Dominance hierarchy refers to the phenomenon that the status of each individual in the animal population has a certain order, which not only reflects the social structure of the animal population, but also plays an important role in survival, reproduction and population regulation. The hierarchical structure of animal populations includes both linear and nonlinear cases. The linearity index h (Landau’s h) proposed by Landau (1951) can be used to quantify the degree of linearity of hierarchical structure, and the calculation method of this index was improved by de Vries (1995). Rank differences between individuals affect the strength of natural selection and sexual selection. However, in most animals, individual rank is not obvious in phenotype, which needs to be calculated using the dominance ranking algorithm based on observational data. Existing dominance ranking algorithms include algorithms based on the win probability, win frequency and opponent strength (such as Clutton-Brock et al.’s index, David’s score), the number of wins (such as I & SI, Elo-rating, Randomized Elo-rating), and methods based on Graph Theory i.e., ADAGIO. Based on the field data, we discussed the correlation between these algorithms, and found that ADAGIO, Randomized Elo-rating, I & SI, and David’s score were relatively similar (Fig. 3). The ranking results obtained by Elo-rating and Clutton-Brock et al.’s index are similar, but there are great differences between algorithms based on the win probability and other algorithms (Fig. 4). According to the characteristics of these algorithms and the relationship between algorithms, the recommended algorithm selection in different cases is summarized, so that researchers can choose the appropriate algorithm according to the practical situation. In addition, we briefly introduced the characteristics of ranking algorithms in other fields, to provide a reference for the introduction of new algorithms in animal behavior research.