God numbers for Graphs, Games and Groups
Pith reviewed 2026-05-21 08:52 UTC · model grok-4.3
The pith
Finite solitaire puzzles and zero-sum games can be axiomatized as directed graphs from which Zermelo's theorem follows directly.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
By representing states of a puzzle or game as vertices in a directed graph and moves as edges, the god number for solitaire puzzles is the shortest path distance from the starting vertex to the set of solved states, or the diameter if unspecified. For two-player games, it is the minimax value that minimizes the longest possible game length over all strategies, where a strategy is a subgraph containing the initial vertex. This framework makes Zermelo's theorem that one player has a winning strategy or the game draws follow directly from the graph definitions.
What carries the argument
The god number, defined as the minimal distance to the solution space A in solitaire cases or as a minimax critical value over strategy subgraphs in two-player cases.
If this is right
- God number computation is NP-complete in general and for concrete sliding puzzles like the 15-puzzle.
- Group games such as Rubik-type problems, the Tower of Hanoi, and transposition sorting all reduce to the same graph distance measure.
- A chess problem labeled 'mate in k' receives god number exactly k under the minimax definition.
- Strategies in two-player games are uniformly modeled as subgraphs containing the initial vertex.
Where Pith is reading between the lines
- The single graph model might let researchers compare solution complexities across unrelated puzzles like sliding blocks and cube twists using the same computational tools.
- Extending the definitions beyond finite cases could show where graph-theoretic distance measures stop capturing real game outcomes.
- Treating all examples through one axiomatization could surface shared hardness patterns between sorting problems and board games that are not obvious from isolated studies.
Load-bearing premise
Every finite solitaire puzzle and zero-sum sequential game can be faithfully represented as a directed graph where distances and minimax values capture the intended notions of solving or winning without needing extra structure.
What would settle it
A concrete finite solitaire puzzle or game whose actual solution lengths or winning conditions cannot be matched by shortest paths or minimax values in any directed graph model of its states and moves.
Figures
read the original abstract
We describe and axiomatize finite solitaire puzzles and zero sum sequential games graph theoretically. Zermelo's theorem telling that there is a win for one of the players or a draw follows from the definitions. The god number is a geometric quantity that quantifies the number of moves necessary to solve the puzzle. In the solitaire case, the god number is the minimal distance from the initial state $v$ to the solution space $A$. If $v$ and $A$ are not specified, the god number is the graph diameter. God number computations are related to combinatorial sorting problems and is a NP-complete problem in general even when restricted to concrete sliding problems. In the two-player case, the god number is a minimax critical value: it minimizes the maximal game event length over the set of all strategies. A strategy is a sub-graph of the game graph that contains the initial vertex. The definition is done so that a ``mate in k" chess problem has god number k. As for examples: in the solitaire case, we look at group games like Rubik type problems, transposition games related to sorting, at sliding puzzles like the 15 puzzle or rainbow ball, or the tower of Hanoi. For two-player games, we illustrate the story using examples of small chess games, a small card game or tic-tac-toe type problems.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript axiomatizes finite solitaire puzzles and zero-sum sequential games as directed graphs with a distinguished solution set A (or terminal positions). It asserts that Zermelo's theorem on the existence of a win for one player or a draw follows directly from the definitions. The god number is introduced as the minimal graph distance from initial vertex v to A in the solitaire case (or the diameter if unspecified), and as a minimax critical value over strategy subgraphs in the two-player case. Examples are provided from group-theoretic puzzles (Rubik-type), sliding puzzles (15-puzzle, rainbow ball, Tower of Hanoi), transposition/sorting problems, and small two-player games (chess variants, card games, tic-tac-toe). The paper notes that god-number computation relates to combinatorial sorting and is NP-complete in general, even for concrete sliding puzzles.
Significance. If the framework holds, it supplies a unified graph-theoretic language for puzzles and games that directly encodes standard notions of distance and minimax, with the god number serving as a geometric measure of solution length. This perspective links to known results in combinatorial game theory and complexity, and the explicit tie to 'mate in k' problems is a clear illustration. However, the manuscript largely restates existing concepts from graph theory and Zermelo's theorem without new theorems, original computations, or machine-checked proofs, so its significance is primarily expository rather than advancing the research frontier.
major comments (3)
- [Axiomatization of games and puzzles] The assertion that Zermelo's theorem follows from the definitions is correct by backward induction on finite directed graphs with terminals, but the paper should specify in the axiomatization section whether graphs are required to be acyclic or how cycles (if present) are handled to avoid infinite play; this is load-bearing for the claim that the result is immediate from the graph model alone.
- [Two-player case and god number definition] The definition of a strategy as a subgraph containing the initial vertex (and the god number as the minimax length over such subgraphs) requires explicit conditions ensuring the subgraph is functional (out-degree exactly 1 at non-terminal positions) to guarantee it corresponds to a valid strategy; without this, the reduction to 'mate in k' is not fully rigorous.
- [Complexity and sorting problems] The claim that god-number computation is NP-complete in general, even restricted to sliding problems, is a strong computational statement; the manuscript should either cite the specific reduction (e.g., from a known sorting or puzzle problem) or provide a brief sketch, as this underpins the complexity discussion.
minor comments (3)
- [Abstract] The abstract and introduction would benefit from a clearer statement of what is novel in the axiomatization versus standard graph-distance and minimax constructions.
- [Examples] Explicit small-scale computations of the god number (e.g., for the 3-disk Tower of Hanoi or a 2x2 sliding puzzle) would strengthen the illustrative examples.
- [Definitions] Notation for the solution set A and initial vertex v should be introduced once and used consistently; avoid shifting between 'solution space' and 'terminal positions' without cross-reference.
Simulated Author's Rebuttal
We thank the referee for the thoughtful and detailed report. The comments highlight areas where the axiomatization and definitions can be made more precise, and we will incorporate revisions to address them while preserving the expository nature of the manuscript. We respond to each major comment below.
read point-by-point responses
-
Referee: The assertion that Zermelo's theorem follows from the definitions is correct by backward induction on finite directed graphs with terminals, but the paper should specify in the axiomatization section whether graphs are required to be acyclic or how cycles (if present) are handled to avoid infinite play; this is load-bearing for the claim that the result is immediate from the graph model alone.
Authors: We agree that explicit handling of potential cycles strengthens the claim. The manuscript already emphasizes finite graphs for both solitaire puzzles and games. We will revise the axiomatization section to state that the directed graphs are finite and that any play continuing indefinitely (via cycles) is classified as a draw. This ensures termination or draw in all cases, allowing Zermelo's theorem to follow directly from the finite graph model and the definitions of terminal positions without requiring acyclicity. revision: yes
-
Referee: The definition of a strategy as a subgraph containing the initial vertex (and the god number as the minimax length over such subgraphs) requires explicit conditions ensuring the subgraph is functional (out-degree exactly 1 at non-terminal positions) to guarantee it corresponds to a valid strategy; without this, the reduction to 'mate in k' is not fully rigorous.
Authors: This is a valid point for rigor in the two-player setting. We will update the definition of a strategy to require that the subgraph has out-degree exactly one at every non-terminal position for the player whose turn it is, while preserving the out-degree of the opponent as in the original game graph. This functional condition ensures the subgraph encodes a deterministic strategy, making the minimax characterization of the god number and its equivalence to 'mate in k' fully rigorous. revision: yes
-
Referee: The claim that god-number computation is NP-complete in general, even restricted to sliding problems, is a strong computational statement; the manuscript should either cite the specific reduction (e.g., from a known sorting or puzzle problem) or provide a brief sketch, as this underpins the complexity discussion.
Authors: We will strengthen this section by adding a reference to established results on the complexity of optimal solutions in sliding-block puzzles (such as generalizations of the 15-puzzle, where shortest-path computation in the configuration graph is known to be NP-hard via reductions from problems like vertex cover on implicit graphs). A brief sketch of the reduction idea will also be included to connect it to combinatorial sorting problems as mentioned in the text. revision: yes
Circularity Check
No significant circularity identified
full rationale
The paper models finite solitaire puzzles and zero-sum sequential games as directed graphs with a distinguished solution set A, from which Zermelo's theorem follows immediately by backward induction on the finite structure. The god number is explicitly defined as the graph distance from initial vertex v to A, the graph diameter when unspecified, or a minimax value over strategy subgraphs; these are direct applications of standard graph distance and game-theoretic minimax without any fitted parameters, self-referential predictions, or load-bearing self-citations. The construction is stipulated as the representation itself, rendering the framework self-contained rather than deriving results from unverified prior claims or reducing equations to their inputs by definition.
Axiom & Free-Parameter Ledger
axioms (2)
- domain assumption Every finite solitaire puzzle and zero-sum sequential game can be represented as a finite directed graph with a designated solution set A.
- standard math Zermelo's theorem holds for finite two-player zero-sum games of perfect information.
invented entities (1)
-
God number
no independent evidence
Lean theorems connected to this paper
-
IndisputableMonolith/Foundation/RealityFromDistinction.leanreality_from_one_distinction unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
The god number of a winnable solitaire is the minimal path length from v to A. ... In the two-player case, the god number is a minimax critical value: it minimizes the maximal game event length over the set of all strategies.
-
IndisputableMonolith/Cost/FunctionalEquation.leanwashburn_uniqueness_aczel unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
Zermelo’s theorem telling that there is a win for one of the players or a draw follows from the definitions.
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
Reference graph
Works this paper leans on
-
[1]
S.B. Akers and B. Krishnamurthy. A group-theoretic model for symmetric interconnection networks.IEEE Trans.Comput., 38:555–566, 1989
work page 1989
-
[2]
M.H. Albert, R.J. Nowakowski, and D. Wolfe.Lessons in Play, An introduction to combinatorial game theory. A.K. Peters, 2007
work page 2007
-
[3]
L.V. Allis. A knowledge-based approach of connect-four. 1988. M.Sc Thesis, VU Amsterdam
work page 1988
-
[4]
L.V. Allis. Searching for solutions in games and artificial intelligence. 1994. Ph.D. Limburg
work page 1994
-
[5]
A.F. Archer. A modern treatment of the 15 puzzle.American Mathematical Monthly, 106:793–799, 1999
work page 1999
-
[6]
The cyclic towers of hanoi.Information Processing Letters, 13:118–119, 1981
Atkinson. The cyclic towers of hanoi.Information Processing Letters, 13:118–119, 1981
work page 1981
-
[7]
J. Beck.Combinatorial Games, Tic-Tac-Toe Theory, volume 114 ofEncyclopedia of Mathematics and its Applications. Cambridge University Press
-
[8]
B. Benesh and R. Cambbell. Element building games onZ n.Arnold Mathematical Journal, 7:599–608, 2021
work page 2021
-
[9]
D. Berend and A. Sapir. The diameter of hanoi graphs.Information Processing Letters, 98:79–85, 2006
work page 2006
-
[10]
E. Berlekamp, J. Conway, and R. Guy.Winning Ways, for your mathematical plays. A.K. Peters, second edition, 2001
work page 2001
- [11]
-
[12]
T. Bousch. La quatri` eme tour de hano¨ ı.Bulletin of the Belgian Mathematical Society, 21:895–912, 2014
work page 2014
-
[13]
A. Br¨ ungger, A. Marzetta, K. Fukuda, and J. Nievergelt. The parallel search bench zram and its applications. Annals of Operations Research, 90:45–63, 1999
work page 1999
-
[14]
Cameron.Permutation Groups, volume 45 ofStudent Texts
P.J. Cameron.Permutation Groups, volume 45 ofStudent Texts. London Mathematical Society, 1999
work page 1999
-
[15]
C. Cashman, S.J. Miller, J. Shuffelton, and D. Son. Black hole zeckendorf games.Hardy-Ramanujan Journal, 48:102–107, 2025
work page 2025
-
[16]
A. Cayley. On the theory of groups, as depending on the symbolic equationθ n = 1.Philosophical Magazine, 7, 1854
-
[17]
X. Chen and J. Shen. On the frame-stewart conjecture about the towers of hanoi.SIAM Journal on Computing, 33(3):584–589, 2004
work page 2004
- [18]
-
[19]
The lean theorem prover (system description)
Leonardo de Moura, Soonho Kong, Floris van Doorn, and Jakob von Raumer. The lean theorem prover (system description). volume 9195, pages 378–388, 08 2015
work page 2015
-
[20]
M.A. Dimand and R.W. Dimand (eds). The foundations of game theory, vol 1-3. An Elgar Reference Collection, 1997
work page 1997
-
[21]
Dudeney.The Canterbury Puzzles
H.E. Dudeney.The Canterbury Puzzles. Dutton and Company, 1908
work page 1908
-
[22]
van den Herik E.C.D van der Werf and J.W.H.M
H.J. van den Herik E.C.D van der Werf and J.W.H.M. Uiterwijk. Solving go on small boards.ICGA Journal, 26:92–107, 2003
work page 2003
-
[23]
N. Fijalkow.Games on Graphs. Cambridge University Press, 2026
work page 2026
-
[24]
J.S. Frame. Solution to advanced problem 3918.American Mathematical Monthly, 48:216–217, 1941
work page 1941
-
[25]
D. Gale. The game of Hex and the Brouwer fixed point theorem.Amer. Math. Monthly, 86:818–827, 1979
work page 1979
-
[26]
A. Ganesan. Automorphism groups of cayley graphs generated by connected transposition sets.Discrete Math- ematics, 313:2482–2485, 2013
work page 2013
-
[27]
The GAP Group.GAP – Groups, Algorithms, and Programming, Version 4.15.1, 2025
work page 2025
-
[28]
M. Gardner. Mathematical carnival: Ticktacktoe games.Sci. Amer., 1975
work page 1975
-
[29]
S.W. Golomb. Rubik’s cube and quarks: Twists on the eight corner cells of rubik’s cube provide a model for many aspects of quark behavior.American Scientist, 70:257–259, 1982
work page 1982
-
[30]
A.W. Hales and R.I. Jewett. Regularity and positional games.Trans. AMS, 106:222–229, 1963
work page 1963
-
[31]
R. B. Hayward.Hex, a playful introduction, volume 54 ofAnneli Lax New Mathematical Library. AMS, 2022
work page 2022
-
[32]
A.M. Hinz. Pascal’s triangle and the tower of hanoi.American Mathematical Monthly, 99:538–544, 1992
work page 1992
-
[33]
Hinz, Klavˇ zar, Milutinovi´ c, and Petr.The Tower of Hanoi — Myths and Maths
A.M. Hinz, Klavˇ zar, Milutinovi´ c, and Petr.The Tower of Hanoi — Myths and Maths. Birkh¨ auser, 2013
work page 2013
-
[34]
A.M. Hinz and Parisse. On the planarity of hanoi graphs.Expositiones Mathematicae, 20:263–268, 2002
work page 2002
-
[35]
Solving 7,7,5-game and 8,8,5 game.The Journal of the Computer Games Community, 40, 2018
W-Y Hsu, C-L Ko, and I-C Wu. Solving 7,7,5-game and 8,8,5 game.The Journal of the Computer Games Community, 40, 2018
work page 2018
-
[36]
W.W. Johnson and W.E. Story. Notes on the “15” puzzle.American Journal of Mathematics, 2:397–404, 1879
-
[37]
Joyner.Adventures in Group Theory, Rubik’s Cube, Merlin’s Machine and Other Mathematical Toys
D. Joyner.Adventures in Group Theory, Rubik’s Cube, Merlin’s Machine and Other Mathematical Toys. Johns Hopkins University Press, second edition, 2008
work page 2008
-
[38]
L. Kalmar. Zur Theorie der abstrakten Spiele. 1928
work page 1928
-
[39]
R.M. Karp. Reducibility among combinatorial problems. In In R. E. Miller and J. W. Thatcher, editors, Complexity of Computer Computations, pages 85–103. Plenum, 1972. 72 MATH/CS 91r, Harvard, Spring 2026
work page 1972
- [40]
-
[41]
M. Kiyomi and T. Matsui. Integer programming based algorithms for peg solitaire problems. InLecture Notes in Computer Science, volume 2063, pages 229–240, 2001
work page 2063
-
[42]
D. E. Knuth.Surreal numbers. Addison-Wesley Publishing Co., Reading, Mass.-London-Amsterdam, 1974
work page 1974
-
[43]
V.L. Kompelmakher and V.A. Liskovets. Sequential generation of arrangements by means of a basis of trans- positions.Kybernetica, 3:17–21, 1975
work page 1975
-
[44]
R.E. Korf. Depth-first iterative-deepening.Artificial Intelligence, 27:97–110, 1985
work page 1985
-
[45]
R.E. Korf and L.A. Taylor. Finding optimal solutions to the twenty-four puzzle. InProceedings of the AAAI Conference on Artificial Intelligence (AAAI-93), pages 756–761, 1993
work page 1993
-
[46]
Richard E. Korf. Linear-time disk-based implicit graph search.Journal of the ACM, 55(6):1–26, 2008
work page 2008
-
[47]
D. Kornhauser, G. Miller, and P. Spirakis. Coordinating pebble motion on graphs, the diameter of permutation groups, and applications. InProceedings of the 25th Annual Symposium, pages 241–250, 1984
work page 1984
-
[48]
Lucas.R´ ecr´ eations math´ ematiques, volume III
E. Lucas.R´ ecr´ eations math´ ematiques, volume III. Paris, 1883
-
[49]
M. Maschler, E. Solan, and S. Zamir.Game Theory. Cambridge University Text, 2013
work page 2013
-
[50]
R.J. Nowakowski M.H. Albert and D. Wolfe.Lessons in Play, An introduction to Combinatorial Game theory. A.K. Peters, Wellesley Massachusetts, 2007
work page 2007
-
[51]
S.J. Miller, M.T. Phaovibul, and M. Shen. God’s number of bi-colored cubes. 2024
work page 2024
-
[52]
Moscovich.The Puzzle Universe, A History of Mathematics in 315 puzzles
I. Moscovich.The Puzzle Universe, A History of Mathematics in 315 puzzles. Firefly Books, 2019
work page 2019
-
[53]
J. Von Neumann and O. Morgenstern.Theory of Games and Economic Behavior. Princeton University Press, 1953
work page 1953
-
[54]
Knill, R.M¨ ader, R.Sanders, J.Simon
O. Knill, R.M¨ ader, R.Sanders, J.Simon. The rotation group of Rubik’s cube.Sigsam Bulletin, 21, 1987
work page 1987
-
[55]
OEIS. A151944: Maximal number of moves required for the m x n generalization of the sliding block 15-puzzle. Anton Kulchitsky, referring Richard E. Korf (2008) and corrections Tomas Rokicki
work page 2008
-
[56]
Sequence a007664 (Reve’s puzzle)
OEIS Foundation. Sequence a007664 (Reve’s puzzle). Accessed: 2026-04-22
work page 2026
-
[57]
Orlin.Math Games with Bad Drawings
B. Orlin.Math Games with Bad Drawings. Black Dog and Leventhal Publishers, 2022
work page 2022
-
[58]
M.J. Osborne and A. Rubinstein.A Course in Game Theory. MIT Press, 1994
work page 1994
- [59]
- [60]
-
[61]
Pritchard.The Classified Encyclopedia of Chess Variants
D.B. Pritchard.The Classified Encyclopedia of Chess Variants. John Beasley, 2007. originally printed Biddles Ltd, Kings’s Lynn
work page 2007
-
[62]
D. Ratner and M. Warmuth. Finding a shortest solution for the nxn extension of the 15-puzzle is intractable. AAAI-86 Proceedings, 1986
work page 1986
-
[63]
D. Ratner and M. Warmuth. Finding a shortest solution for the (n×n)-extension of the 15-puzzle is intractable. Journal of Symbolic Computation, 10:111–137, 1990
work page 1990
- [64]
- [65]
-
[66]
Roberts.Thinking recursively with Java
E. Roberts.Thinking recursively with Java. Wiley, 2006
work page 2006
-
[67]
T. Rokicki, H. Kociemba, M. Davidson, and J. Dethridge. The diameter of the Rubik’s cube group is twenty. SIAM Journal on Discrete Mathematics, 27:1082–1105, 2013
work page 2013
-
[68]
A. E. Roth. Intoduction to the Shapley value. InThe Shapley value. Cambridge University Press, 1988
work page 1988
-
[69]
Rubik.Cubed, The puzzle of us all
E. Rubik.Cubed, The puzzle of us all. Flatiron Books, New York, 2020
work page 2020
-
[70]
U. Schwalbe and P. Walker. Zermelo and the early history of game theory.Games and Economic Behavior, 34:123–137, 2001
work page 2001
-
[71]
R.S. Scorer, P.M. Grundy, and C.A.B. Smith. Some binary games.Mathematical Gazette, 28:96–103, 1944
work page 1944
-
[72]
C. E. Shannon. Programming a computer for playing chess.Philosophical Magazine, 41(314), 1950. National IRE Convention on March 9, 1949
work page 1950
-
[73]
Shapley.A value for n-person games
L.S. Shapley.A value for n-person games. Rand Coorporation, 1952
work page 1952
-
[74]
D. Singmaster.Rubik’s magic cube. Enslow Publishers, 1981
work page 1981
-
[75]
S. Steinerberger. On the number of positions in chess without promotion.International Journal of Game Theory, 44:761–767, 2015
work page 2015
-
[76]
B.M. Stewart. Solution to advanced problem 3918.American Mathematical Monthly, 48:217–219, 1941
work page 1941
-
[77]
R. Uehara and S. Iweta. Generalized hi-q is np-complete.Trans IEICE 73, pages 270–273, 1990
work page 1990
-
[78]
A. van Zuylen, J. Bieron, F. Schalekamp, and G. Yu. An upper bound on the number of circular transpositions to sort a permutation, 2014
work page 2014
-
[79]
J. von Neumann. Zur theorie der gesellschaftsspiele.Mathematische Annalen, pages 295–320, 1928. Translated by S. Bargmann to English. 73 GRAPHS,GAMES,GROUPS
work page 1928
-
[80]
R. M. Wilson. Graph puzzles, homotopy, and the alternating group.Journal of Combinatorial Theory, Series B, 16:86–96, 1974
work page 1974
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.