The model dynamics repeatedly eliminates the least adapted species and mutates it and its neighbors to recreate the interaction between species. A comprehensive study of the details of this model can be found in Phys. Rev. E53, 414–443 (1996). A solvable version of the model has been proposed in Phys. Rev. Lett.76, 348–351 (1996), which shows that the dynamics evolves sub-diffusively, driven by a long-range memory.
We consider N species, which are associated with a fitness factor f(i). They are indexed by integers i around a ring. The algorithm consists in choosing the least fit species, and then replacing it and its two closest neighbors (previous and next integer) by new species, with a new random fitness. After a long run there will be a minimum required fitness, below which species don't survive. These "long-run" events are referred to as avalanches, and the model proceeds through these avalanches until it reaches a state of relative stability where all species' fitness are above a certain threshold.[2]
^Wei1, Li; Yang, Luo; YuanFang, Wang & AiPing, Cai. "A mean-field Bak-Sneppen model with varying interaction strength". Chinese Science Bulletin, 2011, p. 3639.
Bak, P. (1996). How Nature Works: The Science of Self-Organized Criticality. New York: Copernicus. ISBN0-387-94791-4.