Development of a chess engine using various techniques and heuristics

Simjanovski, Nikola (author), Sadikov, Aleksander (mentor)

This thesis presents the overall design, along with different algorithms, heuristics, and methods used to develop FRIChallenger, a strong competitive chess engine. During the development of the engine, the process of creating the move generator is explained, responsible for generating all legal moves in a given position. Then, with the help of algorithms (e.g., alpha-beta pruning, quiescence search, iterative deepening, and other alpha-beta selective enhancements), a selective search is created that explores only the most promising paths in the search tree. Using an extensive evaluation function, a position is evaluated to determine whether it is better for the White or Black side and by what margin. Finally, the engine’s performance was tested against other established engines using the CuteChess program, and the results indicate that the final implementation reaches a playing strength of approximately 2500 Elo.

• Language: English
• Keywords: Chess Engine, Alpha–Beta Search, Evaluation heuristics, Rust
• Work type: Bachelor thesis/paper
• Typology: 2.11 - Undergraduate Thesis
• Organization: FRI - Faculty of Computer and Information Science
• Year: 2026
• PID: 20.500.12556/RUL-181936
• COBISS.SI-ID: 278262019
• Publication date in RUL: 20.04.2026

Secondary language

• Language: Slovenian
• Title: Razvoj šahovskega programa z uporabo različnih tehnik in hevristik

Abstract

Ta diplomska naloga predstavlja celotno zasnovo ter različne algoritme, hevristike in metode, uporabljene pri razvoju FRIChallengerja, močnega in konkurenčnega šahovskega pogona. V okviru razvoja pogona je opisan postopek izdelave generatorja potez, ki je odgovoren za generiranje vseh legalnih potez v dani poziciji. Nato je s pomočjo algoritmov (npr. alfa-beta obrezovanja, quiescence iskanja, iterativnega poglabljanja in drugih selektivnih izboljšav alfa-beta iskanja) oblikovan selektivni iskalni postopek, ki preučuje le najbolj obetavne poti v iskalnem drevesu. Z uporabo obsežne evaluacijske funkcije se pozicija ovrednoti z namenom določanja, ali je ugodnejša za belo ali črno stran ter za kolikšno prednost. Na koncu je bila zmogljivost pogona preizkušena proti drugim uveljavljenim pogonom z uporabo programa CuteChess, pri čemer rezultati kažejo, da končna implementacija dosega igralno moč približno 2500 Elo.

• Keywords: Šahovski motor, Iskanje alfa–beta, Evalvacijske hevristike, Rust