This document discusses uncertainty analysis and the importance of recording assumptions. It explains that single point estimates are often inaccurate, and it is better to estimate variables as ranges or distributions. Tools like sensitivity analysis can help identify the key drivers of uncertainty. Monte Carlo analysis incorporates the uncertainty ranges into simulations to generate outputs. All estimates and assumptions must be thoroughly documented in files like the Master Data Assumption List, as models may be audited. Recording assumptions provides evidence for results and allows others to understand and validate the analysis.