Loadout as budget

Lemma: Power lives inside a fixed budget — points, cells, slots, or hand size. Adding more of one thing necessarily means having less of another. Vertical growth is gated by something the player must spend, not just earn.

What it solves

The naive instinct in progression design is "more = better": more health, more damage, more skills. That instinct produces power-creep — players accumulate everything until the game becomes trivial.

A budget mechanic refuses the more-is-better drift. Power isn't accumulated; it's packed. Every choice is also a sacrifice.

This solves several problems at once:

• Active opportunity cost at every loadout decision.
• Bounded power growth without nerfing content.
• Identity for individual builds — different budgets produce visibly different playstyles.
• Re-engagement — players love rearranging budgets they've outgrown.

Variants across games

Game	Budget shape	What you "pack"	The packing constraint	Trade-off cost
Sparklite	Spatial — 3×3 → 5×5 grid	Patches with Tetris footprints (1–4 cells)	Cells available on the patch board	Fitting a 4-cell damage patch costs 4 cells you can't use for HP / map / energy
Warframe	Numeric + categorical — 60 mod points + polarity slots	Mods costing N points each	Total points + matching polarity = packing efficiency	A non-matching slot costs +25%; matching slot costs −50%
Slay the Spire	Combinatorial — 3 energy / turn + ~5-card hand from a thin deck	Cards costing energy	Energy per turn + draw consistency	Bigger deck dilutes draws; energy can't carry to next turn
Moonlighter 2	Adjacency-driven — fixed backpack grid where placement triggers synergies	Relics with conditional placement tags (POINTED, ABSORB, corner, row)	Adjacency rules + grid position	Where you put a relic determines whether it triggers a synergy or wastes its tag
Path of Exile 1	Massive numeric — 1500 nodes shared across classes; ~122 points at level 100	Single-point passives, multi-point notables, keystones, jewel sockets	Total points + pathing — points spent traversing also count	Respec costs Orbs of Regret; full re-spec is endgame-currency-expensive
Mega Man Battle Network	Layered fixed budgets — 30-chip Folder + 4×4 NaviCust grid + ~5-card Hand drawn from Folder	Battle Chips (codes A–Z + ★), NaviCust programs (Tetris shapes), opening Hand from random draw	Folder size (30) + Hand size (5) + NaviCust cells; chip codes constrain co-play	Fewer co-playable codes = more dead Hand draws; more bug-tolerant NaviCust = more debuffs
Xenoblade Chronicles 3	Combinatorial: 8-art palette × class × cross-class master — pick a class for each character; pick which 3 arts of that class are active; pick alt-arts from a different mastered class	Arts (class-bound), Master Arts (from another mastered class), accessory slots, gem slots	8 art slots × ~20 classes × N other classes mastered for cross-slotting	Pick the wrong cross-class slot and you've burned 60+ hours of class mastery; the loadout is the build
Hollow Knight	Pure numeric — 11-notch charm budget	~45 charms with integer costs (1–5 notches each)	Total notches; opt-in overcharm state lets you exceed at the cost of 2× damage taken	A 5-notch charm eats nearly half the early budget; cheap-charms-only vs. one-big-charm builds play very differently
Nier: Automata	Numeric storage budget — starts ~64, max 256 — that includes the HUD	Plug-in chips: combat passives, Auto-* accessibility, AND the HUD elements (HP bar, minimap, damage numbers). Same chip exists at multiple storage costs depending on quality	Total storage; mandatory OS Chip cannot be removed without dying; three saved Sets (A/B/C) for fast role-swap	Showing the minimap costs storage you can't spend on Auto-Heal. Frame and action share one budget — the only entry where the UI itself is in the loadout
Pokémon RSE	Tiny 4-slot moveset per Pokémon, drawn from hundreds of typed options; plus 510 EVs across 6 stats per Pokémon	Battle moves (typed; with PP); Effort Values (cap 252 per stat); held items (1 slot per creature)	4 move slots — adding a 5th forces forgetting an existing move permanently; EVs hard-capped at 510 total / 252 per stat	A 5th move replaces a 4th — coverage vs power vs status vs finisher; EVs in two stats means none in the others. The smallest-grain canonical case in this knowledge base
Balatro	5-slot Joker loadout from a pool of 150+; slot order matters (Jokers fire left-to-right)	Jokers (passive triggers — flat adders, scalers, conditional triggers, chain meta-jokers like Blueprint)	5 slots; expansion via Negative editions (rare) and Antimatter voucher; Spectral cards can destroy all Jokers as a "swap" cost	A Joker dropped is 5 turns of build identity gone. Slot order doubles the optimization space — Blueprint copies the Joker to its right, so the chain composition matters

Visual contrast

Sparklite	Warframe	Slay the Spire	Moonlighter 2	Path of Exile
Patches collected, awaiting placement on the Med Bay grid	Mod points + polarity slots; matching polarity halves cost	4/3 energy + 5-card hand; every turn is allocation	Backpack grid where adjacency triggers POINTED / ABSORB synergies	Atlas tree (≈700 nodes) — the meta-budget; the character passive tree is even bigger (~1500 nodes)

Hollow Knight	Nier: Automata	Pokémon RSE
11 notches, ~45 charms with integer costs. Pure numeric.	72 storage, chip costs 1–17. Same numeric form — but the HUD elements (HP bar, minimap, damage numbers) are also chips in the same budget.	4 move slots, hundreds of typed moves. The smallest-grain budget — and the constraint that's anchored Pokémon battling for ~30 years.

When to use this pattern

• Roguelites / run-based games where each run wants distinct identities.
• Looter-shooters / RPGs with long-term build progression that risk power-creep otherwise.
• Deckbuilders, deck-light card games, ability-bar designs.
• Anywhere "more upgrades" would otherwise just keep stacking.

The pattern requires:

• A clear, visible packing space (the grid, the mod points, the hand size).
• A way to expand the budget at real cost (Forma re-leveling, med-bay tier-ups, deck thinning).
• Distinct items with non-uniform "weight" — if everything is the same size, the budget doesn't bite.

Pitfalls

• Budgets that are too generous defeat the point. If the player can fit everything they want, the constraint isn't a constraint. Sparklite's mid-game flatness comes from the 5×5 board being roomy enough to fit "the answer."
• Budgets that never expand feel punishing. Players need progression somewhere — the budget itself is a good axis of progression (Sparklite's med-bay tiers, Warframe's Reactor/Catalyst).
• Spatial budgets are intuitive but author-heavy. Designing patches/items with multiple footprints takes more art and tuning than a flat numeric cost.
• Numeric budgets can feel abstract. Warframe's polarity-matching is initially opaque to new players. The math has to be explained well or the depth is invisible.

Adjacent patterns

• power-creep-mitigation — broader category. A loadout budget is one mechanism for mitigating power creep; there are others (e.g. enemy scaling, content gating).
• opportunity-cost-loadout — restates the trade-off framing.
• subtractive-deckbuilding — Spire's specific implementation: a deck that's too big is worse, so the budget pressure is internal to the deck rather than external to it.
• grid-inventory — the specifically-spatial flavor.

Why this matters as a design lesson

The variants above show that the same design problem (preventing power-creep while keeping progression interesting) can be solved with completely different math — spatial, numeric, or combinatorial. The choice of math shapes the feel of the constraint:

• Spatial budgets feel puzzly and tactile.
• Numeric budgets feel algebraic and optimization-driven.
• Combinatorial (deck-based) budgets feel probabilistic and adaptive.

Pick the math that fits your game's tone. The pattern is the same; the texture is yours to design.