A synthesis of deck-construction sources built around one idea: consistency is a math problem. Below are the copy counts, ink ratios, mulligan rules, and a worked archetype that keep a 60-card deck doing what it's supposed to, turn after turn.
15–19The one number to rememberNon-inkable cards to keep, so you can ink every single turn.
1
The Consistency Math
Variance control
Non-inkable thresholds
Every deck lives or dies on hitting an ink drop each turn. Model it with two targets: 90% certainty is the competitive standard for your opening sequence; 98% is what you want when a miss is an instant loss (you stall out with no ink to play). In practice that means holding non-inkables to 15–19 cards — leaving 41–45 inkable sources.
Consistency goal
Curve top
Max non-inkables (no mulligan)
Max non-inkables (reasonable mulligan)
98% certainty
7 ink by Turn 7
16
18
98% certainty
6 ink by Turn 6
17
21
90% certainty
7 ink by Turn 7
20+
24+
The more your plan depends on hitting every ink drop, the fewer non-inkables you can afford. A reasonable mulligan buys you 2–4 extra non-inkable slots.
2
Copy Counts
How many of each card
4 · 3 · 2 · 1 logic
4×
Must see by T1
Essential engine pieces or early staples. You accept drawing dead multiples to guarantee the early play.
e.g. Simba – Protective Cub
3×
Inkable mid-threats
High draw odds, but being inkable defuses the risk of clogging your early hand with multiples.
e.g. Aladdin – Heroic Outlaw
2×
Non-inkable effects
You want the effect once per game. Multiples are high-variance and force off-curve turns.
e.g. Rafiki – Mysterious Sage
1×
Toolbox / search
Single silver-bullet targets in decks with enough recursion or filtering to find them.
search & recursion shells
The cost of four: running 4 copies maximizes your Turn 1 hit, but pushes the chance of drawing multiple dead copies to 23% after 14 looks. That's the tax you pay for consistency on non-inkable staples.
3
1-Drops & Search Density
Hitting your targets
Guaranteeing a Turn 1 play
To open with a 1-drop at 90% certainty, your deck needs a critical mass of them:
Mulligan style
Copies needed
Hyper-aggressive
8
Moderate
10
Search targets (top-4 look)
How dense the pool must be depends on the fail state of the searcher:
Ariel — 70%Min 15 Songs. A miss is survivable — she still puts a body on the board.
Be Our Guest — 95%Min 31 Characters. A miss is a dead draw: no board, no resources.
Combo reliability (hyper-aggressive mulligan)
Line
Plan
Success rate
Lilo → Simba
Turn 1 Lilo into Turn 2 Simba aggro curve
42.5%
Tinker Bell → Grab Your Sword
"Giant Fairy" shift into a Turn 4 board wipe
32.1%
Even hyper-optimized two-card lines land under half the time. Build a deck that functions without the combo, and treat the combo as upside.
4
Ratios & the Ratio Paradox
Structure
Functional deck ratio
Starting from a 60-card minimum and a strict two-ink limit:
3 · SupportFill the last 8–12 slots with utility items and removal.
The Ratio Paradox: the 22–24 character guideline collides with the math — Be Our Guest needs 31 characters to hit 95%. If you run it, you must exceed the beginner character count; if you can't, cut the card.
What each ink brings
Ink
Primary strengths
Amber
Card draw and board control
Amethyst
Disruption and removal effects
Emerald
Character enhancement and ramping power
Ruby
Aggressive strategies and direct damage
Sapphire
Resource manipulation and utility
Steel
Defensive options and protective abilities
5
The Mulligan Matrix
Opening decisions
Curve priorityTurns 1–3 come first: secure a play for each. If more than half your hand costs 3+, mulligan aggressively.
Repeat removalGame 1 (blind), mulligan repeated cards. Game 2 vs. Aggro, override it — keep a removal-heavy opener.
UninkablesKeep a non-inkable only if it's an essential Turn 1–3 play; otherwise ship it to avoid resource stalls.
Tech cardsHold matchup tech (Benja vs. Items, Evasives vs. low-interaction boards) only once you've confirmed the archetype.
6
Archetype: Sapphire/Steel Ramp-Control
A worked example
The 2-4-6 rule
The engine is a precise resource-to-threat transition — ramp early, land an oversized threat on curve, then lock the board.
Turn 2
Ramp
Accelerate ink
Tipo · Sail the Azurite Sea
Turn 4
Threat
High-impact play
Goliath (6/5)
Turn 6
Lock
Board control
Hades · Vincenzo
Early — accelerate & select
Develop Your BrainTurn 1 filtering to find ramp pieces.
Captain HookChallenger +2 deterrent vs. early Aggro.
Tipo / Azurite SeaPrimary ramp engines for the 2-4-6.
Kida (3/3, Ward)Untargetable body + draw vision.
CinderellaTurn 3 Princess "ramp + draw" cycle.
Mid — stabilize & pressure
Goliath (6/5)Forces opponents to discard to two cards.
NamaariDraw/discard engine that scales off the discard pile.
Tinker BellWide-board cleanup; on-entry and challenge damage.
PlutoConverts Steel challenges to 2 lore; kills items.
He Hurled His Thunderbolt4-damage removal song.
Late — close the engine
HadesHard removal — puts threats into the opponent's inkwell.
Vincenzo (8 will)3 lore; deals 3 damage whenever they play a character.
Arthur (Chal/Resist +2)Mirror tool — forces challenges on ready characters.
Headless HorsemanDamage scaling off discarded actions.
Beyond the HorizonDraw engine that disrupts their hand.
Matchup tactics
vs AggroDelay ramp; prioritize Hook and Tinker Bell interaction.
vs Amethyst/SteelGoliath + Headless Horseman to deny resources; Inkrunner onto evasives (Elsa, Genie).
MirrorPrioritize Arthur; keep threat density high to counter Hades.
vs Ruby/AmethystLean on Namaari and Vincenzo; Inkrunner past evasive blockers.
★
Pre-Flight Checklist
Variance: 15–19 non-inkables for 98% Turn-7 ink reliability.
Search density: 31+ Characters for Be Our Guest; 15+ Songs for Ariel.
Curve: a Turn 1, 2, and 3 play secured through the mulligan.
Pivot: ID the opposing archetype by Turn 2 — interact vs. Aggro, ramp vs. Control.
Density: the 2-4-6 rule so threats outpace removal.
Copy audit: 4× Turn-1 essentials; cap high-cost non-inkables at 2.