MORANA › How It Is Balanced
A BUILD-IN-PUBLIC DEV LOG

How MORANA Is Balanced

A printed card game has no patch button. Once the box ships, every number on every card is final. So MORANA is balanced before it prints, with a simulator that has played the full ruleset more than 130,000 times and caught four real balance flaws that would otherwise have reached your table.

Most modern card games balance after release. They watch the metagame, then push a digital update or a ban list to fix whatever turned out broken. A physical, fixed-distribution game does not get that luxury. There is no server to patch and no card to silently re-word. The set you receive is the set forever. That constraint sounds scary, and it is exactly why we treat pre-print balance as the single most important job on the whole project.

The promise of fixed distribution is that every player owns the same complete set and nothing is pay-to-win. That promise is only worth making if the set is actually fair. A broken card in a game you cannot patch is not an inconvenience, it is a permanent flaw in a box someone paid real money for. So we go looking for those flaws with more force than a printed game usually gets.

The simulator: a patch button that runs before the box exists

We wrote a headless simulator that plays MORANA end to end by the real paper rules: Embers by the turn, simultaneous secret commits, every one of the set's card effects, Omen and Vigil and Clash, the Turning, the 13-Slava race, and the turn-8 cap. A heuristic plays both sides. Because the randomness runs from a fixed seed, every game is fully reproducible: a result we see once, we can see again exactly, which is what makes it evidence rather than an anecdote.

Then we run it at volume. Thousands of games per matchup, across every archetype we can build, looking for four things:

130k+
full-rules games simulated across three balance passes
3
balance patches, each re-verified across the full matchup suite before shipping
0
changes shipped without re-running the full suite to check nothing else moved

What the first 46,000 games caught

The very first large run was humbling, which is the point. It surfaced four real problems, each of which would have hurt a printed game:

  1. A degenerate scoring deck that stacked score-amplifier cards and won by turn six while barely contesting the board.
  2. An undefined core rule: the price to Reborn a Spirit was never actually stated anywhere, even though several cards reduced it.
  3. An inverted matchup where one faction beat the other far harder than intended.
  4. A signature mechanic that did not fire: against decks with little removal, nothing died, so the Turning, the heart of the whole game, simply never happened.

Three of the four were fixed with small, careful card and rule changes before any print run. The fourth, the lopsided matchup, corrected itself in an instructive way: once the heuristic learned to play the printed cards more realistically, the matchup measured near-even on its own, and the nerf we had drafted to force it got thrown out. Catching a false alarm from our own tool is exactly the kind of check this process exists for. Writing "we killed a degenerate deck with tens of thousands of simulated games" is only satisfying because the alternative was shipping it.

Two later fixes, in the open

Balance is not one event, it is a habit. Two more passes followed, and they are worth showing because they demonstrate the method rather than just the result.

Patch two · the Warden fix

When a choice is not really a choice

Every player picks a Warden, so the Wardens in a faction should be genuine side-grades, not one obvious pick. The simulator showed one Hearth Warden was quietly a must-play: swapping her in for her partner swung a matchup by about fifteen points on nothing but a free, repeating point of score for holding the center with a single body. She also propped up the score-amplifier deck.

The fix was one clause. Her bonus now asks you to genuinely mass the center with two or more of your own Spirits, a real tempo cost instead of a free tap. The result: her matchup fell back to even, her faction partner became a real alternative again, and the amplifier deck was trimmed to fair, all with zero collateral on the other matchups.

Patch three · the signature-mechanic payoff

Making the marquee mechanic worth building around

The Turning is the soul of MORANA, so the deck built entirely around Reborn should be competitive. It was the weakest deck in the game. An early idea, a universal reward for reborning, was tested and rejected, because the score-racing deck also reborns and would have pocketed the same reward and grown stronger. The simulator caught that before it shipped.

The answer was to gate the reward on a card the racing deck does not run: Veles the Cattle-God now grants a point of score on the first Spirit you Reborn each turn. It lifts the Reborn engine into fair territory, and because no other deck runs that card, every other matchup came out byte-for-byte identical in testing. A precise fix, and the suite confirmed it.

The honest limits, stated plainly

We do not pretend a simulator is the whole answer. A heuristic plays good, not perfect, so the last few points of faction balance are noise to it, and those belong to human hands. Bluffing depth, the feel of a close finish, and whether the game is genuinely fun are things only real players at a real table can judge. What the simulator does is separate the two jobs: it removes the broken math with overwhelming volume, so that human playtesting can spend its scarce time on feel instead of on catching a 20-point skew nobody enjoyed losing to.

So the rule we hold ourselves to is simple. The machine kills the degenerate and the dead. The table decides the fun. Neither ships a change the other has not seen.

This is what building in public means to us: not a highlight reel, but the actual process, including the decks we broke and the fixes we rejected. If you back MORANA, this is the discipline you are backing.

Why this matters if you are considering the game

Two reasons. First, it is proof the skill-over-luck promise is real: a game is only decided by decisions if the cards underneath are actually fair, and this is how we make sure they are. Second, it is proof of seriousness. A solo maker who runs a hundred thousand simulated games before asking anyone to print a box is telling you exactly how much they care about the thing being good.

If that is the kind of game you want to exist, the best thing you can do is stand with it early. You can play the First Duel free to feel the design for yourself, then join the founders list on the home page to help kickstart the First Turning with us, early-bird perks included.

Frequently asked

How is a trading card game balanced before it is printed?
With a headless simulator that plays the complete ruleset against itself thousands of times per matchup, on a fixed seed so every result is reproducible. We measure win rates across every archetype, hunt for degenerate decks and dead cards, change the fewest cards that fix the problem, and re-run the whole suite to confirm nothing else broke. More than 130,000 games have been simulated across three balance passes.
Can you patch card games after they are printed?
No. A printed card cannot be edited after it ships, unlike a digital game. That is exactly why MORANA is a fixed-distribution game balanced before printing. The simulator is our patch button, and it runs before the box exists instead of after.
What is a simulated game of MORANA?
One complete game played by a computer following the real paper rules end to end: Embers, simultaneous secret commits, every card effect, the Turning, the 13-Slava race, and the turn cap. Running thousands per matchup reveals win rates and broken lines far faster than humans can, so human playtesting can focus on feel.
Is the simulator a replacement for human playtesting?
No. It reliably kills degenerate strategies and large matchup skews, which are far beyond heuristic noise. The last few points of faction balance, bluffing depth, and whether the game is fun belong to human playtests. The simulator tells us what to playtest first; it does not replace the table.
Does balancing with a simulator make the game solved or boring?
The opposite. The goal is a wide field of viable decks with no single dominant list and no dead archetype. Our audits confirm several viable archetypes, near-even mirrors, and a signature Reborn engine that is now competitive rather than a trap. A balanced game is a deeper game.

This dev log describes the design method for the planned First Turning set (62 cards) at campaign stage. Win rates and simulation counts describe MORANA's own internal testing and reference no other specific product. Numbers are directional and rules may change before the campaign.