Recognise the Pattern First

Before touching code, ask: is this a subarray problem (sliding window)? Do I need pairs (two pointers)? Is the data sorted (binary search)? Matching to one of ~15 patterns is 60% of the work.

State Complexity Upfront

Verify brute force works first. Derive its O(n²) complexity, then think: can a hash map (O(1) lookup) or sorting (O(n log n)) do better? Never jump to optimise without a baseline.

Master the 15 Core Patterns

Sliding window, two pointers, fast/slow pointers, merge intervals, cyclic sort, in-place reversal, BFS, DFS, two heaps, subsets, binary search, bitwise XOR, top-K, K-way merge, 0/1 knapsack.

Draw Before You Code

Sketch the tree, the graph, the pointers. Visual thinking catches bugs before they happen and shows interviewers your mental model clearly. A whiteboard drawing is worth 100 lines of rushed code.

Handle Edge Cases Early

Empty array, single element, all duplicates, negatives, integer overflow — state these before coding. FAANG interviewers dock points for solutions that break on trivial edge cases.

5 Problems Per Pattern

Don't grind 500 random problems. Solve 5 representative problems per pattern — 75 total. After that, 80% of new problems will feel familiar. Pattern depth beats volume every time.