The optimization problem
Information moves fast now. A boss kill on Tuesday night is a wiki page by Wednesday morning and a meta build by the weekend. Twenty years ago an MMO could ship with a few obviously broken setups and live happily for months while the playerbase poked at it. Today that runway is hours. As a developer, balance issues, bugs, and dominant strategies all surface at internet speed.
For a game as intricate as Doomborne, this is the question I've been chewing on: how do you build a system rich enough to be interesting, that doesn't collapse into a single solved answer within a week of release?
The two roads modern MMOs have taken
Most modern MMOs land in one of two camps.
The first is curated and straightforward. The mechanics are clear, the optimal play is mostly obvious, and content is designed as bridges to the next piece of content. FFXIV and modern WoW live here. This isn't the same as "theme park" — there's still real moment-to-moment decisionmaking, the progression path is just linear. The advantage is enormous: you can balance precisely, ship predictably, and the player gets a tightly curated experience. The cost is that the game ends up feeling like a single tuned ride rather than a sandbox.
The second is open and creative. Lots of meaningful choices, lots of interaction between systems, the developer trusts the player to evaluate tradeoffs. The risk here is well-known. When a system rewards optimization but can't withstand it, you get the "optimizing the fun out of the game" trap — players find the best line and then feel obligated to run it, even when they'd rather not. The genre is littered with games that wanted this and didn't pull it off.
Doomborne is going for the second. I want to be honest about why that's hard before I explain why I think it's possible.
What League of Legends gets right
The model in my head is League of Legends.
On paper League should be solved. Each champion has five abilities. The item shop has a finite list. Most matches drop into a small set of recurring shapes. Compared to a modern MMO it looks tiny. And yet — 15 years in, the meta still shifts, players still innovate, "off-meta" picks still win games. League is not solved, despite every reason to expect it should be.
My read on why: it isn't the size of the system that resists solving, it's how contextual the optimal answer is. The right ability sequence depends on what the other nine players picked. The right item depends on the last five minutes of the game, not the matchup chart. The best answer is rarely a build — it's a build for this specific game state, right now. The information players have to integrate is so high-dimensional that no single solution generalizes.
That's the property I want.
Doomborne's bet
You can preserve that property in an MMO by making it hard to reach the "solved gamestate" in the first place. Three things I'm leaning on:
Subclasses you actually swap. Every class has a base kit and picks one of three subclasses keyed to a stat. The three subclasses pull from completely different skill pools — not three flavors of the same role, but three actual answers to "what does this class do." A Warrior tanking a boss with a heavy phase mechanic and a Warrior clearing a wave room shouldn't be running the same loadout. The swap isn't a min-max checkbox at character creation; it's a decision about what you're walking into.
Wide stat spread on gear. Items don't sit on a clean DPS-tank-healer triangle with a single dominant roll. The pool is broad enough that "best in slot" depends heavily on what you're stacking and what you're fighting. Two builds that look superficially identical on the character screen can carry very different stat distributions and play very differently.
Gear that gives you skills. Some gear slots carry active abilities that go on your hotbar. That makes equipping a piece of gear a kit decision, not just a stat decision. The same character can show up to two encounters with the same skills, the same subclass, and a different effective toolkit, because the gear changed which buttons they have.
The bet is that these three systems, multiplied against varied boss and encounter design, produce enough situational variance that "optimal" is always for this fight, not forever. Same way League's optimal is always for this game.
What this means in practice
If this works, the player experience is: you have a build you're comfortable with, you walk into a new encounter, and at least one of your decisions — subclass, stat priority, hotbar skills from gear — feels worth reconsidering. Not every fight. But enough that the game keeps asking you to think instead of letting you settle.
If it doesn't work, the community finds the dominant configuration in week one and the rest of this is academic. I'm aware of that. The reason I think it's worth trying is that the alternative — designing for the curated path — is a different game than the one I want to make.
More on this once the systems hit playtest and I get to see what people actually do with them.
— Kevin