Computer Analysis Tools

Serious competitive players use computer engines to analyze positions after games, run simulations across thousands of iterations, and quantify the equity cost of any play. The two tools most used in the North American competitive community are Macondo and Magpie.

Macondo

Free · Open Source · Runs in Browser

Macondo is a full-featured Scrabble analysis engine written in Go by César Del Solar. It runs directly in your browser — no installation required. It supports NWL and WOW lexicons and can analyze any board position, suggest the highest-equity plays, and run Monte Carlo simulations to evaluate the long-term consequences of a move.

What you can do with Macondo

  • Position analysis. Enter a board position and rack; Macondo ranks every legal play by equity.
  • Simulation. Run thousands of simulated continuations from a position to find the play that wins most often — especially useful when equity alone doesn't distinguish between plays.
  • Endgame solver. Macondo has a dedicated endgame solver that finds the optimal move sequence when the bag is empty.
  • Game import. Load a .gcg file (the standard Scrabble game log format) and step through each move, analyzing any position in the game.
  • Pre-endgame analysis. Evaluate positions with a few tiles left in the bag where blocking or setting up plays can be decisive.

Getting started

  1. Go to domino14.github.io/macondo — the engine loads in your browser.
  2. Select your lexicon (NWL23 or WOW24) from the settings.
  3. Either enter a position manually using the board editor, or use Load GCG to import a game file from Woogles or another source.
  4. Click Analyze to see all legal plays ranked by equity. The top play is highlighted.
  5. To run a simulation, select a candidate play and click Simulate. Macondo will run thousands of random continuations and report win rates for each play under consideration.

GCG files can be downloaded from your game history on Woogles.io after each game.

Magpie

Free · Open Source · Command-Line

Magpie (MAGPIE) is a high-performance Scrabble engine written in C by Josh Castellano, one of North America's top-rated competitive players. It is designed for speed: simulations that take minutes in other tools can run in seconds in Magpie. It is primarily a command-line tool, best suited for players comfortable in a terminal environment.

What you can do with Magpie

  • Fast simulation. Magpie's simulation engine is among the fastest available — useful for running large numbers of iterations to statistically separate close plays.
  • Position analysis. Generate and rank all legal plays for any position by equity.
  • Endgame solving. Find optimal endgame sequences.
  • Inference and inference-based simulation. Magpie can simulate with partial information about the opponent's rack, based on their observed plays.
  • Scripting and batch analysis. Because it's command-line driven, Magpie is easy to script for bulk analysis of many positions.

Getting started

  1. Visit the Magpie GitHub repository and follow the build instructions in the README for your operating system (macOS, Linux, or Windows).
  2. Download the required lexicon data files as described in the repository documentation.
  3. Run ./magpie from the build directory to start the interactive prompt.
  4. Use the help command to see available commands for loading positions, analyzing plays, and running simulations.
  5. Positions can be entered in standard notation or loaded from GCG files.

The GitHub README is the authoritative source for current installation steps and command syntax, as Magpie is under active development.

Which Tool Should You Use?

Easier to Start
Macondo

Best for most players. Runs in the browser with no setup, has a graphical board interface, and handles the full analysis workflow — from position entry through simulation. A good first tool for anyone new to computer-aided Scrabble analysis.

Faster & More Scriptable
Magpie

Best for players who want maximum simulation speed, batch analysis, or inference-based simulation. Requires a build step and comfort with the command line, but repays the setup cost with significantly faster iteration times.

Many competitive players use both: Macondo for its convenient browser interface during casual post-game review, and Magpie when they need to run large simulations or script position analysis.