From root: How can I trust the actions I take in simulations?

Can I backtest the actions a simulation recommends?

Cover these topics: - Lack of data on this generally because its hard to control for cause-to-effect - Quasi-experiments all kinds and the openaction2outcome project

References and further reading

Angrist, Joshua D., and Jörn-Steffen Pischke. 2009. Mostly Harmless Econometrics: An Empiricist’s Companion. Princeton University Press.
Hardwick, Robert J. 2026a. “How can I tell luck from skill?” umbralcalculations, November. https://umbralcalc.github.io/posts/how_can_i_tell_luck_from_skill.html.
Hardwick, Robert J. 2026b. “How can I trust the actions I take in simulations?” umbralcalculations, August. https://umbralcalc.github.io/posts/how_can_i_trust_the_actions_i_take_in_simulations.html.
Hardwick, Robert J. 2026c. “Openaction2outcome: Real-World Reference Points for Testing Models That Predict the Effects of Decisions.” GitHub. https://github.com/umbralcalc/openaction2outcome.
Imbens, Guido W., and Donald B. Rubin. 2015. Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction. Cambridge University Press.