Skip to content

balance(doomsday-clock): reshape wave cadence — flatten the front, steepen the endgame#2

Open
Zixer1 wants to merge 1 commit into
feat/doomsday-warship-decayfrom
balance/doomsday-clock-pacing
Open

balance(doomsday-clock): reshape wave cadence — flatten the front, steepen the endgame#2
Zixer1 wants to merge 1 commit into
feat/doomsday-warship-decayfrom
balance/doomsday-clock-pacing

Conversation

@Zixer1

@Zixer1 Zixer1 commented Jul 6, 2026

Copy link
Copy Markdown
Owner

What

Reshape the Doomsday Clock wave cadence so the required-share curve can be shaped, not just paced. WaveSchedule.rampSeconds / pauseSeconds become per-wave arrays (were single uniform values); levels are softened at the open and nudged at the top.

Why (OFM 2026 Summer data)

Analysis of all 9 tournament games under the current uniform fast cadence:

  • ~75% of eliminations happened in the first half; the last third saw only ~7% → a rushed early game followed by a drawn-out endgame stalemate.
  • The rush was worst in the biggest lobbies (up to 58% of deaths in the first third at 90 players) and disproportionately culled weaker/newer players.
  • 85% of deaths are doom-driven (only ~15% pre-doom combat), so the wave shape is the effective lever.

Change

  • Early waves lower + later — longer grace and early ramps give breathing room before the first cull.
  • Late waves steeper, no pauses — a continuous endgame squeeze that breaks the stalemate and ends games sooner.
  • Levels: 3/5/10/20/30/55%2.5/4.5/10/20/32/55% (softer open, slight top bump; final 55% squeeze unchanged). Same levels across presets — presets only change pace.

When each wave lands (mm:ss), current → proposed

Preset W1 W2 W3 W4 W5 W6 (win)
veryfast — cur 5:00 7:30 10:00 12:30 15:00 17:30
veryfast — new 6:00 8:50 10:55 12:45 13:53 14:43
fast — cur 7:20 10:40 14:00 17:20 20:40 24:00
fast — new 8:40 12:40 15:40 17:40 19:10 21:00
normal — cur 10:00 15:00 20:00 25:00 30:00 35:00
normal — new 11:40 17:00 21:10 23:55 25:55 28:25
slow — cur 12:00 18:00 24:00 30:00 36:00 42:00
slow — new 14:00 20:20 25:20 28:40 31:10 34:10

Notes

  • Contained to DoomsdayClock.ts — no other code reads the schedule shape; the HUD companion (doomsdayClockWaveState) is updated in lockstep so sim + HUD still agree.
  • Tests updated to the new schedule; tests/DoomsdayClockExecution.test.ts passes (33/33).
  • Values are data-informed playtest starting points, not final — the shape (flatten front, steepen back) is the point; exact grace/ramp/pause want tuning against live dev-lobby curves.
  • Base is feat/doomsday-warship-decay (that branch introduces DoomsdayClock.ts; it isn't on main yet), so this is a clean balance-only diff on top of it.

…eepen the endgame

Per-wave ramps/pauses (was a single uniform ramp/pause for all six waves)
so the required-share curve can be SHAPED, not just paced.

Why (OFM 2026 Summer data): under the uniform cadence ~75% of eliminations
fell in the first half while the last third saw only ~7% — a rushed early
game (worst in big lobbies, where it disproportionately culled weaker/newer
players) followed by a drawn-out endgame stalemate. 85% of deaths are
doom-driven, so the curve shape is the lever.

Change:
- Early waves are LOWER and come LATER (grace + longer early ramps) — more
  breathing room before the first cull.
- Late waves are STEEPER with no pauses — a continuous endgame squeeze that
  breaks the stalemate and ends games sooner (fast: 55% at ~21:00 vs 24:00).
- Levels softened at the open (3/5% -> 2.5/4.5%), nudged at the top
  (30% -> 32%); final 55% squeeze unchanged. Same levels across presets;
  presets only change pace.

Values are data-informed PLAYTEST STARTING POINTS — tune against live curves.
Tests updated to the new schedule.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant