The ground conditions prevailing on the day of a cricket match is an important confounding variable that results in the majority of cricket analyses requiring qualification. We present a Bayesian method for estimating the value of ground conditions in the absence of a direct measure. We use dynamic programming techniques to estimate models of both the first and second innings and we outline an application for each model. We extract a proxy variable for risk from our first-innings model and we use this variable to successfully estimate the trade-off between scoring rate and the probability of survival for individual batsmen. This enables us to decompose a batsman’s performance into ability and strategic nous. Our second-innings model gives an estimate of a team’s probability of winning at any point in the second innings of the match. We use this variable in conjunction with our ground-conditions variable to outline a new method for adjusting the target score in rain-affected matches. We introduce a simple metric for comparing the performance of various rain rules and we find that our proposed rule outperforms the incumbent Duckworth/Lewis method.