In this work we propose a novel Ising model for benchmarking optimization algorithms called the Corrupted Biased Ferromagnets (CBFM). A detailed evaluation of the CBFM instance class across six classical algorithms suggest that it is a challenging problem for both heuristics and global optimization algorithms.