SYSTEMS / WEB APPLICATIONS

Back to Systems
Live

NeuralGambit

A portal to chess psychology: research, training, and data tools that look at chess through the lens of cognitive science.

NeuralGambit turns my research on chess expertise into something players can use. It explains how strong players think, and gives them tools to test their own perception, memory, and problem solving.

Visit neuralgambit.com Explore features ↓
Type
Full-stack web application
Domain
Chess psychology
Data
500,000+ FIDE players
Languages
English · Croatian

THE IDEA

Research players can use.

“Chess is the Drosophila of the human mind.”

Psychologists have studied chess for decades because it makes expertise visible and measurable. My own research looks at how strong players recognise patterns, and why that same skill can sometimes lead them astray.

NeuralGambit brings these findings out of journals and into practice. Its mission is to investigate the human mind using chess as a method, and to give players training grounded in cognitive science rather than guesswork.

FEATURES

Four ways into the expert mind.

Research & analysis

Research blog

In-depth articles on the psychology of chess, such as the hidden architecture of chess thinking and why strong players make mistakes, written for players rather than academics.

  • Explains studies on perception, memory, and intuition
  • Newsletter for new insights
Read the blog
Training

Chess exercises

Training modes designed by a psychologist to show how your mind solves problems, not just whether you found the move.

  • Chess Puzzler: puzzles filtered by tags, with optional time limits
  • Memory Reconstruction: study a position for 30–90 seconds, then rebuild it from memory
Start training
Data & statistics

Rating predictor

Enter your rating, age, title, and years active to see a projected improvement path, modelled on data from more than 500,000 FIDE-rated players.

  • 15-year rating trajectory with an 80% prediction interval
  • Plateau probability and rating ceiling analysis
  • Likelihood of being a fast improver
Predict a rating
Pattern recognition

Perception

Paste a position and a logic engine finds the checkmate patterns and positional motifs in it, such as the bishop pair, discovered attacks, or a bad bishop.

  • Imports any position in FEN notation
  • Highlights each motif on the board with an explanation
Train your eye

UNDER THE HOOD

How it is built.

Back end

Python server with API endpoints for the rating model and pattern detection.

  • Python
  • REST endpoints

Front end

Responsive interface with interactive chess boards and move validation.

  • JavaScript
  • chess.js
  • Bootstrap
  • Tailwind CSS

Data & modelling

Rating predictions from a dataset of 500,000+ FIDE players, with interactive charts.

  • Statistical modelling
  • Chart.js

Deployment

Hosted in the cloud with a custom domain, available in English and Croatian.

  • Heroku
  • Internationalisation

MY ROLE

From research to product.

01

Concept and content. Defined the idea and wrote the research articles, drawing on my background in chess psychology.

02

Training design. Designed the exercises around how experts perceive, remember, and solve problems.

03

Data and modelling. Built the rating predictor on a large dataset of FIDE players.

04

Development and deployment. Built the front end and back end and deployed the application.

Confirm these points match what you did, and name your back-end framework and data sources if you want to show them.

Visit neuralgambit.com