Roguelite Tuesday — Pocket Raid Captain

The trigger

The Game Business is reporting that MMO interest is rising again, which means raid-role fantasy still has pull even outside the traditional subscription crowd. At the same time, the broader industry mood in Axios and Game Developer is about contraction, layoffs, and teams avoiding content-heavy bets. On the player side, the Midnight Suns discussion is a useful warning: people like squad-based combat, but they do not necessarily want hours of connective tissue around it. That points to a small, reusable structure: compress the satisfaction of being a tank/healer/support/DPS in a raid into short solo runs with boss mechanics as the content backbone.

The idea

A single-player roguelite where each run is one procedural "raid wing" made of 3-5 pulls and a boss. You choose a role at the start—tank, healer, support, or DPS—and command a small AI party through telegraphed mechanics: stack markers, interrupts, add waves, cleanse checks, soak zones, enrage timers. Between encounters, you draft cooldown cards, passive relics, and one party order upgrade that changes how your AI teammates behave. The fantasy is not "be an MMO"; it is "solve raid problems under wipe pressure" in 15-25 minutes.

The core loop

Move through a compact encounter node map. Enter a pull. Your character auto-attacks or has a basic spammable action, while your real decisions are cooldown timing, positioning, target priority, and issuing one-button squad orders like Focus Adds, Hold Defensive, Spread, Stack, or Save Interrupts. Survive the mechanic pattern, collect loot, draft one new tool, then face a boss whose script remixes the same mechanic language with higher stakes. Meta-progression unlocks new role kits, relic pools, and boss modifiers rather than permanent stat inflation.

Why now

There is current demand for MMO-style role fantasy without MMO scheduling. The MMO resurgence signal says the appetite is back; the player does not need 200 hours of questing to want a clean healer or tank fantasy. A run-based structure also fits today's smaller-budget reality better than trying to build a "tiny MMO." It is streamable because bosses create readable failure stories. It also side-steps one common roguelite complaint seen in the autobattler feedback thread: players dislike feeling arbitrarily constrained. Here, the drafting should increase options run over run, not lock them down—especially around cooldown combinations and squad commands.

Scope

First playable milestone: one role, one biome, three trash encounter templates, two bosses, eight cooldown cards, twelve relics, and four squad orders. One full run should take 15 minutes and already answer the key question: is solving mechanics with a semi-autonomous party fun more than once? Art can be minimalist top-down or side-view silhouettes with strong telegraphs, role-color coding, and readable floor markers. For Early Access, the realistic expansion is four roles, 4-6 bosses, 30-40 cards/relics total, and mutators that reshuffle mechanic cadence.

What kills it

If the AI party feels dumb or unfair, the whole project dies. Raid mechanics only feel good when failure is legible. If players wipe because the healer bot stood in fire or the interrupt order was ignored, they will blame the game, correctly. The other kill shot is overbuilding MMO trappings—hub areas, dialogue trees, gear spreadsheets—instead of shipping a boss-mechanics toy box. Do not compete with WoW or FFXIV on world scale; compete on readable encounter design in short sessions.

Pick one role and one boss this week. If the wipe feels fair and the second attempt feels smarter, keep going.