We propose a comparative study of four algorithms dedicated to the lexicographic unranking of combinations. The three first ones are algorithms from the literature. We give some key ideas how to analyze them in average and describe their strengths and weaknesses. We then introduce a new algorithm based on a new strategy inside the classical factorial numeral system (or factoradics). We conclude the paper with an experimental study, in particular when n = 10000; k = 5000, our algorithm is 36 (resp. 53) times as fast as our C++ implementation of Matlab's (resp. Sagemath's) algorithm.