle. One stint abroad is worth more than years of new faces at home.
- Max. The distance to the single furthest player. It sounds like the better measure but it can't rank anyone: across 1,140 players it takes only six values, so most of the table ties. It's also decided entirely by whoever is most remote, so one obscure roster can move it. The ranking uses the average, which is measured against everyone and so can't be swung by one straggler. The distance to the single furthest player. It sounds like the better measure but it can't rank anyone: across 623 players it takes only seven values, so most of the table ties. It's also decided entirely by whoever is most remote, so one obscure roster can move it. The ranking uses the average, which is measured against everyone and so can't be swung by one straggler.
- Wtd. The same calculation with links weighted by how many series a pair played together, so a four-year duo counts as closer than two players who shared one map. It barely changes the order, which is the point: the ranking doesn't depend on picking a weighting.
- Eras. The '23+ view is a separate graph, not a filter: dropping a player's later teams changes everyone's distances, not just theirs. 2023 is the split because franchising stitched the scene together in one offseason. There is no 2021-22 view because that graph barely holds together: its largest group contains only 26% of the era's players, so "closest to everyone" would mean closest to a quarter of them. In the '23+ window it's 94%.
- Who's missing. 87 players sit outside the main web, in 15 islands. 14 are a single team's roster whose players never played a map for any other team in the data; the largest, 16 players across 4 teams (FULL SENSE, Made in Thailand, NG.Black, Sharper Esports), is held together by players who moved between those teams, none of whom ever shared a map with anyone in the main web. With no path to anyone here, they're left out of the ranking. 39 players sit outside the main web, in 7 islands. 6 are a single team's roster whose players never played a map for any other team in the data; the largest, 8 players across 2 teams (Eternal Fire, ULF Esports), is held together by players who moved between those teams, none of whom ever shared a map with anyone in the main web. With no path to anyone here, they're left out of the ranking.