Draft pick value charts are widely used by NFL teams to evaluate trade proposals during the draft. Traditional charts provide point estimates for the relative value of each selection but do not quantify the uncertainty associated with those values. This paper introduces a statistical framework for estimating the uncertainty in draft pick values using historical trade data.Starting from the Rich Hill draft chart as a prior mean valuation, deviations between observed trade outcomes and chart values are analyzed to estimate the variability associated with each pick. Because individual picks have limited observations, a correlation structure between neighboring selections is introduced through a custom kernel function. Kernel regression is then applied to produce smooth estimates of the standard deviation associated with each pick.The resulting model generates probabilistic confidence intervals for draft pick values and allows trades to be evaluated in terms of statistical percentiles. Using historical trade data, the Rich Hill chart is shown to provide a better baseline representation of observed trade behavior than the traditional Jimmy Johnson chart. Model calibration using normalized residuals, log-likelihood, and confidence interval coverage indicates that the proposed framework produces well-calibrated uncertainty estimates while preserving the empirical structure of draft trades.