I'd like to contribute an algorithm to this field of study. I call it, "Spend several months researching a novel algorithm when brute force would have been effective for all real-world problem sizes."