Filling Crosswords Is Very Hard

Authors Laurent Gourvès, Ararat Harutyunyan, Michael Lampis , Nikolaos Melissinos

Thumbnail PDF


  • Filesize: 0.75 MB
  • 16 pages

Document Identifiers

Author Details

Laurent Gourvès
  • Université Paris-Dauphine, Université PSL, CNRS, LAMSADE, 75016, Paris, France
Ararat Harutyunyan
  • Université Paris-Dauphine, Université PSL, CNRS, LAMSADE, 75016, Paris, France
Michael Lampis
  • Université Paris-Dauphine, Université PSL, CNRS, LAMSADE, 75016, Paris, France
Nikolaos Melissinos
  • Université Paris-Dauphine, Université PSL, CNRS, LAMSADE, 75016, Paris, France


We thank Dominique Taton for communicating the puzzle to us.

Cite AsGet BibTex

Laurent Gourvès, Ararat Harutyunyan, Michael Lampis, and Nikolaos Melissinos. Filling Crosswords Is Very Hard. In 32nd International Symposium on Algorithms and Computation (ISAAC 2021). Leibniz International Proceedings in Informatics (LIPIcs), Volume 212, pp. 36:1-36:16, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2021)


We revisit a classical crossword filling puzzle which already appeared in Garey&Jonhson’s book. We are given a grid with n vertical and horizontal slots and a dictionary with m words and are asked to place words from the dictionary in the slots so that shared cells are consistent. We attempt to pinpoint the source of intractability of this problem by carefully taking into account the structure of the grid graph, which contains a vertex for each slot and an edge if two slots intersect. Our main approach is to consider the case where this graph has a tree-like structure. Unfortunately, if we impose the common rule that words cannot be reused, we discover that the problem remains NP-hard under very severe structural restrictions, namely, if the grid graph is a union of stars and the alphabet has size 2, or the grid graph is a matching (so the crossword is a collection of disjoint crosses) and the alphabet has size 3. The problem does become slightly more tractable if word reuse is allowed, as we obtain an m^{tw} algorithm in this case, where tw is the treewidth of the grid graph. However, even in this case, we show that our algorithm cannot be improved to obtain fixed-parameter tractability. More strongly, we show that under the ETH the problem cannot be solved in time m^o(k), where k is the number of horizontal slots of the instance (which trivially bounds tw). Motivated by these mostly negative results, we also consider the much more restricted case where the problem is parameterized by the number of slots n. Here, we show that the problem does become FPT (if the alphabet has constant size), but the parameter dependence is exponential in n². We show that this dependence is also justified: the existence of an algorithm with running time 2^o(n²), even for binary alphabet, would contradict the randomized ETH. Finally, we consider an optimization version of the problem, where we seek to place as many words on the grid as possible. Here it is easy to obtain a 1/2-approximation, even on weighted instances, simply by considering only horizontal or only vertical slots. We show that this trivial algorithm is also likely to be optimal, as obtaining a better approximation ratio in polynomial time would contradict the Unique Games Conjecture. The latter two results apply whether word reuse is allowed or not.

Subject Classification

ACM Subject Classification
  • Theory of computation → Parameterized complexity and exact algorithms
  • Theory of computation → Approximation algorithms analysis
  • Crossword Puzzle
  • Treewidth
  • ETH


  • Access Statistics
  • Total Accesses (updated on a weekly basis)
    PDF Downloads


  1. Anbulagan and Adi Botea. Crossword puzzles as a constraint problem. In Principles and Practice of Constraint Programming, 14th International Conference, CP 2008, Sydney, Australia, September 14-18, 2008. Proceedings, pages 550-554, 2008. Google Scholar
  2. Edward K. Crossman and Sharyn M. Crossman. The crossword puzzle as a teaching tool. Teaching of Psychology, 10(2):98-99, 1983. Google Scholar
  3. Marek Cygan, Fedor V. Fomin, Lukasz Kowalik, Daniel Lokshtanov, Dániel Marx, Marcin Pilipczuk, Michal Pilipczuk, and Saket Saurabh. Parameterized Algorithms. Springer, 2015. Google Scholar
  4. Jakob Engel, Markus Holzer, Oliver Ruepp, and Frank Sehnke. On computer integrated rationalized crossword puzzle manufacturing. In Fun with Algorithms - 6th International Conference, FUN 2012, Venice, Italy, June 4-6, 2012. Proceedings, pages 131-141, 2012. Google Scholar
  5. Michael R. Garey and David S. Johnson. Computers and Intractability: A Guide to the Theory of NP-Completeness. W. H. Freeman, 1979. Google Scholar
  6. Matthew L. Ginsberg, Michael Frank, Michael P. Halpin, and Mark C. Torrance. Search lessons learned from crossword puzzles. In Proceedings of the 8th National Conference on Artificial Intelligence. Boston, Massachusetts, USA, July 29 - August 3, 1990, 2 Volumes, pages 210-215, 1990. Google Scholar
  7. Heather Hulett, Todd G. Will, and Gerhard J. Woeginger. Multigraph realizations of degree sequences: Maximization is easy, minimization is hard. Oper. Res. Lett., 36(5):594-596, 2008. Google Scholar
  8. Russell Impagliazzo, Ramamohan Paturi, and Francis Zane. Which problems have strongly exponential complexity? J. Comput. Syst. Sci., 63(4):512-530, 2001. Google Scholar
  9. Subhash Khot and Oded Regev. Vertex cover might be hard to approximate to within 2-epsilon. J. Comput. Syst. Sci., 74(3):335-349, 2008. Google Scholar
  10. Michael Lampis, Valia Mitsou, and Karolina Soltys. Scrabble is PSPACE-complete. J. Inf. Process., 23(3):284-292, 2015. URL:
  11. Michael L. Littman, Greg A. Keim, and Noam M. Shazeer. Solving crosswords with PROVERB. In Proceedings of the Sixteenth National Conference on Artificial Intelligence and Eleventh Conference on Innovative Applications of Artificial Intelligence, July 18-22, 1999, Orlando, Florida, USA, pages 914-915, 1999. Google Scholar
  12. Télé Magazine. Publications Grand Public. Google Scholar
  13. Gary Meehan and Peter Gray. Constructing crossword grids: Use of heuristics vs constraints. In In: Proceedings of Expert Systems 97: Research and Development in Expert Systems XIV, SGES, pages 159-174, 1997. Google Scholar
  14. Jagan A. Pillai, Charles B. Hall, Dennis W. Dickson, Herman Buschke, Richard B. Lipton, and Joe Verghese. Association of crossword puzzle participation with memory decline in persons who develop dementia. Journal of the International Neuropsychological Society, 17(6):1006-1013, 2011. Google Scholar
  15. Leonardo Rigutini, Michelangelo Diligenti, Marco Maggini, and Marco Gori. Automatic generation of crossword puzzles. Int. J. Artif. Intell. Tools, 21(3), 2012. Google Scholar
  16. Christopher D. Rosin. Nested rollout policy adaptation for Monte Carlo tree search. In IJCAI 2011, Proceedings of the 22nd International Joint Conference on Artificial Intelligence, Barcelona, Catalonia, Spain, July 16-22, 2011, pages 649-654, 2011. Google Scholar