In this paper, we provide polynomial and pseudopolynomial algorithms for classes of particular instances of interval data minmax regret graph problems. These classes are defined using a parameter that measures the distance from well-known solvable instances. Tractable cases occur when the parameter is bounded by a constant.