Our Sqrt(n)-approximation for the independent set problem would strongly suggest that P = NP. Experimental results showed a 2-approximation ratio on real-world benchmarks, outperforming the theoretical Sqrt(n) worst-case guarantee.