Branching Paths: the Fractal Three-Arm Arcade Map with a 3-Team Spawn

Branching Paths is the arcade map that turns a free-for-all into three coordinated fronts: a 2480-by-2148 board of 2,103,911 land tiles where fractal waterways split the board into three arms from a central spine, 19 nations packed along the lower-center, and a 105-player FFA ceiling — half of The Box’s 210. It has been in the game since v0.33.0, and since v0.35.5 it is one of very few maps with a fixed 3-team spawn that fires 75% of the time. Real water exists at spawn, so ports and boats are available from minute one; the decision is whether to play the water arms or fight for the land spine, and whether to expect two teams or three.

Version boundary: when the map landed and the 3-team spawn arrived

Branching Paths shipped in v0.33.0, the 2026-08-03 release, under the new-maps section as “Branching fractals form waterways.” The map itself — its 2480-by-2148 canvas, its 2,103,911 land tiles, its 19 named nations, and its three-armed fractal geometry — has not changed since that introduction; the geometry has been identical in every release from v0.33.0 forward. What changed after launch is how the game treats the board, and the most important change for your decision-making is the 3-team spawn added in v0.35.5 (2026-10-26).

Before v0.35.5, Branching Paths behaved like every other arcade map that did not define a fixed team layout: when you queued into a 2-team game on it, the map fell back to random team spawns, the same as The Box and the rest of the arcade roster. That meant the map’s “three arms” were just geography, and the team count you got depended on the standard team-selection roll. In v0.35.5 the release added fixed team spawns for several maps — two-team spawns for Madagascar, Rio de Janeiro, Bab el-Mandeb, and New Zealand — and a 3-team spawn specifically for Branching Paths. The map’s definition now carries a special_team_count of 3, and the playlist treats that as a signal to force three teams most of the time.

This is the version boundary you should anchor on. If you have played Branching Paths on a build between v0.33.0 and v0.35.4, your experience was a random-spawn arcade board with no fixed team structure. From v0.35.5 onward, the same board can land you in a 3-team game where roughly 75% of the time all three arms are assigned to one team each, with 35 players per side on the full 105-player board. The lobby size itself (105-player FFA ceiling) is unchanged by that edit — it is driven purely by land-tile count and was set when the map’s size was fixed in v0.33.0 — but the mode you are most likely to queue into changed materially. When reading any older discussion or stream about Branching Paths, check the build date: anything before late October 2026 does not reflect the 3-team behavior this guide is built around. One more detail from the map’s definition that does not change between v0.33.0 and v0.35.10 is its multiplayer frequency of 7, which places it in the upper tier of the arcade pool for how often it can appear in a multiplayer rotation — higher than the base frequency of most arcade maps, so the board is not a rare pick. That frequency was set when the map shipped and has not been touched since, so the “you will actually see this map in your rotation” expectation is unchanged by the 3-team edit; only the mode you land in changed.

The board: a triangle split into three arms by fractal water

The map reads as a triangle with its apex toward the top and its base across the lower band, and the fractal waterways carve that triangle into three roughly equal arms that branch away from a central vertical spine. The decoded terrain puts the numbers behind the shape: of the full 2480-by-2148 canvas, 39.5% is playable land, 10.3% is open water, and 50.3% is impassable background. That last half is the key to the “fractal” description — the board is not a filled rectangle like The Box. It is a web of land corridors separated by water channels, and the water is what forces the board into arms.

The 19 named nations confirm the shape from the other direction. Their coordinates cluster along the central column (11 of the 19 sit within the middle third of the board’s width) and along the lower band (15 of the 19 sit in the lower third of the board’s height), with the centroid at roughly 50% across and 67% down — dead center, low. Only one nation, “Scorching Rays,” sits anywhere near the apex. So the named entities — the places the game will seed AI players, the places the early economy concentrates — are all down the middle, in the base of the triangle, not up the arms. That is not a coincidence: the arms are the expansion direction, and the spine is the contested direction. The lone apex nation, “Scorching Rays,” at the top of the triangle is the single spawn point that sits out of the central crowd — if the game seeds a named nation there, that player starts with the most room and the least immediate pressure, which is why the apex is a quietly desirable early position even though it is farthest from the board’s economic center.

The contrast with The Box is the whole point of the map. The Box is 100% land at spawn — no water, no ports, no boats until a water nuke carves the first crater. Branching Paths is the opposite: the water is already there, already branched, already forming the three arms. That means a navy is legal from the first tick. A port on one of the waterways lets you move a trade ship or a warship down the channel into an arm, and the channel is narrow enough that a single boat can contest it. The map’s identity is therefore “land spine + water arms”: you can play a land war down the central column, a naval war up the arms, or the two at once — and the arms are where the water advantage compounds because the channels re-branch and re-branch, multiplying your reach.

The lobby math: why Branching Paths runs half the size of The Box

The lobby ceiling is not a design number someone picked; it is computed from the board’s land area. The playlist counts the map’s land tiles, multiplies by 50 and divides by a million, then rounds to the nearest five to get the base player count for the largest size tier. Branching Paths has 2,103,911 land tiles, so 2,103,911 × 50 ÷ 1,000,000 is 105.19, which rounds to 105 — the 105-player FFA ceiling. The size tiers then split that base: the first tier (30% of the pool) takes the full 105, the 60% tier takes 75% of the base rounded to five (80), and the 30% tier takes 50% of the base rounded to five (55). So a Branching Paths lobby lands at 105, 80, or 55 players depending on the size roll, with 105 being the most memorable figure because it is the “full” board.

Put that next to The Box and the relationship is exact. The Box is 4,194,304 land tiles on a 2048-by-2048 square — roughly double the land area of Branching Paths — so its base player count is 210. Branching Paths is, in lobby-size terms, The Box at half scale: the same arcade family, the same land-tile-driven formula, about half the board, about half the players. That has three practical consequences. First, borders are shorter and early contact is earlier — on a 210-player board you can spend twenty minutes before the fronts meet, on a 105-player board the central spine starts touching near the opening. Second, the win threshold (80% of the map) is the same percentage but a smaller absolute area, so a decisive swing weapon or a collapsed front flips the map faster. Third, the 105 figure sits at the top of the “not crowded” band — the playlist only applies the “crowded” player cap to maps whose first-tier count is 60 or below, so Branching Paths never gets the crowded modifier; it is a large map by definition.

The team-mode math is where the map gets interesting. The playlist caps a team game’s player count at the smaller of 1.5× the base (rounded up) and the base itself, per size tier. On the full 105 tier that is min(ceil(105 × 1.5), 105) = min(158, 105) = 105 — so a full-size team game is still 105 players, and on the 60% tier it is also 105 (because 80 × 1.5 = 120, capped back to the base of 105). Only the small 55 tier drops, to min(ceil(55 × 1.5), 105) = min(83, 105) = 83. The upshot: team games on Branching Paths are the same 105 players as the FFA board, not a bigger number — the team mode concentrates the same population into fewer, coordinated sides rather than inflating the lobby. That is the opposite of what a naive reading of “team mode multiplies players” would suggest, and it is the single most common misread of this map.

The 3-team spawn: the mode that changes everything

The special_team_count: 3 on Branching Paths is the most decision-relevant property of the map, and it is the reason you should not treat it like The Box or a normal arcade board. The playlist keeps a separate pool of maps that define a special_team_count; when one of those maps is chosen, a 75% roll forces the team count to that special value instead of using the normal 2-team default. For Branching Paths that special value is 3, so roughly three out of every four Branching Paths team games you queue into will be 3-team games — three sides, not two.

The capacity check explains why this works at 105 players. A team game is only valid when each side has at least 2 players and there are at least 2 teams. Split the full 105-player board three ways and you get 35 players per team — far above the 2-player floor, so the 3-team split is valid at the full tier and at the 80 tier (80 ÷ 3 ≈ 26 per team, still valid). That means the full 105-player Branching Paths board can comfortably run as 3 × 35. The three arms of the board are then almost perfectly matched to the three teams: the playlist assigns each team a spawn zone, and on a three-armed board those zones line up with the three arms, so each team starts in one arm and fights outward along its channel toward the spine.

This is the fork that changes your entire plan. On a 2-team game you think about a single front and a single flanking route. On a 3-team game you are one of three sides, and the two other sides are not free agents — they are coordinated teams with their own navy, their own economy, and their own push up an arm. The central spine becomes a three-way junction rather than a two-front line: whoever holds the center can threaten the other two arms, and the two arm-teams will often converge on the spine to deny it. The tactical question stops being “which side do I flank” and becomes “which two of the three do I fight, and which do I leave for the other team.” If you queue into Branching Paths and the game lands at 3 teams, your opening moves should be the three-teams version of your normal plan — hold your arm, contest the spine with a team, and treat the third team as a variable you cannot fully control. If you want a two-front game on this board, you are out of luck most of the time: the 75% roll means the 3-team outcome is the norm, and the 2-team fallback is the exception.

Two scenarios: a 105-player FFA and a 3-team game

Scenario A — a full 105-player FFA (no fixed teams). You spawn into the central spine, which is where 15 of the 19 named nations sit. Your immediate neighbors are real players, not bots, because the named-nation density puts human spawns down the middle. The three arms are open to you: the waterways are yours to run because no team owns them. Assumptions: FFA mode (the 25% of games where the 3-team roll does not fire), you spawn on the spine with a named nation, no starting-gold modifier. In this scenario your plan is land spine, water arms: build a factory first (the spine is land, and the map has no port advantage that beats a factory on the central column in the early economy), hold your spine tile with a border wall, and push a trade ship or a boat down the nearest arm to establish a second economy line off the water. The spine is where the early game is won or lost — 15 named nations means 15 potential early neighbors and 15 potential early trade or attack vectors. The arms are where the mid-game is won: a boat that reaches the end of an arm can out-flank a land player who never went to water. The failure mode is committing all your power to the spine and losing the arms to a navy that started one factory earlier.

Scenario B — a 105-player 3-team game. You spawn into one arm (the 3-team spawn zones line up with the arms, so this is the most likely placement on a 3-team roll). Your team is 35 players, the other two teams are 35 each, and the central spine is unowned at the start. Assumptions: 3-team mode (the 75% of team games), you are in an arm team, standard starting gold. In this scenario your plan is own the arm, contest the spine: your team’s economy runs down your arm’s channel (water economy, ports on the waterway), and your team’s objective is to be the first to reach the spine before the other two teams do. The spine is the three-way junction — whoever holds it first can choose which arm-team to attack next, so the first team to the spine sets the terms for the other two. The 35-player team depth matters here: with 35 players you have the population to run a navy and hold the spine and defend the arm simultaneously, which is the luxury a 2-team split of the same 105 players does not have as cleanly. The failure mode is a team that commits its whole navy to the spine and leaves its own arm undefended, letting the third team take the arm’s economy and cut the first team’s supply line.

Failure and counter: when the central axis collapses

The map’s central failure mode is the spine collapse: a moment when the central column, which 11 of the 19 named nations and the centroid of the board occupy, is taken by one player or team and the three arms become supply lines to that center instead of independent economies. In an FFA, if one player walls the spine and holds the center, the arm players become tributaries — their water economies feed the spine player’s land economy, and the arms stop being a counter to the center and start being the center’s reach. The counter is to not let the spine become one person’s problem: in FFA you want at least two or three strong players contesting the spine so no single player holds it alone, and you want your own army to keep one foot on the spine and one foot on an arm so you are never purely a water economy that can be cut. In a 3-team game the spine collapse is the endgame: the team that holds the spine last wins the junction and can pick off the other two arms, so the counter is coordinated spine denial — your team’s navy holds the water approaches to the spine while your team’s land holds the central tile, and you do not let the other two teams meet each other on the spine without meeting you first.

The second failure mode is the arm overcommit: a player or team that pushes deep into an arm, establishes a port, and then gets cut off when the waterway behind them is taken. Because the fractal waterways branch, an arm is a cul-de-sac — if the channel connecting it to the spine is blocked, the arm economy is stranded. The counter is to keep a retreat line: do not build your port so deep that a single enemy boat can sever the channel behind you, and keep one boat as a mobile retreat/escort rather than committing all your navy to the deep arm. In a 3-team game this is also the “who fights whom” question — if you overcommit to an arm, the team that is not fighting you takes your spine approach while the team you are fighting holds the channel.

The third failure mode is the spine race stall in a 3-team game: all three teams reach the central junction at roughly the same time and none commits to holding it, so the spine becomes a no-man’s-land where every team probes but nobody takes. Because the spine is the three-way junction, the team that does not take it first concedes the choice of which arm to attack next — so a team that hedges too long is the team that ends up on the defensive. The counter is a committed spine hold with a reserve: one team unit takes the central tile early and holds it with a border wall and a boat, while the rest of the team economy keeps running down the arm. The cost of that hold is the team unit’s attention (it cannot also run the deep arm economy), which is exactly the trade the arm-overcommit failure punishes in reverse. The two failure modes — overcommitting to an arm and stalling on the spine — are the same decision from the other side: you must split your 35-player team between holding the junction and running the arm, and the team that gets that split right first wins the race to set the terms for the other two.

Mode and map choices: Branching Paths versus The Box

The honest comparison is that Branching Paths and The Box are the same arcade family at two scales, and the choice between them is a choice of water and mode, not of difficulty. Choose Branching Paths when you want: (a) real water at spawn, so a navy and trade ships are legal from minute one rather than gated behind a water nuke; (b) a smaller, faster board (105-player FFA) where the early game reaches contact sooner and a swing flips the map quicker; (c) the 3-team mode — this is the map where three coordinated fronts are the default rather than an exception, and it is the best board for playing a 3-team game with a real team; and (d) a board where the arms reward a water player who wants to out-flank along a channel. Choose The Box when you want: (a) a pure land war with no water and no boats until a nuke, so the game is a land-and-train economy with no naval variable; (b) the bigger 210-player board, where borders are longer and the early game is a slow snowball; (c) the 75% forced water-nukes modifier, which makes the first crater the map’s opening event and the only source of water; and (d) a two-front game, because The Box has no fixed team spawn and the team count falls back to the standard 2-team roll.

If your goal is specifically a 3-team game with your team, Branching Paths is the default pick: it is the arcade board where the 3-team outcome is the norm (75% of the time) and the spawn zones match the three arms. If your goal is the biggest land war in the arcade pool, The Box’s 210-player all-land board is the pick, and you should expect to fight the first water nuke rather than a navy. If your goal is a compact arcade map with water routes and fast contact but you do not want the 3-team default, Branching Paths is still the map — you just accept that 75% of team games will be 3-team and plan your team around that. The two boards share the land-tile lobby formula, the 80% win threshold, and the arcade nation seeding; what they do not share is the water, the scale, and the team mode — and those three differences are the entire decision. If you still cannot decide, the tiebreaker is the mode: if the queue you want to join is a 3-team game, Branching Paths is the only arcade board where that outcome is the default, so every other arcade map is strictly worse for a 3-team plan.

Quick reference: at a glance

This table is the mode decision without the full breakdown, and the row you will most often sit in is the 3-team row — that is precisely the reason you should plan Branching Paths like a three-front board rather than like The Box or a normal arcade map. Three numbers carry the whole call and are worth memorizing. First, 105: that is the FFA ceiling, half of The Box’s 210, and it is what makes the board play smaller and faster, with shorter borders and earlier first contact. Second, 35: that is the number of players per team on the full 105-board, and it is why a 3-team game on Branching Paths gives a genuine 35-deep roster instead of a thin 2-team split of the same population — a team that fields 35 can simultaneously field a navy, hold the spine, and defend the arm, which a 2-team split of the identical 105 players cannot do as cleanly. Third, 75: that is the chance the 3-team spawn fires, which means the 3-team outcome is the norm and the 2-team fallback is the exception, so the safe planning assumption is always three opposing teams. Pick Branching Paths over The Box when you want water routes, ships from the first minute, and a compact board; pick The Box when you want the biggest land war and the first water nuke as the opening moment. And remember the row you will most often be in: plan your team, your arm, and your spine approach as if you have three opposing teams, because that is what the game will most of the time give you.

For the broader map-choice context, see The Box map guide for the 210-player waterless board and the map-size and compact-mode guide for how the land-tile lobby formula works across the arcade pool.