> The appropriate approach depends on the expected percentage of duplicates. And problems that seem simple… aren’t. tldr once you get past 30% duplicates the f

> The appropriate approach depends on the expected percentage of duplicates. And problems that seem simple… aren’t. tldr once you get past 30% duplicates the fastest algo switches