Recently, I was doing Cow Run (http://www.usaco.org/index.php?page=viewproblem2&cpid=265), and I at first used the Policy Based Data Structure faster map (https://codeforces.com/blog/entry/60737), but I TLE'd on the last 4 test cases. I then switched back to using the unordered_map, and I passed all the test cases. Can anyone tell me the reason why the unordered_map was faster? I was under the impression that the PBDS map was faster. Thanks!
To run this code, simply submit them in a file to the USACO website above, and use C++11.