DARWIN PRO LEAGUE — SEASON 1
DPL S1 is the competitive Darwin Project ladder, running on this site. Scrim sets are recorded as they're played, every set is priced into the ladder the moment it publishes, and the full game-by-game record is public — the LADDER, the set boards, and every player profile update live.
S2 — the community scrim season that came before — is closed and archived. Its final standings are permanent: the top 1 / 10 / 25 finishers carry season trophies on their profiles, and the whole S2 ladder and set history remain browsable under the S2 tabs.
DPL S1 is rated with Glicko-1. Two numbers per player: a rating, and an uncertainty (RD) that measures how sure the system is about it.
- The number you see is conservative. Displayed rating = rating − 1.5 × uncertainty. Play more, uncertainty drops, your displayed rating climbs toward your true skill. The displayed number is also exactly what the ladder sorts by — no hidden math.
- Provisional players are dimmed. Until your uncertainty settles (roughly your first handful of sets), your rating is marked provisional and can't leapfrog established players on the board.
- Inactivity costs certainty, not rating. Step away and your uncertainty slowly widens — your displayed rating drifts down until you play again, and recent form counts for more than ancient results.
The S2 formula had two known flaws, confirmed by backtesting the full season: it rewarded volume (grinding sets beat playing well) and it was blind to opponent strength (beating a stacked lobby paid the same as beating a quiet one). Glicko-1 fixes both — every result is weighed against who was actually in the set, and the uncertainty model keeps small samples from outranking a season's work. Across the whole S2 record, it predicted set outcomes measurably better than the shipped formula.