The theory of hashing never matched up well with how I've experienced it in the real world. For example almost all analysis of the hash table (for proving stuff

The theory of hashing never matched up well with how I've experienced it in the real world. For example almost all analysis of the hash table (for proving stuff about load factor, chain length, or probe clustering sizes, etc) starts off with a "uniform hashing" assumption. But that assumption have basically never held true given how languages define their default hashes.