Obsidian design director Josh Sawyer says RPG towns can cost about four times as much as dungeons per comparable map area. Here is how that tradeoff shapes quest density, pacing, exploration, and player expectations.
Image: wnhub.io
Sawyer puts a production number on a familiar RPG problem
Obsidian design director Josh Sawyer has put a blunt estimate on one of the most expensive promises in role-playing games: the believable town. In comments from a Q&A vlog reported by PC Gamer and summarized by BestGamesWorld, Sawyer said that, compared with a dungeon, “generally speaking, a town is four times the expense per square foot, basically.” He clarified that he was not using square footage as a literal budgeting unit, but as a way to compare maps of similar size. BestGamesWorld reports that Sawyer framed the estimate as a broad generalization drawn from experience at Obsidian and Black Isle.
That is the concrete development here, and it cuts into a long-running expectation players bring to RPGs. A large city or town reads as content-rich before the player has done anything. Doors imply interiors. Named NPCs imply backstories. Factions imply questlines. Shops imply economies. A plaza suggests overheard rumors, guards, social class, and consequences for player choices. A dungeon of the same footprint can ask the player to read layout, threat, loot, and encounter rhythm. A town has to pass a different test: it has to feel like people live there.
For readers searching around Josh Sawyer RPG cities, the value of the comment is that it turns a taste debate into a production tradeoff. Players often ask why an RPG does not simply add another major hub, make every neighborhood enterable, or give every settlement the density of a classic city map. Sawyer’s estimate suggests the answer is rarely a lack of affection for RPG town design. It is often that every square of urban space makes demands on writing, art, scripting, animation, navigation, quest logic, and QA at the same time.
A town consumes budget because it has to pretend to be normal
Sawyer’s explanation, as reported by BestGamesWorld, centers on population and specificity. A town needs many non-player characters, and many of those NPCs need dialogue. Even when every resident does not receive a unique model or a deep questline, the act of making a place feel populated creates custom work that a dungeon can often avoid.
That distinction matters because RPG city development cost is not only a matter of visual assets. A dungeon room can be excellent because it supports combat geometry, a locked chest, a trap, a lore note, or a boss setup. Reuse can be a strength if the player understands the rules. Corridors, chambers, enemy patrols, and environmental storytelling can operate within a tight mechanical language.
A town is judged by a less forgiving standard. If a market has six stalls and only one vendor speaks, the silence becomes visible. If a temple has priests but no ritual, the space can feel staged. If a noble district has mansions with locked doors and no social routes into them, players read absence as cut content. If a peasant says one line forever, players may accept it in one game and mock it in another, depending on how strongly the rest of the fiction has trained them to expect reactivity.
That is where RPG quest design becomes expensive. Dialogue is rarely isolated text. A small town quest may require a named NPC, alternate responses for different stats or reputations, item states, journal updates, fail conditions, combat fallback, reward tuning, voice or barks depending on the game, and checks for whether the player killed someone, stole something, or arrived after a main-story milestone. The more systemic the RPG, the more a harmless-looking conversation can become a branch point the team must support for the rest of the game.
Quest density is the real pressure point
Sawyer’s four-times comparison helps explain why RPG hubs are often dense in uneven ways. A city that looks large but offers few meaningful interactions disappoints players. A city that gives every corner a quest can become a scheduling trap for the studio and a pacing trap for the campaign. The problem is not whether towns are valuable. It is how much of the game’s total progression economy they absorb.
For completionist players, this is where the friction becomes obvious. A hub promises agency, but it also interrupts forward momentum. Before leaving town, many RPG players feel compelled to check every shop, exhaust every dialogue tree, recruit every companion, identify every faction, collect every task, and mark every locked door for later. When the writing and questing are strong, that ritual is satisfying. When the hub arrives at an awkward point in the campaign, it can feel like the main story has been placed in storage while the player clears errands.
BestGamesWorld reports that Sawyer pointed to Twin Elms in Pillars of Eternity as an example of the challenge. He called the area a “big pacing landmine” that many players did not appreciate, and apologized for it. That is a notable admission because Twin Elms is not being criticized here as an idea. The issue, in Sawyer’s framing, is placement and production weight. An expensive urban or settlement area can arrive late enough that the team has spent considerable effort on maps the player may meet when they are already pushing toward resolution.
The design tension is familiar across party-based RPGs. Early hubs teach systems and establish social stakes. Midgame hubs widen faction choice. Late hubs must justify slowing the player down after the build is online, companion arcs are advanced, and the central conflict is near its peak. If they cannot offer decisions that feel equal to that moment, their density works against them.
Exploration changes when the map is full of people
Sawyer’s comparison also reveals why dungeon exploration and town exploration solve different problems. Dungeons are usually built around legibility under pressure. The player asks where enemies are, which path loops back, which resource to spend, and whether to press onward. A good dungeon can create drama through depletion: health, spells, camping supplies, ammunition, consumables, and patience.
Towns usually create drama through interpretation. Who is lying? Which faction’s version of history should be trusted? Which merchant is hiding a supply chain? Which locked house is a quest hook rather than set dressing? Which local law will punish a build that solved every wilderness problem with violence? These are expensive questions because they require authored context.
Sawyer’s examples, as reported by BestGamesWorld, help mark the range. He named Athkatla in Baldur’s Gate 2, Disco Elysium, and the recent Zero Parades as examples of RPGs with believable, dense urban environments. He also noted that Disco Elysium and Zero Parades have an advantage because they do not need to include as many believable combat encounters as other RPGs.
That observation is especially useful for understanding player expectations. An urban RPG built around investigation, dialogue, and social navigation can spend its density budget differently from a combat-heavy party RPG. If the game does not need a steady supply of plausible fights in alleys, mansions, sewers, and taverns, the city can concentrate on conversation, clues, and character logic. A traditional combat RPG has to make the city feel civilized while also finding credible places to stage danger. That is a harder fiction to sustain.
The hidden cost is reactivity, not decoration
The easiest mistake is to hear Sawyer’s estimate as an art-production complaint. The prettier the city, the higher the cost. That is part of it, but the RPG-specific burden is reactivity. Towns concentrate the player’s desire to be recognized by the world.
A dungeon can often reset the player’s social identity at the door. Inside, the character is an adventurer solving a hostile space. In town, the character is a reputation, a class fantasy, a moral history, and a bundle of prior choices walking past civilians. A paladin, thief, cipher, barbarian, priest, or silver-tongued diplomat may all expect different routes through the same civic problem. Players who invest in non-combat skills expect those investments to surface in urban play.
That is where Obsidian RPG development history gives Sawyer’s comment extra weight. Wikipedia identifies Joshua Eric Sawyer as an American video game designer known for role-playing games and says he has been employed by Obsidian Entertainment since 2005, where he holds the position of studio design director. The same public biography lists his work as project director on Pillars of Eternity and Pillars of Eternity II: Deadfire, and as game director and narrative director of Pentiment. In other words, these remarks come from a designer whose games are closely associated with character builds, dialogue checks, faction texture, and historically grounded worlds.
From that perspective, a town is expensive because it is where the character sheet meets society. If a game supports stealth, speech, reputation, class identity, companion interjections, crime, and faction allegiance, a town is where all those systems should plausibly matter. When they do not, players notice. When they do, the studio pays for it in implementation and testing.
What is confirmed, and what players should not assume
There is no new Obsidian project announcement attached to these comments in the provided source material. The confirmed information is narrower and more useful: PC Gamer reported on Sawyer’s Q&A remarks, BestGamesWorld quotes the four-times comparison, and the report says Sawyer treated it as a broad production generalization rather than a universal rule. No budget, staffing total, schedule, or platform-specific detail was given for any current game.
That distinction matters because comments like this can be flattened into “developers do not want to make cities” or “cities are too expensive now.” Sawyer’s reported examples point in the opposite direction. He named admired urban RPG spaces and said towns can be made. The warning is about tradeoffs. A dense town may mean fewer other maps, a smaller wilderness, fewer dungeons, a slower act structure, or a late-game area that asks for attention the player is no longer prepared to give.
For players, the practical takeaway is to judge RPG scope by density and timing, not map count alone. A single hub with strong quest layering, believable NPC routines, and meaningful build checks may represent a larger development bet than several combat zones. Conversely, a giant city with locked doors, repeated barks, and thin quest consequences may be a sign that the game is spending on scale without enough support for interaction.
For developers and modders, Sawyer’s estimate is a reminder to design towns around the choices they must carry. If a settlement exists to introduce a faction, it needs different density than a capital city that will anchor half the campaign. If a late-game town interrupts the finale, it needs stakes that belong to the finale. RPG town design succeeds when the player’s instinct to explore is rewarded at the same intensity the story is asking them to continue.
The city fantasy survives, but it has to be budgeted honestly
Sawyer’s comments land because RPG players love cities for good reasons. Towns are where lore becomes daily life. They let a build speak before it fights. They turn treasure into equipment, rumors into quests, and companions into people with opinions about the world. They are also where production optimism goes to be tested.
A dungeon can be efficient and still excellent. A town that feels efficient can look hollow. That asymmetry is the heart of Sawyer’s reported four-times estimate. Urban RPG spaces are asked to carry atmosphere, systems, quest density, social credibility, and player fantasy at once. When they work, they become the places players remember for decades. When they miss, they become the stretch of the campaign players quietly dread on replays.
The best use of Sawyer’s remark is not as a defense of smaller RPGs or a dismissal of ambitious cities. It is a clearer lens for reading them. Every believable inn, temple, slum, court, dock, and market is competing with combat spaces, companion content, enemy variety, cinematic support, and the final act. If a future RPG promises a vast city, the sharper question is not how large it is. The question is what the game is prepared to make that city do.