#Vault Breakers: Super Score

DEV ONLY · NOT FOR PRODUCTION · NOT FOR CERTIFICATION. Draft math: mutable, unfrozen, unpublished, not enabled for any operator. Internal tests and simulations are not external certification.

draft-v0.1 — APPROVED 2026-08-19 as the UNBALANCED REFERENCE BASELINE. Mechanics and runtime approved, whole-game estimate recorded, Bonus Buy 100× / 500× deliberately left unbalanced, not approved for DEV publication. Superseded by draft-v0.2; its evidence is retained unchanged.

draft-v0.2 — DRAFT MATH APPROVED 2026-08-19. Not frozen, not final, not certified. Free spins run their own charge-dense reel table, so the reference Bonus Buy prices of 100× and 500× hold at 96.42% and 97.05%, and the whole game is back on 96%.

The high-volatility, charge-dominant free-spins shape is approved for Draft and must pass a separate DEV playtest before Freeze. Do not smooth it or retune it without playtest results.

Published to DEV 2026-08-20 as vault-breakers-super-score-v1. All four identities matched.

✅ FIXED 2026-08-22 in the backend, not yet re-verified on DEV. See the third addendum in evidence/manifest-draft-v0.2.md and backend/doc/vault-breakers-super-scatter-runtime-fix.md. The config hash did not move; every figure below is now a correct statement about the runtime as well as about the math. The paragraph below is retained as the record of what was wrong.

⚠️ The Super Scatter instant prize does not pay on the player path. A base spin evaluates a config with free_spins stripped, and the executor resolves through the free-spins trigger, so it never fires. Over 400,000 drawn rounds: full config 55 prizes, player config 0. What dev actually serves is Super Scatter at 0% instead of 1.8246%, whole-game ≈94.07% instead of 95.8973%, and a 50,000× max win that is unreachable.

Every figure below is a correct statement about the approved math and none of it is rewritten — the gap is in the runtime. Freeze blocker. See evidence/manifest-draft-v0.2.md, second addendum, and backend TODO GE6.

The game is left reachable on dev deliberately: playtest on pacing and free-spins shape is still worth collecting; playtest on how big wins feel is not.

An original 6×5 count-anywhere tumble slot about a modern bank heist.

#Documents

#Originality

The game takes publicly confirmed RULES from Gates of Olympus Super Scatter as a reference baseline. It takes none of its branding, characters, art, symbol designs, presentation or undisclosed mathematics. Every weight, payout and probability here is authored. No figure in any Vault Breakers document is a claim about that game's paytable, weights or RTP.

#The game at a glance

grid            6 reels x 5 rows, 30 cells
draw            every cell independent, one weight table, initial grid and refills alike
evaluation      count-anywhere -- copies visible ANYWHERE, no adjacency, no paylines
bands           8-9 / 10-11 / 12-30.  Seven copies pay nothing
tumble          remove all winning cells, drop, refill, repeat until a snapshot pays nothing
charges         2x-500x, one-shot, ADDED across a spin's sequence, multiply that sequence's win
scatter class   SCATTER + SUPER_SCATTER count together
free spins      4+ scatter-class on ONE snapshot -> 15 spins;  3+ inside -> +5, repeatable
super scatter   1/2/3/4+ -> 100x / 500x / 5,000x / 50,000x, only on a snapshot that triggered
buys            Standard (base pool) and Super (pool floors at 10x)
ceiling         50,000x per game cycle -- reaching it ends generation, not just the number

#Draft results — draft-v0.2

Exact and estimated are separate. Nothing sampled is labelled exact.

Target 96.00% sits inside the confidence interval, 0.76 standard errors above the estimate. Both Bonus Buys hold their reference prices — not because the prices were moved, but because the feature is now worth them.

#Against the approved baseline

#Symbols

H1 diamond case · H2 gold bars · H3 cash stacks · H4 bearer bonds · L1 thermal cutter · L2 electronic keycard · L3 crew radio · L4 lockpick set · SCATTER robber mask · SUPER_SCATTER golden master key · MULTIPLIER hacking charge.

Eight regular symbols rather than the six the design named. That is a measured requirement, not a preference — see the GDD and Math Spec. At six, a 96% game can only afford a top band of 1.68× and a lowest band of 0.013×.

#Reproduction

cd /Users/admin/kiro/backend

# Contract, exact results, pinned identity
go test -race -count=1 ./libs/mathengine/games/vault_breakers_super_score

# Reference-to-runtime parity, both halves
make test-parity

# Draft evidence runs
go test -tags vbdraft -count=1 -v -timeout 180m \
    ./libs/mathengine/games/vault_breakers_super_score -run Draft

#Not in Draft v0.1

• Any art, audio or UI asset. This is math only.
• A math-studio authoring preset — the studio domain has no multiplier_charge or super_scatter feature type yet, so the pack cannot be seeded through versions/from-pack.
• A game-engine DEV admission gate. Deliberate: publishing is out of scope until approval.
• A product decision on the free-spin shape: a free-spin snapshot pays on 8.23% of snapshots against the base game's 19.76%, and 95%+ of a sequence's award is the charge part.