Matroid varieties are the closures in the Grassmannian of sets of points defined by specifying which Plucker coordinates vanish and which don;;t --- the set of nonvanishing Plucker coordinates forms a well-studied object called a matroid. Many problems in enumerative geometry could be solved if there were an efficient way to compute the cohomology class of a matroid variety in an efficient way just from the combinatorial data contained in the matroid itself. Unfortunately, in full generality, this problem is known to be completely intractable. In this thesis, we explore two attempts to get a handle on this problem in the special case of a better-behaved class of matroids called positroids. First we examine a method based on degenerations on the Grassmannian that;;s known to be successful for an even smaller class of matroids called interval rank matroids, showing that it can be made to work with some effort in rank 3 but that there is a counterexample demonstrating that it is hopeless beyond that. Finally, we give a way of producing a number for each matroid variety called its expected codimension that can be computed combinatorially solely from the matroid itself and show that it agrees with the actual codimension in the case of a positroid variety.