Hi - when running a multiple regression, I am looking at all stepwise regression steps (backwards, forward, mixed and adjusted r2).

Hi - when running a multiple regression, I am looking at all stepwise regression steps (backwards, forward, mixed and adjusted r2). I am getting the same ideal variables for backwards and forwards but when I run mixed and adjusted r2, I am getting the same variables from back/forward PLUS another variable that is an ideal fit. Is this normal? Wouldn't all four steps give the same exact ideal variables?