The article introduces a method to estimate posterior distributions from data alone, without adding extra information. By using operational priors based on information theory, researchers can create reference posteriors that closely resemble those obtained with proper priors. This approach unifies and expands on previous work, addressing previous criticisms.