We have no solo film.
We have 18 hours of people playing alone.

Solo format

Chapter 7 Season 4 — Override, internal tag S42 — went live 20 August 2026 on patch v42.00. Solos are on the calendar this season, but they are not where the money is. Here is the whole shape of it, with the parts nobody has confirmed marked as such.

The Override competitive calendar

Every row below is an all-region online event unless the location says otherwise. Times are the session windows as listed; solo events are the highlighted rows.

Sources: docs/season-s42-override-plan.md (season facts, verified 2026-08-20 against Epic’s in-game Compete tab on launch day) and the schedule entries in lib/sample-data.ts, each carrying confidence: “confirmed”. Antwerp dates re-verified 2026-08-15 against BLAST, Lotto Arena and Liquipedia.

Which of our drop dossiers still describe live ground

Override rebuilt the centre of the same Chapter 7 island. Roughly ten of the thirteen named POIs carried over, so most of the 31 drop dossiers still describe terrain you can land on — but four of them do not, and a solo player planning a drop should know which.

Why the corpus cannot be filled from ranked or pubs

Solo vs duos film

Everything in this room was measured in duos. That does not make it useless for solos — it makes it partly useless, and the interesting work is saying which part. Below, each claim is tied to a mechanic and, wherever a number exists, to the games it came from.

What holds in solos

What inverts

1 · The trade almost never comes — and in solos it is exactly zero

The habit duos teaches worst is the fight you take because your partner can clean it up. In the Grand Finals there were 958 eliminations where the victim’s partner was still alive at the moment of death. Only 36 of them — 3.8% — were answered by that partner eliminating the killer within thirty seconds. When an answer did come it came fast (median 8 s), which is the point: either the trade is immediate or it never happens.

2 · Shorthanded play is where the endgame eliminations actually are

361 of the 2,380 Grand Finals eliminations (15.2%) were taken by a player whose partner was already dead. That share is not spread evenly — it climbs as the lobby empties, until in the last five teams nearly two eliminations in five are being taken by somebody playing alone.

Read the other way round: the closing minutes of a duos tournament are already being played substantially solo. If you only ever practise the paired version of the endgame, you have not practised the version that decides most Grand Finals games.

3 · Third-party maths gets worse, and it is a density problem

Counting every elimination that landed within 20 seconds and 150 metres of an earlier elimination involving neither the killer’s team — a proximity definition, not a mind-reading one — the Grand Finals look like this:

In duos a third party costs you a partner and you keep playing. In solos it costs you the tournament. The early-game figure is the one to train against: even when the island is still wide, one elimination in seven was already somebody arriving late to a fight they did not start.

4 · No rebuild, no revive, no refresh — and we cannot measure any of it

The mechanics that most obviously invert are the ones our parse cannot see. The decoded replay stream gives positions, eliminations, damage, storm circles and stop durations. It does not give builds, materials, heals, shields or reboots. So everything in this paragraph is a mechanic argument and is labelled as one: in solos nobody boxes you up while you heal, nobody carries the mats when you run dry, nobody rebuilds the wall you just lost, and a reboot card has nobody to bring back. Each of those changes the cost of taking a fight; none of them is a number we can show you, and we are not going to invent one.

5 · Being alone is survivable — three Grand Finals were won that way

Of the 766 duos that spent time shorthanded, 76.2% finished ahead of the number of teams that were still alive when their partner died, gaining a median of two placement spots while a player down. Read that carefully: staying alive means climbing as others fall, so this is not proof that going shorthanded is good. It is proof that it is not a reason to give up the game.

Three of the twenty-four Grand Finals were won outright by a player whose partner had been dead for more than five seconds:

Muz’s film page opens on exactly that game — NA Day 1, Game 3 — so you can scrub the last minute and watch a Grand Finals won by one player. It is the nearest thing to a solo endgame receipt in the room.

Loot Paths

The build, dated

Steps one and two are already done — they are what the Major 2 room is made of. Everything from step three needs a solo tournament to exist first.

1. Done · Aug 10
   The stop detector, proven on 24 games

   53,258 stops are already logged across the Grand Finals — each with a position, a duration and the point of interest it sits inside. The measurement Loot Paths needs already works; it has only ever been run on duos.

2. Done · Aug 10
   Per-spot pages to hang the routes on

   The 31 drop dossiers and the consensus opening routes are the surface a loot path lands on. No new page architecture is needed — a loot path is a new section inside a dossier.

3. Next · Sep 5
   Pull Victory Cup #1 the night it runs

   Epic keeps replay chunks for roughly three months, and the download is the only irreversible step. Cup one is pulled and archived on the night of 5 September whether or not anything is built on it yet.

4. Sep 5–19
   Three cups, then count the sample per spot

   A route is only a consensus if enough players walked it. The gate is per POI, not per event: a spot gets published when it has enough landings to show agreement, and the spots that do not clear it stay blank rather than showing a route derived from four people.

5. After cup 3
   Re-fit the stop and activity definitions for solos

   Half of the S41 activity model is about two players — duo split, who took the fight, proximity contact between teammates. Those labels mean nothing with one player, so the model gets re-fitted against solo games rather than reused. Same bar as Major 2: labels publish only after they check out against a cup the model never saw.

6. Late Sep
   Publish per-POI loot paths, versioned to the Override map

   Loot pools and the island change every season, so a loot path carries the season it was measured in and expires with it. This is an Override-era view from day one; the S41 archive never gets one.

What you will be able to do with it

Six things, listed in full on the Loot Paths room page along with the one that matters most — what version one still will not know. The short version: it names the building, not the POI; it gives you the clock at which the placing players stopped looting; and it cannot tell you what came out of a chest, because the replay stream does not carry inventory. That last limit is permanent, not a v1 gap.

How it will be sourced

Where this lives. Loot Paths is the room page — what it will be, what it will not be, and the nearest views that already exist. This tab is the build plan behind it: the dated steps, the sourcing rules and the solo calendar the whole thing is gated on. The argument about the design runs in the Elite Arena Discord.

Notes & method

This page mixes three kinds of statement: numbers measured from replay data, facts taken from a published calendar, and analyst opinion. They are not interchangeable, so here is which is which.

Every claim on this page, and where it came from

How to check the work

Everything in the measured rows comes from twenty-four JSON files that ship with this site. Each index.json holds players[] (with team, placement, kills), deaths[] as {p, t} in seconds, events[] as {t, k, v, x, y} for every final elimination, and zones[]. A duo is two players[] rows sharing a team; a shorthanded stint is the gap between their two death timestamps. No other input was used, and no number on this page was rounded up.