CARPET-BOMB DEFENSE

We Monte-Carlo'd this.

140 attack configurations. 2,000 random targets per cell. Deterministic (seed=42). Run it yourself.

The attack we worried about

A whale claims K slots strategically to extract value from honest players. We tested four strategies:

  • Random slot picks
  • Uniform spread (evenly across the bracket)
  • Cluster LOW (cheap slots 0-K)
  • Cluster HIGH (expensive slots N-K to N)

Each tested against 5 to 500 honest players. Whale-as-late-arrival (the actual extraction case · whale captures zero seed dividend).

Result · 115 of 140 configs lose for the whale

StrategyOutcomeEV range
RandomALWAYS loses-3% to -15%
High (expensive)ALWAYS loses (catastrophic)-39% to -96%
Uniform spreadMostly loses; some marginal gain at small K-10% to +11%
Low (cheap)Profits when H ≥ K+5% to +68%

The only winning whale strategy is LOW (cheap-slot carpet-bomb). And it only works under specific conditions.

What the cheap-low attack requires

  • ~125 ETH committed capital · 500 cheap slots cost $375k. Real money, not gas-cheap.
  • 10+ sybil wallets · per-address cap is 50. Real funding gas across every wallet.
  • Irrational honest participation · whale needs 500+ honest piling into expensive slots despite obvious whale presence in the heatmap.
  • Variance tolerance · single-round outcome has wide spread; whale can lose 125Ξ outright on bad target draws.
  • Late arrival (no seed bonus) · if honest seeds first, whale captures 0% seed dividend.

What we DON'T claim

❌ "Whales can never profit." Wrong. Cheap-low against many irrational honest can profit.

❌ "Symmetric play always loses." Wrong. Uniform spread vs comparable H can win ~10% in some configs.

What we DO claim (empirically validated)

✅ Random-strategy carpet-bombing always loses 10-15%.

✅ Expensive-slot carpet-bombing always loses 60-96%.

✅ Cheap-slot carpet-bombing requires honest H ≥ K AND irrational opponents AND multi-wallet sybil setup AND late-arrival (no seed bonus).

✅ Per-address cap (50 slots) bounds single-wallet damage.

✅ Whichever side wins, 3% of every pot flows to $BURN. The protocol always wins.