Abstract
Game tree search is the core of most attempts to make computers play games. We present a fairly general theoretical analysis on how leaf evaluation errors influence the value estimation of a game position at the root. By an approach using prime factorization arguments in the ring of polynomials, we show that in this setting the maximum number of leaf-disjoint strategies proving a particular property is a key notion. This number precisely describes the quality of the heuristic game value in terms of the quality of the leaf evaluation heuristics. We extend this model to include random nodes (rolls of a die). Surprisingly, this changes the situation: utill the number of leaf-disjoint strategies ensures robustness against leaf evaluation errors, but the converse is not true. An average node may produce additional robustness similar to additional leaf-disjoint strategies. This work extends earlier ones which only deal with 0, 1 valued nodes, or without randomness.
Similar content being viewed by others
References
Althöfer I (1988) Root evaluation errors: how they arise and propagate. ICCA J 11(3):55–63
Anantharaman TS (1991) Extension heuristics. ICCA J 14(2):47–63
Ballard BW (1982) A search procedure for perfect information games of chance.In: proceedings of AAAI, pp 111–114
de Bruin A, Plaat A, Schaeffer J, Pijls W (1999) A minimax algorithm better than SSS*. Artif Intell 87:255–293
Donninger C (1993) Null move and deep search. ICCA J 16(3):137–143
Ehrhoff J, Grothklags S, Halbsgut J, Lorenz U, Sauerwald T (2003) The repair game: robust plans and disturbation management in aircraft scheduling by the help of game tree search. In: Proceedings of 43rd Annual symposium of the airline Group of the International Federation of Operational Research Societies (AGIFORS), Paris, France
Kaindl H, Scheucher A (1989) The reason for the benefits of minmax search. In: Proceedings of the 11th IJCAI, pp 322–327, Detroit
Knuth DE, Moore RW (1975) An analysis of alpha-beta pruning. Artif Intell 6(4):293–326
Lorenz U (2000). Controlled conspiracy-2 search. In: Reichel H, Tison S (eds). Proceedings of the 17th STACS, LNCS. Springer, Berlin Heidelberg New York, pp. 466–478
Lorenz U (2004). Beyond optimal play in two-person-zerosum games. In: Albers S., Radzik T. (eds), Proceedings of the 12th Annual European symposium on algorithms (ESA) LNCS 3221. Springer, Berlin Heidelberg New York, pp. 749–759
Lorenz U, Monien B (2002). The secret of selective game tree search, when using random-error evaluations. In: Alt H., Ferreira A. (eds). Proceedings of the 19th STACS LNCS. Springer, Berlin Heidelberg New York, pp. 203–214
McAllester DA (1988) Conspiracy numbers for min-max searching. Artif Intell 35(1):287–310
Pearl J (1984) Heuristics – intelligent search strategies for computer problem solving. Addison-Wesley Publishing Co., Reading
Reinefeld A (1989) Spielbaum - Suchverfahren. Springer, Berlin Heidelberg New York
Reibmann AL, Ballard BW (1983) Non-minimax search strategies for use against fallible opponents. AAAI, pp 338–342
Schaeffer J (1990) Conspiracy numbers. Artif Intell 43(1):67–84
Schrüfer G (1988) Minimax-Suchen Kosten, Qualität und Algorithmen. PhD Thesis, University of Braunschweig
Shannon CE (1950) Programming a computer for playing chess. Philos Mag 41:256–275
Author information
Authors and Affiliations
Additional information
The first author was partially supported (associate member) by the graduate school ‘Effiziente Algorithmen und Mehrskalenmethoden’, Deutsche Forschungsgemeinschaft. The second author was partially supported by the Future and Emerging Technologies programme of the EU under contract numbers IST-1999-14186 (ALCOM-FT) and IST-2001-33116 (FLAGS).
Rights and permissions
About this article
Cite this article
Doerr, B., Lorenz, U. Error Propagation in Game Trees. Math Meth Oper Res 64, 79–93 (2006). https://doi.org/10.1007/s00186-006-0064-6
Received:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00186-006-0064-6