An order statistics model based on the list of top m scores after th change

Consider an infinite sequence of independent identically distributed (i.i.d.) continuous random variables. We begin with an initial set of m random variables and keep in our list only the top m values as new variables are observed sequentially and added to the data set. We obtain the joint distribution of the ordered observations in our list after the ℓ th change occurs to the list. We also obtain the probability of a change in the list given that there are already ℓ changes. We discuss the relationship between the proposed model, ordinary order statistics, and m-records.