Abstract

Communicating results between experimentalists and theorists in particle physics has been long and varied. From efficiency maps to selection cut flows – the collaboration between the two communities has continued to grow and evolve. Now, particularly more than ever, a stronger effort has been led within the ATLAS Collaboration taking advantage of the existing technologies used, such as containerization and plain-text data formats, to make analyses fully reproducible. One such data product from an experimental analysis relies on the statistical model used to derive the results published in papers, and these statistical models are essential information for analysis preservation and reuse. The ATLAS Collaboration is starting to publicly provide likelihoods associated with statistical fits used in searches for new physics. These likelihoods adhere to a specification first defined by the 𝙷𝚒𝚜𝚝𝙵𝚊𝚌𝚝𝚘𝚛𝚢 p.d.f. template. This talk will focus on how these statistical models came about, the technical developments to make this possible, and illustrate how detailed information on the statistical modeling can enhance the short- and long-term impact of experimental results.