Remarks on history and presence of game tree search and research

  • Authors:
  • Ingo Althöfer

  • Affiliations:
  • Fakultät für Mathematik und Informatik, Friedrich-Schiller-Universität Jena, Jena, Germany

  • Venue:
  • Information Theory, Combinatorics, and Search Theory
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

One hundred years ago, in 1912 game tree search was introduced as a scientific field by Ernst Zermelo in a concise 4-page paper. Almost four decades later the first computers were there, and three more or less concrete proposals for a Chess computer program were made by Norbert Wiener, Claude Shannon, and Alan Turing. After a long march of craftsmanship, in 1997 computer Deep Blue beat the best human Chess player in a match with six games. The other big classic in the world of games is Go from Asia. The approach from computer Chess does not work in Go. But in 2006 a Monte Carlo tree search procedure became the starting point of a triumph march. Within the following six years computer Go programs have reached a level near to that of the best western amateur players. Also in other games like Havannah, Monte Carlo search led to tremendous progress in computer playing strength. We describe the origins of game tree search in the early 20th century and discuss some of the waves of progress. With the help of C. Donninger we also meditate about the twilight role of science and scientific research for progress in game programming.