When to Use BSTs

In competitive programming, you rarely implement balanced BSTs from scratch. Instead: Use language libraries:

• C++: set, map, multiset (Red-Black trees)
• Java: TreeSet, TreeMap
• Python: sortedcontainers (not built-in, but available)

When you need:

• Ordered iteration: BST gives sorted order naturally
• Range queries: find all elements in $[L, R]$
• Predecessor/successor: find closest smaller/larger value
• Rank queries: augment with subtree sizes

When hash tables are better:

• Only need insert/delete/lookup (no ordering)
• Keys aren't naturally comparable

When arrays + sorting is better:

• Static data (no insertions after initial load)
• Binary search on sorted array is simpler