The article introduces a new method to improve the accuracy of estimating distribution and hazard functions for nonnegative random variables. Traditional methods struggle near x=0, but the proposed technique, inspired by previous research, reduces bias by using a generalized reflection approach. Through simulation studies, the new estimators were found to outperform traditional methods and the classical reflection method.