A quantitative study on the uncertainty of ablation heat transfer is conducted, in which a dual platform model is used for ablation prediction, and the heat transfer process under the ablation dynamic boundary is numerically solved using the finite element method. Firstly, an orthogonal experimental study was conducted to analyze the influence of parameters on the target variable. It was discovered that while the input parameters had consistent impacts on the backside temperature, they had distinct effects on the ablation quantity under two common heating situations (high heat flow, short time and low heat flow, long time). In order to obtain more accurate uncertainty quantification results, Monte Carlo (MC) and polynomial chaos (PC) methods were further used to conduct uncertainty analysis on the ablation heat transfer problem. Through sensitivity analysis, it was found that under two heating conditions, heat flux was the key factor affecting the ablation amount. The back temperature is most affected by the material’s thermal conductivity in state 1 (high heat flow, short time), but heat flow in state 3 (low heat flow, long time) has a comparatively bigger effect. Compared to the MC method, the PC method can effectively reduce computational costs and obtain satisfactory computational results.