Abstract
We consider the game of proper Nim, in which two players alternately move by taking stones from n piles. In one move a player chooses a proper subset (at least one and at most \(n-1\)) of the piles and takes some positive number of stones from each pile of the subset. The player who cannot move is the loser. Jenkyns and Mayberry (Int J Game Theory 9(1):51–63, 1980) described the Sprague–Grundy function of these games. In this paper we consider the so-called selective compound of proper Nim games with certain other games, and obtain a closed formula for the Sprague–Grundy functions of the compound games, when \(n\ge 3\). Surprisingly, the case of \(n=2\) is much more complicated. For this case we obtain several partial results and propose some conjectures.
Similar content being viewed by others
References
Albert MH, Nowakowski RJ, Wolfe D (2007) Lessons in play: an introduction to combinatorial game theory, 2nd edn. A K Peters Ltd., Wellesley, MA
Beideman C, Bowen M, Muyesser NA (2018) The Sprague-Grundy function for some selective compound games. arXiv:1802.08700 [math.CO]
Berlekamp ER, Conway JH, Guy RK (2001–2004) Winning ways for your mathematical plays, vol.1–4, second edition, A.K. Peters, Natick, MA
Boros E, Gurvich V, Ho NB, Makino K, Mursic P (2018) On the Sprague–Grundy function of exact \(k\)-Nim. Discrete Appl Math 239:1–14
Boros E, Gurvich V, Ho NB, Makino K, Mursic P (2019a) Tetris hypergraphs and combinations of impartial games. arXiv:1701.02819
Boros E, Gurvich V, Ho NB, Makino K, Mursic P (2019b) Sprague-Grundy function of symmetric hypergraphs. J. Comb. Theory. Ser. A 165:176–186
Boros E, Gurvich V, Ho NB, Makino K, Mursic P (2019c) Sprague-Grundy function of matroids and related hypergraphs. Theor Comput Sci 799:40–58
Bouton CL (1901–1902) Nim, a game with a complete mathematical theory. Ann. Math, 2nd Ser., 3:35–39
Conway JH (1976) On numbers and games. Acad. Press, London, NY, San Francisco
Grundy PM (1939) Mathematics of games. Eureka 2:6–8
Grundy PM, Smith CAB (1956) Disjunctive games with the last player losing. Proc Cambridge Philos Soc 52:523–527
Gurvich V, Ho NB (2019) Slow k-Nim, RUTCOR Research Report, RRR-03-2015. Rutgers University. arXiv:1508.05777
Jenkyns TA, Mayberry JP (1980) The skeletion of an impartial game and the Nim-Function of Moore’s Nim\(_k\). Int J Game Theory 9(1):51–63
Moore EH (1910) A generalization of the game called Nim. Ann Math, Second Series, 11(3):93–94
Smith CAB (1966) Graphs and composite games. J Combin Theory 1:51–81
Sprague R (1935–1936) Über mathematische Kampfspiele. Tohoku Math J 41:438–444
Sprague R (1937) Über zwei Abarten von Nim. Tohoku Math J 43:351–354
Acknowledgements
The authors thank Rutgers University and RUTCOR for the support to meet and collaborate. The authors thank the anonymous reviewers and the editor for the detailed reports and helpful advice.
Author information
Authors and Affiliations
Corresponding author
Additional information
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Vladimir Gurvich was partially funded by the Russian Academic Excellence Project ‘5-100’. Kazuhisa Makino was partially supported by JSPS KAKENHI Grant Numbers JP24106002, JP25280004, JP26280001, and JST CREST Grant Number JPMJCR1402, Japan.
Appendix: Slow Nim
Appendix: Slow Nim
In this section, we consider a variant of Nim, so called slow Nim. A move in a Hypergraph Nim game Nim\(_\mathcal{H}\) is called slow if each pile is reduced by at most one token. Let us restrict both players by their slow moves, then the obtained game is called slow hypergraph Nim. We study SG functions and losing positions of slow Moore Nim and slow exact Nim, where they respectively correspond to hypergraphs \(\mathcal{H}= \{H \subseteq I \mid 1 \le |H| \le k\}\) and \(\mathcal{H}=\{H \subseteq I \mid |H|=k\}\) for some \(k \le n\). We provide closed formulas for the SG functions of both games when \(n = k = 2\) and \(n = k+1 = 3\), where we remak that the SG function for slow exact Nim when \(n = k = 2\) is trivial. We also characterize losing positions for slow Moore Nim if either \(n \le k+2\) or \(n = k+3 \le 6\) holds.
Here we only present the results, where all the proofs can be found in the preprint by Gurvich and Ho (2019).
Given a position \(x = (x_1, \ldots , x_n) \in {\mathbb {Z}}_+^I\), we will always assume that its coordinates are nondecreasing \(x_1 \le \cdots \le x_n\). The parity vector p(x) is defined as the vector \(p(x) = (p(x_1), \ldots , p(x_n)) \in \{0,1\}^I\) such that \(p(x_i) = 0\) if \(x_i\) is even, and \(p(x_i) = 1\) if \(x_i\) is odd. It appears that the status of a position x in the slow Moore Nim in the cases below is defined by p(x) .
Proposition 2
The SG function \({\mathcal {G}}\) for slow Moore Nim for \(n = k = 2\) and \(n = 3, k = 2\) are uniquely defined by p(x) as follows:
-
(i)
For \(n = k = 2\),
$$\begin{aligned} {\mathcal {G}}(x) = {\left\{ \begin{array}{ll} 0, \quad \text { if } p(x) = (0,0) \\ 1, \quad \text { if } p(x) = (0,1) \\ 2, \quad \text { if } p(x) = (1,1) \\ 3, \quad \text { if } p(x) = (1,0). \end{array}\right. } \end{aligned}$$ -
(ii)
For \(n = 3\) and \(k = 2\),
$$\begin{aligned} {\mathcal {G}}(x) = {\left\{ \begin{array}{ll} 0 \quad \text { if } p(x) \in \{(0,0,0), (1,1,1)\} \\ 1 \quad \text { if } p(x) \in \{(0,0,1), (1,1,0)\} \\ 2 \quad \text { if } p(x) \in \{(0,1,1), (1,0,0)\}\\ 3 \quad \text { if } p(x) \in \{(0,1,0), (1,0,1)\}. \end{array}\right. } \end{aligned}$$
We next consider losing positions of slow Moore Nim.
Proposition 3
Consider a slow Moore Nim when \(n \le k+2\) or \(n = k+3 \le 6\). Then for a position \(x \in {\mathbb {Z}}_+^I\), we have the following five cases.
-
(1)
for \(n=k\), x is losing if and only if \(p(x) = (0, 0, \ldots , 0)\).
-
(2)
for \(n=k+1\), x is losing if and only if \(p(x) \in \{(0, 0, \ldots , 0), (1, 1, \ldots , 1)\}\).
-
(3)
for \(n=k+2\), x is losing if and only if \(p(x) \in \{(0, 0, \ldots , 0), (0,1, \ldots , 1)\).
-
(4)
for \(n=5\) and \(k=2\), x is losing if and only if
$$\begin{aligned} p(x) \in \{(0,0,0,0,0), (0,0,1,1,1), (1,1,0,0,1), (1,1,1,1,0)\}; \end{aligned}$$ -
(5)
for \(n=6\) and \(k=3\), x is losing if and only if
$$\begin{aligned} p(x) \in \{(0,0,0,0,0,0), (0,0,1,1,1,1), (1,1,0,0,1,1), (1,1,1,1,0,0)\}. \end{aligned}$$
We note that Moore Nim games satisfy that \(1\le k \le n\), and the case in which \(n=4\) and \(k=1\) is a standard 4-pile Nim.
We finally consider slow exact Nim. Note that this game is trivial when \(k=1\) or \(k=n\). We show that the game is tractable if \(n=3\) and \(k=2\). Again the parity vector plays an important role, although it does not define the SG function uniquely.
Define six sets of positions \(x \in {\mathbb {Z}}_+^3\):
and let \(C = C_1 \cup C_2\), \(D = D_1 \cup D_2\).
Proposition 4
For a slow exact Nim with \(n=3\) and \(k=2\), the SG function \({\mathcal {G}}\) can be represented by
Rights and permissions
About this article
Cite this article
Boros, E., Gurvich, V., Ho, N.B. et al. On the Sprague–Grundy function of extensions of proper Nim. Int J Game Theory 50, 635–654 (2021). https://doi.org/10.1007/s00182-020-00707-3
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s00182-020-00707-3