i know best case , worst case avl tree logn find, insert, , delete. best case , worst case hashtable chaining? how differentiate 2 if given 2 mystery data structures? what best case , worst case hashtable chaining? find, insert, delete operation best time o(1) or constant time. worst time, insert: o(1), assuming chaining done adding end of chain delete: o(k), k = number of elements on longest chain. k == n when hashing function poor , n values map same entry , gets chained. so, absolute worst case o(n). find: o(n), using same reasoning delete how differentiate 2 if given 2 mystery data structures? just idea: insert data in both structures , plot insertion time insertion operation vs n. choose data make mimic worst case avl tree (i.e. if integers, insert in order 1000000, 99999, 999998...) not hashtable. notice plot being generate. avl tree 1 have insertion time log(n) beginning while should constant hashtable. around 1000th insertion alv tree take 1...