Kill Rewards
Kill rewards follow a log-weighted schedule where early kills pay more than late kills.
Core Concept
Kill rewards are calculated using a logarithmic weight formula where the reward decreases as fewer players remain alive. This means early kills (when more players are alive) earn more than late kills.
Weight Function
The weight increases logarithmically with alive count:
| Alive Before | w(A) = 1 + log₂(A) | Interpretation |
|---|---|---|
| 2 | 2.00 | Final kill |
| 3 | 2.58 | |
| 4 | 3.00 | |
| 5 | 3.32 | |
| 6 | 3.58 | |
| 8 | 4.00 | |
| 10 | 4.32 | |
| 16 | 5.00 | |
| 32 | 6.00 | |
| 64 | 7.00 |
Why Log-Weighting?
- Rewards aggression: Early kills when it's hardest pay most
- Diminishing returns: Prevents snowballing from late kills
- Strategic depth: Risk/reward for early engagement
- Sum preservation: All bounty gets distributed
Exact Sum Property
The schedule guarantees that the exact sum equals the bounty pool. Any rounding remainder is distributed to the earliest kills.
Schedule Precomputation
The kill schedule is computed once at round start and never rescaled mid-round. This ensures determinism regardless of elimination order.
Edge Cases
Self-Elimination
If a player eliminates themselves, no kill credit is given and the bounty for that slot remains in the pool, going to the winner at finalization.
Multiple Kills Same Frame
When multiple kills happen in the same frame, they are processed in order determined by the game engine.
Bounty Depletion
If all bounty is distributed, later kills pay nothing, though EE can still trigger if the target was already met.
Scaling with Player Count
| Players | First Kill | Last Kill | Ratio |
|---|---|---|---|
| 5 | ~1.25x avg | ~0.75x avg | 1.66x |
| 10 | ~1.40x avg | ~0.65x avg | 2.16x |
| 20 | ~1.55x avg | ~0.55x avg | 2.82x |
| 50 | ~1.70x avg | ~0.45x avg | 3.78x |
Larger rounds have steeper reward curves.
Next Steps
- Bounty Inheritance - Kill chaining
- Reward Distribution - Pool structure
- Take Profit - Securing kill earnings