NYT Games Solver

Automated puzzle solvers for New York Times word games, turning daily puzzles into algorithmic challenges.

Supported Games

LetterBoxed

A puzzle where you connect letters on a square to form words, with each word starting where the last one ended.

The solver:

• Finds optimal solutions minimizing total words used
• Explores valid word chains connecting all letters
• Uses efficient graph traversal algorithms

Spelling Bee

Create words using 7 letters, with one center letter required in every word.

The solver:

• Identifies all valid words from the dictionary
• Highlights pangrams (words using all 7 letters)
• Ranks words by point value

Technical Implementation

Dictionary Data Structures

Implemented efficient word lookup using:

• Trie structures for prefix matching
• Hash sets for O(1) word validation
• Pre-computed valid word lists

Search Algorithms

• Depth-first search for LetterBoxed chains
• Constraint satisfaction for valid word generation
• Pruning strategies for performance optimization

Why Build This?

Beyond solving daily puzzles, this project explores:

• Combinatorial optimization
• Word graph algorithms
• Efficient search strategies