Arcane 101 Decks: Cost Curves, Card Counts and What to Cut
Deck building is where Arcane 101 runs are won. Here is how to set a cost curve for a 1-5 player party and which cards to cut before the first boss.
Every Arcane 101 run is decided by a deck, and the developer's own framing puts deck building at the centre of the game: choose your school, build your deck, every defeat teaches you something. There is no published deck size, no official mana curve, and no card database. What there is, is a set of constraints you can reason about.
The two constraints nobody publishes
Your school caps your pool. Eight schools share 150+ spells, so your character holds roughly twenty cards' worth of options. That is enough for one coherent plan and not enough for two.
A turn is finite. Whatever the exact energy rule turns out to be, a card that costs more than your typical turn is conditional by definition. Every deck is therefore a statement about how many conditional cards you tolerate.
Those two constraints produce the same conclusion from different directions: the ceiling on a deck is its consistency, not its best card.
Setting a cost curve
A workable early curve, in plain terms rather than fake percentages:
If your opening hand can produce a dead turn, the curve is wrong. That is the whole test. Play five opening hands in your head: if two of them do nothing, cut an expensive card and add two cheap ones.
What to cut
Cutting is the skill the game actually rewards. Three passes, in order:
- Cut duplicates of role. Two cards that both answer the same pattern are one card plus a liability.
- Cut the card you never play. Not the card you think is weak — the one that sits in hand for two runs. That is data, not opinion.
- Cut the card that needs a partner you do not reliably draw. Setup cards are only as good as their payoff's draw rate.
Most players add cards when a run fails. The correct response to a failed run is usually one cut and one addition, not a rebuild.
Building for a party, not for a screenshot
Arcane 101 runs with one to five players, and the party changes what a good deck is:
- Solo. Your deck must contain all four roles: damage, defence, setup, utility. That is why solo decks are tighter and less flashy.
- Duo. Split roles explicitly. One player carries defence so the other can run the expensive plan.
- Full party. Each deck can be lopsided. Five players with five narrow decks outperform five balanced ones, provided the roles are actually covered.
The failure mode in co-op is not a weak deck — it is five decks that all answer the same problem. Before the first boss, ask the group one question: who covers what.
The 11-boss test
Eleven bosses across eight lands is the practical test of a deck. A deck that clears land one comfortably and stalls in land three is not unlucky; it is missing a role. Treat each boss as a diagnostic:
What is not confirmed
Deck size, energy rules, card rarities and any "best deck" list are all unpublished. The developer confirms 150+ spells across eight schools and nothing more granular. Any page that quotes exact deck sizes or tiered card rankings is extrapolating — including pages that frame their numbers as community testing without a source.
The method above survives that gap because it does not depend on numbers. Cost before damage, consistency before power, roles before favourites. Those three rules hold whatever the patch does.
Next steps
Read the spell guide for the five-second card test, the schools page if you have not committed to a pool yet, or the boss list to see the encounter order your deck has to survive. If you are still deciding whether the game is worth the time, the beginner guide covers what the first hour actually looks like.
