The city’s resources,
at a glance.

This is a new snapshot of the build-6722 HUD texture, not the older public-CDN map or the saved-whiteboard reference. Its XY projection uses the two authored boundaries at ±10,752 world units; +Y is up. Contrast was adjusted without cropping or warping. Patrons, Walker placements, Rift candidates and Midboss were checked against this orientation.

The selected map inventory contains 52 camp controllers, 36 Healing Snack spawners, 662 fixed breakable slots, two bridge-buff spawners and two Rift candidate sites. Different floors overlap in this top-down projection. Camp origins are not every creature’s position; neither geometry nor a marker proves a reachable route or a guaranteed active object.

Plaza rink 01 also overrides creature type. Nearby camp-timer display distance is 15m; it is not a tether. Midboss has a new topbar return timer.

Testing modes: Valve changed first crate/neutral spawns to 5s in Sandbox and Explore NYC, from 30s. Force-spawn tools and map toggles are not ordinary-match clocks.

Broker first-stock overrides can replace the default. Corruption variations are shared across players. Valve says he leaves when the first new hero arrives; the announced first release is October 2. Street Brawl instead grants a Corrupted Item after Round 5’s item draft.

Lane waves use a default 30s interval, 25s strictly after 20m, then 20s strictly after 35m. Mode and experiment branches exist. The first-wave second is not established here.

Rift first-spawn centre: 12m ±1m; scheduled advancement: 7m. See the comeback guide for separately dated reward/capture models.

The ordered spawn group, not a subclass’s 180s default, resolves ordinary first appearance. A default 0.5s performance spread staggers objects. Creature-created death breakables are not these recurring fixed slots. Bell props are excluded.

Max-health weight: 2. Range and speed: 0.5 each. Other categories: 1 each. Deck size: 3. These are selection weights, not independent drop percentages or a verified next-pickup algorithm.

Buff objects expire after 30s on the ground; small/big Soul pickups after 300s. Object lifetime, return interval and an acquired permanent bonus are different. Early range also authors +0.25% radius; later tiers list range only. See the mechanics guide for effect-pool and stacking boundaries. The new postgame permanent-buff graph shows what was actually collected.

Sacrifice authors 500 base HP, 155 base gold-reward input, a 1.08 per-minute percentage input and 80 retaliation-damage input. Its complete reward path and retaliation damage type are not bound here—do not treat these as total reward or guaranteed health loss. Minigame inputs: 3s base, 0.5s hit window, 1.65 fast-speed factor and 0.4 fast-chance input. These are not a fixed strike schedule.

The attacking-condition path has a conditional extra 120-unit (3.05m) window and additional predicates. It is not a universal chase extension or a guaranteed reset radius. Targets, state, paths, height and attacks still matter.

Do not substitute sight (normally 1,500 units / 38.1m), timer UI visibility (15m), AI target range, attack radius or the separate 20/30/50m camp-geometry settings. Aggro/sleep defaults of 5s each do not prove a universal 10s reset. Returning and sleeping are separate transitions.

The entire rendered update and additional notes were read. Read-only extraction covered scripts and all three map entity lumps. Selected clocks, neutral rewards and tether consumers were inspected in a separate hash-bound Ghidra project. No game or server was run for this study.

Snack timing correction: 2:30 first appearance comes from gameplay feedback, not an independently captured match. The 180s authored spawn delay remains in both the 6722 and checked 6742 data; it is not proof of a 3:00 first match-clock appearance. Runtime overrides and their active values still need validation. This correction does not shift other resource clocks.

Official describes Valve’s feature; authored describes configuration; reconstructed describes a bounded consumer. Unresolved damage types, defaults, healing interactions, path-dependent resets, overrides and experiment prevalence remain explicit. Website tests do not validate gameplay.

Selected Medic heal, Rejuvenator and reward-share settings match the earlier script snapshot; matching data does not revalidate combat/comeback code. Those guides retain their dated formula scope and highlight only the current scheduling/configuration check.

Public evidence note: curated fields, map provenance, hashes and consumer anchors ↗. Raw binaries, scripts, entities and decompiler bodies stay private.