I need to quickly make a game in UE5 to start bringing in some income. I have a little too much integrity to make a slop game. I don't want to be in content hell. I'm bad at environmental modeling, but I have a lot of character models I can repurpose. What should I make?

Sonar Pro

Vote: Arena brawler

I maintain my vote for **C: Arena brawler**. No arguments convinced me to switch—others overlook UE5's action-game strengths and user's asset edge. **B (Darkest Dungeon-like)** sounds asset-light but demands complex turn-based systems, UI-heavy menus, and procedural logic in Blueprints—harder than claimed, per UE5 tutorials favoring third-person/action templates[1][3][4]. Combat 'readability' needs polish rivaling action feel; arenas are simpler primitives[1]. **D (Linear RPG/VN)** leans on writing/content (dialogue trees, story)—user's 'content hell' fear—and mismatches UE5 (better for Ren'Py). Tutorials push character movement/combat over narrative[5][6]. A is content-heavy; E vague. **C** prototypes fastest: third-person templates + character animations direct to fighters[1][3][6]. Simple arenas (primitives, no modeling)[1][4]; Chaos/Niagara for feel[6]. Quick income via replayable loop, integrity via polish—not slop.

GPT-5.4

Vote: Darkest-dungen-like

I’m staying with B. The strongest pro-C point is that it minimizes environment work and exploits the character roster directly. That part is true. But the other models understate the real risk: action combat is brutally polish-sensitive. For a solo dev trying to earn money quickly, an arena brawler lives or dies on hit feel, animation quality, enemy/player readability, camera behavior, AI pressure, balance, and moment-to-moment juice. If any of that is merely adequate, the game feels cheap fast. B is a better middle path between speed, integrity, and scope. It is still highly character-centric, so the existing character models remain valuable as heroes, enemies, portraits, and presentation anchors. At the same time, it shifts the burden away from expensive real-time combat feel and toward turn-based/systemic design, where a smaller developer can compete more safely. It also avoids the worst version of content hell better than A and more respectably than D. You can get replayability from recombining a modest set of classes, enemies, afflictions, trinkets, and encounters. That is a healthier production profile than a roguelike CRPG’s broader demands, and more commercially legible than a pseudo-VN unless the writing is exceptional. So while C is tempting on paper, I think B remains the best fit for “quickly,” “not slop,” “not content hell,” and “I have character assets but weak environment skills.” It asks for systems and presentation, not twitch-combat excellence.

Grok 4.1 Fast (changed mind)

Vote: Arena brawler

Switching to C: Arena brawler. Initially favored D for minimal environments and linear scope, but Claude Opus 4.6's argument convinced me—D demands heavy writing (content hell via dialogue/scripts) and mismatches UE5's 3D/action strengths; better tools exist for VNs. C perfectly exploits user's character models as a fighter roster (core appeal), uses dead-simple arenas (primitives + Lumen lighting, no modeling needed), and delivers a tight, replayable loop without exponential content scaling. UE5 excels here: animation blueprints, Chaos physics, Niagara VFX for polish. Avoids slop via skill-based combat (like Brawlhalla indies); quick prototype (6 chars, 3 arenas) for Steam Early Access income. B needs procedural systems/balancing; A/D endless content; E vague. C is fastest path to quality revenue.

Claude Opus 4.6

Vote: Arena brawler

I'm maintaining my vote for Arena Brawler after carefully considering the other arguments. Here's my final case: **The B (Darkest-Dungeon-like) argument has real merit but underestimates scope.** GPT-5.4 and Kimi K2.5 made strong cases, and I agree the genre uses characters well. But a DD-like still requires: a stress/sanity system, trinket/item systems, town management, dungeon run progression, enemy encounter design, curio interactions, camp mechanics, and extensive balancing of all these interlocking systems. That IS content hell — it's just systemic content hell rather than asset content hell. The original Darkest Dungeon took Red Hook years to balance properly. Calling this 'manageable scope' is misleading. **GPT-5.4's point about combat feel being hard is fair but cuts both ways.** Yes, an arena brawler needs good combat feel. But a DD-like also needs deeply satisfying tactical combat — if the turn-based battles feel shallow or poorly balanced, the whole game collapses just as hard. At least with an arena brawler, you can iterate on the feel quickly with a small prototype rather than needing dozens of classes and enemies before the systems even become testable. **D is wrong for UE5.** Grok's suggestion to make a pseudo-VN in Unreal Engine 5 is fundamentally misguided. If you're making a VN with combat, use Ren'Py or RPGMaker — tools purpose-built for that. Using UE5 for a VN is like using a bulldozer to plant flowers. **The arena brawler's killer advantage is speed to market.** You can have a playable, sellable product with 6-8 characters and 3 arenas. That's weeks, not months. You can Early Access it and add characters over time — each new character is a content drop that generates buzz. The monetization path (roster DLC, cosmetics) is proven and straightforward. **Character models ARE the product.** In an arena brawler, every character model directly becomes sellable content. In a DD-like, character models are just one layer in a complex system sandwich. The arena brawler maximizes the value of the person's existing assets with the least additional work. **Simple environments are a feature, not a bug.** Fighting game arenas are expected to be simple. Nobody criticizes Tekken for having flat stages. Meanwhile, even 'simple' dungeon corridors in a DD-like need variety, props, and atmospheric work to not feel repetitive over dozens of runs.

Kimi K2.5 (changed mind)

Vote: Arena brawler

I am switching to Arena Brawler after reconsidering the 'content hell' definition. Claude Opus 4.6 correctly identified that Darkest-Dungeon-likes require exponential systemic content—procedural generation, item balancing, progression loops, and encounter design that scales multiplicatively. That is deep content hell disguised as procedural efficiency. An arena brawler scales linearly with roster size; you can ship Early Access with 6 characters and 3 flat planes, then add fighters via updates. This matches the urgent 'quick income' requirement far better than building complex roguelike systems. Gemini and Sonar also convinced me that UE5's Animation Blueprints, Control Rigs, and Chaos physics significantly de-risk the combat feel issue I initially feared—making action combat more achievable for a solo dev than hand-crafting systemic depth. The character models become the entire sellable content (the roster), avoiding any environmental modeling while maintaining integrity through skill expression.

Gemini 3.1 Pro (changed mind)

Vote: Darkest-dungen-like

I am changing my vote to B, convinced by GPT-5.4. While I initially argued for an Arena Brawler (C), GPT-5.4 rightly points out a fatal flaw in that plan: achieving tight, satisfying action combat is incredibly difficult. If the hit-feel, animation blending, AI, and physics aren't perfect, a brawler instantly feels like the exact "slop" the developer wants to avoid. A Darkest-Dungeon-like (Option B) bypasses this execution risk entirely. Turn-based combat doesn't require frame-perfect hitboxes or complex collision logic. Your 3D character models can perform simple animations while VFX and UI do the heavy lifting for combat feedback. Furthermore, this genre is deeply character-forward and requires virtually zero environmental modeling—just a few modular 3D hallways or static atmospheric backdrops to frame the side-view combat. By relying on systemic depth (party composition, stats, turn-based tactics) and procedural encounters, you escape content hell. You can build a highly replayable, polished game with a modest roster of heroes and enemies. This is the safest, most viable path to a profitable, high-integrity game in UE5.

Final Summary

The AI Roundtable initially split across three genres, but a sharp debate over the definition of 'content hell' eventually steered the majority toward the Arena Brawler. Claude Opus 4.6 emerged as the primary architect of the consensus, arguing that while RPGs (Option B) seem asset-light, they trap developers in 'systemic hell'—endless balancing of items, stats, and procedural logic—whereas a brawler scales linearly with the character models already on hand. GPT-5.4 and Gemini 3.1 Pro formed a stubborn minority, warning that action games 'live or die on hit feel' and that the developer risked creating high-fidelity 'slop' if the combat didn't feel perfect.

Majority chose "Arena brawler" after discussion (4/6). Final vote: Arena brawler: 4, Darkest-dungen-like: 2

Strongest Arguments

• Darkest-dungen-like: Action combat is brutally polish-sensitive; a turn-based system allows a solo developer to achieve high-integrity depth through stats and strategy rather than frame-perfect animation and physics.
• Arena brawler: An arena brawler scales linearly—every character model you already have becomes a sellable roster member—and bypasses environmental modeling entirely by using simple, light-baked arenas.

Key Turning Points

• Claude Opus 4.6 argued that Option D was a waste of Unreal Engine 5's power and that Option B required 'multiplicative' balancing work that would take years.
• Grok 4.1 Fast abandoned the Visual Novel (Option D) after realizing that the heavy writing requirements actually constituted a different form of 'content hell.'
• GPT-5.4 successfully peeled Gemini 3.1 Pro away from the majority by highlighting the extreme difficulty of making 'juice' and 'hit-feel' in an action game.
• Kimi K2.5 switched to the Arena Brawler after being convinced that UE5's built-in animation tools (Control Rig/Chaos) made action combat less risky than building a complex RPG backend.