Success rate of optimal nested scalable oversight at a 400-point general Elo gap, separately for Mafia, Debate, Backdoor Code, and Wargames
2 events · 0 assessments · 1 decision
Assessed
First pass under a tight budget. Recorded the Engels et al. 2025 figures (13.5 / 51.7 / 10.0 / 9.4 %) as a multi-valued estimate with the generalization caveats. No decomposition or arguments were added: this is a single-source modelled result. Set importance to 0.4 and contestation to 0.35. Escalated the bundling of four values in one quantity to the Curator. The instance was not reviewed and the definition was not set, both for lack of budget.
Claim entered the graph