Log in

Minimal retentive sets in tournaments

  • Original Paper
  • Published:
Social Choice and Welfare Aims and scope Submit manuscript

Abstract

Tournament solutions, i.e., functions that associate with each complete and asymmetric relation on a set of alternatives a nonempty subset of the alternatives, play an important role in the mathematical social sciences at large. For any given tournament solution \(S\), there is another tournament solution  which returns the union of all inclusion-minimal sets that satisfy \(S\)-retentiveness, a natural stability criterion with respect to \(S\). Schwartz’s tournament equilibrium set (\({ TEQ }\)) is defined recursively as . In this article, we study under which circumstances a number of important and desirable properties are inherited from \(S\) to . We thus obtain a hierarchy of attractive and efficiently computable tournament solutions that “approximate” \({ TEQ }\), which itself is computationally intractable. We further prove a weaker version of a recently disproved conjecture surrounding \({ TEQ }\), which establishes —a refinement of the top cycle—as an interesting new tournament solution.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Subscribe and save

Springer+ Basic
EUR 32.99 /Month
  • Get 10 units per month
  • Download Article/Chapter or Ebook
  • 1 Unit = 1 Article or 1 Chapter
  • Cancel anytime
Subscribe now

Buy Now

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Fig. 1
Fig. 2
Fig. 3
Fig. 4
Fig. 5
Fig. 6

Similar content being viewed by others

Notes

  1. NP-hardness is commonly seen as strong evidence that a problem cannot be solved efficiently. The interested reader is referred to the articles by Woeginger (2003) and Brandt and Fischer (2008) for a more detailed discussion.

  2. This definition slightly diverges from the common graph-theoretic definition where \(\succ \) is defined on \(A\) rather than \(X\). However, it facilitates the sound definition of tournament solutions.

  3. \({\widehat{\gamma }}\) is a variant of the better-known expansion property \(\gamma \), which, together with Sen’s \(\alpha \), figures prominently in the characterization of rationalizable choice functions (Brandt and Harrenstein 2011).

  4. Our terminology differs slightly from those of Laslier (1997) and others. Independence of unchosen alternatives is also called independence of the losers or independence of non-winners. The weak superset property has been referred to as \(\epsilon ^+\) or as the Aïzerman property.

  5. See, e.g.,  Laslier (1997) for definitions of these tournament solutions.

  6. It can easily be shown that \(S^{(\ell )}(T) = { TEQ }(T)\) for all \(T \in \mathcal T _n\) and \(\ell \ge k_{S}(n)\).

  7. A set \(B \subseteq A\) is \(R\)-undominated if \((a,b) \in R\) for no \(b \in B\) and \(a \in A \setminus B\).

References

  • Alon N (2006) Ranking tournaments. SIAM Journal on Discrete Mathematics 20(1):137–142

    Article  Google Scholar 

  • Arrow KJ, Raynaud H (1986) Social Choice and Multicriterion Decision-Making. MIT Press

  • Basu K, Weibull J (1991) Strategy subsets closed under rational behavior. Economics Letters 36:141–146

    Article  Google Scholar 

  • Bouyssou D, Marchant T, Pirlot M, Tsoukiàs A, Vincke P (2006) Evaluation and Decision Models: Step** Stones for the Analyst. Springer-Verlag

  • Brandt F (2011) Minimal stable sets in tournaments. Journal of Economic Theory 146(4):1481–1499

    Article  Google Scholar 

  • Brandt F, Fischer F (2008) Computing the minimal covering set. Mathematical Social Sciences 56(2):254–268

    Article  Google Scholar 

  • Brandt F, Harrenstein P (2010) Characterization of dominance relations in finite coalitional games. Theory and Decision 69(2):233–256

    Article  Google Scholar 

  • Brandt F, Harrenstein P (2011) Set-rationalizable choice and self-stability. Journal of Economic Theory 146(4):1721–1731

    Article  Google Scholar 

  • Brandt F, Fischer F, Harrenstein P (2009) The computational complexity of choice sets. Mathematical Logic Quarterly 55(4):444–459

    Article  Google Scholar 

  • Brandt F, Fischer F, Harrenstein P, Mair M (2010) A computational analysis of the tournament equilibrium set. Social Choice and Welfare 34(4):597–609

    Article  Google Scholar 

  • Brandt F, Chudnovsky M, Kim I, Liu G, Norin S, Scott A, Seymour P, Thomassé S (2013) A counterexample to a conjecture of Schwartz. Social Choice and Welfare 40:739–743

    Article  Google Scholar 

  • Conitzer V (2006) Computing Slater rankings using similarities among candidates. In: Proceedings of the 21st National Conference on Artificial Intelligence (AAAI), pages 613–619. AAAI Press

  • Duggan J, Le Breton M (1996) Dutta’s minimal covering set and Shapley’s saddles. Journal of Economic Theory 70:257–265

    Article  Google Scholar 

  • Dung PM (1995) On the acceptability of arguments and its fundamental role in nonmonotonic reasoning, logic programming and n-person games. Artificial Intelligence 77:321–357

    Article  Google Scholar 

  • Dunne PE (2007) Computational properties of argumentation systems satisfying graph-theoretic constraints. Artificial Intelligence 171(10–15):701–729

    Article  Google Scholar 

  • Dutta B (1990) On the tournament equilibrium set. Social Choice and Welfare 7(4):381–383

    Article  Google Scholar 

  • Fisher DC, Ryan J (1995) Tournament games and positive tournaments. Journal of Graph Theory 19(2):217–236

    Article  Google Scholar 

  • Good IJ (1971) A note on Condorcet sets. Public Choice 10:97–101

    Article  Google Scholar 

  • Houy N (2009a) Still more on the tournament equilibrium set. Social Choice and Welfare 32:93–99

    Article  Google Scholar 

  • Houy N (2009b) A few new results on TEQ. Mimeo

  • Laffond G, Laslier J-F, Le Breton M (1993a) More on the tournament equilibrium set. Mathématiques et sciences humaines 31(123):37–44

  • Laffond G, Laslier J-F, Le Breton M (1993b) The bipartisan set of a tournament game. Games and Economic Behavior 5:182–201

    Article  Google Scholar 

  • Laffond G, Lainé J, Laslier J-F (1996) Composition-consistent tournament solutions and social choice functions. Social Choice and Welfare 13:75–93

    Article  Google Scholar 

  • Laslier J-F (1997) Tournament Solutions and Majority Voting. Springer-Verlag

  • Moulin H (1986) Choosing from a tournament. Social Choice and Welfare 3:271–291

    Article  Google Scholar 

  • Schwartz T (1990) Cyclic tournaments and cooperative majority voting: A solution. Social Choice and Welfare 7:19–29

    Article  Google Scholar 

  • Woeginger GJ (2003) Banks winners in tournaments are difficult to recognize. Social Choice and Welfare 20:523–528

    Article  Google Scholar 

Download references

Acknowledgments

This material is based on work supported by the Deutsche Forschungsgemeinschaft under grants BR 2312/3-3, BR 2312/6-1, BR 2312/7-1, and FI 1664/1-1. Preliminary versions of the results were presented at the Workshop on Algorithmic Aspects of Game Theory and Social Choice (Auckland, February 2010), the Dagstuhl Seminar on Computational Foundations of Social Choice (Dagstuhl, March 2010), the Doctoral School on Computational Social Choice (Estoril, April 2010), and the 9th International Joint Conference on Autonomous Agents and Multi-Agent Systems (Toronto, May 2010).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Markus Brill.

Appendix: Proof of Theorem 3

Appendix: Proof of Theorem 3

Theorem 3

Let \(S\) be a tournament solution such that \(\mathcal R _S\) is pairwise intersecting, and let \(\mathsf{{P}}\) be any of the properties \(\mathsf{SSP }, \mathsf{WSP }, \mathsf{IUA }, \mathsf{MON } \wedge \mathsf{SSP },\) or \({\widehat{\gamma }}\wedge \mathsf{SSP } \). Then, \(\mathsf P\) is satisfied by \(S\) if and only if it is satisfied by .

Proof

Assume that \(\mathcal R _S\) is pairwise intersecting. We need to show that each of the properties \(\mathsf{SSP }, \mathsf{WSP }, \mathsf{IUA }, \mathsf{MON } \wedge \mathsf{SSP } \), and \({\widehat{\gamma }}\wedge \mathsf{SSP } \) is satisfied by \(S\) if and only if it is satisfied by . The direction from right to left follows from Theorem 2. We now show that the properties are inherited from \(S\) to .

Assume that \(S\) satisfies SSP. Let \(T=(A,\succ )\) be a tournament, and consider an alternative . We need to show that , where \(T^{\prime }={(A\setminus \{x\},\succ )}\). We will show that

(1)

This implies that is a minimal \(S\)-retentive set in \(T^{\prime }\). Since \(\mathcal R _S\) is assumed to be pairwise intersecting, is indeed the unique minimal \(S\)-retentive set in \(T^{\prime }\), i.e., .

To show (1), consider an arbitrary . If \(x\notin \overline{D}_A(a)\), then obviously \(\overline{D}_A(a)=\overline{D}_{A\setminus \{x\}}(a)\) and thus \(S(\overline{D}_A(a))=S(\overline{D}_{A\setminus \{x\}}(a))\). Assume on the other hand that \(x\in \overline{D}_A(a)\). Since and , it follows that \(x\notin {S}(\overline{D}_A(a))\), as otherwise would not be \(S\)-retentive. Now, since \(S\) satisfies \(\mathsf{SSP } \), we obtain \(S(\overline{D}_A(a))=S(\overline{D}_{A\setminus \{x\}}(a))\) as desired.

Assume that \(S\) satisfies WSP. Let \(T=(A,\succ )\) be a tournament, and consider an alternative . We need to show that , where \(T^{\prime }=(A\setminus \{x\},{\succ })\). Since \(\mathcal R _S\) is assumed to be pairwise intersecting, it suffices to show that is also \(S\)-retentive in \(T^{\prime }\). To this end, consider an arbitrary . Since \(S\) satisfies WSP, we have that \(S(\overline{D}_{A\setminus \{x\}}(a))\subseteq S(\overline{D}_A(a))\). Furthermore, by \(S\)-retentiveness of , and thus .

Assume that \(S\) satisfies IUA. Let \(T=(A,\succ )\) and \(T^{\prime }=(A,{\succ ^{\prime }})\) be tournaments such that for all \(a\in A\). We need to show that . We will show that

(2)

This implies that is a minimal \(S\)-retentive set in \(T^{\prime }\). Since \(\mathcal R _S\) is assumed to be pairwise intersecting, we have .

To show (2), consider an arbitrary . \(S\)-retentiveness of in \(T\) implies that . Therefore, it follows from the definition of \(T\) and \(T^{\prime }\) that \(T|_{S(\overline{D}_{\succ }(a),\succ )\cup \{b\}}=T^{\prime }|_{S(\overline{D}_{\succ }(a),\succ )\cup \{b\}}\) for all \(b \in \overline{D}(a)\). Now, since \(S\) satisfies \(\mathsf{IUA } \), we obtain \(S(\overline{D}_{\succ }(a),\succ ) = S(\overline{D}_{\succ ^{\prime }}(a),\succ ^{\prime })\) as desired.

Assume that \(S\) satisfies \(\mathsf{MON } \) and \(\mathsf{SSP } \). We have already seen that \(\mathsf{SSP } \) is inherited, so it remains to be shown that satisfies \(\mathsf{MON } \). The following argument is adapted from the proof of Proposition 3.6 in Laffond et al. (1993). Let \(T=(A,{\succ })\) be a tournament, and consider two alternatives \(a,b\in A\) such that and \(b\succ a\). Let \(T^{\prime }=(A,\succ ^{\prime })\) be the tournament with \(T|_{A\setminus \{a\}}=T^{\prime }|_{A\setminus \{a\}}\) and \(D_{\succ ^{\prime }}(a) = D_{\succ }(a) \cup \{b\}\).

We have to show that . To this end, we claim that for all \(c\in A\setminus \{a\}\),

$$\begin{aligned} a \notin S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime }) \quad \text{ implies } \quad S(\overline{D}_{\succ }(c),\succ ) = S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime }). \end{aligned}$$
(3)

Consider the case when \(c\ne b\) and assume that \(a\notin S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime })\). It follows from monotonicity of \(S\) that \(a \notin S(\overline{D}_{\succ }(c),\succ )\). To see this, observe that monotonicity of \(S\) implies that \(a\in S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime })\) whenever \(a\in S(\overline{D}_{\succ }(c),\succ )\).

Now, since \(S\) satisfies \(\mathsf{SSP } \),

$$\begin{aligned} S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime })&= S(\overline{D}_{\succ ^{\prime }}(c) \setminus \{a\},\succ ^{\prime }) \quad \text{ and } \\ S(\overline{D}_{\succ }(c),\succ )&= S(\overline{D}_{\succ }(c) \setminus \{a\},\succ ). \end{aligned}$$

It is easily verified that \((\overline{D}_{\succ ^{\prime }}(c)\setminus \{a\},\succ ^{\prime }) = (\overline{D}_{\succ }(c)\setminus \{a\},\succ )\), thus we have \({S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime })}=S(\overline{D}_{\succ }(c),\succ )\).

If \(c=b\), then \(a\notin S(\overline{D}_{\succ ^{\prime }}(b),{\succ ^{\prime }})\) together with \(\mathsf{SSP } \) of \(S\) implies \({S(\overline{D}_{\succ ^{\prime }}(b),\succ ^{\prime })}=S(\overline{D}_{\succ ^{\prime }}(b)\setminus \{a\},\succ ^{\prime })\). Furthermore, by definition of \(T\) and \(T^{\prime }\), \((\overline{D}_{\succ ^{\prime }}(b)\setminus \{a\},\succ ^{\prime }) = (\overline{D}_{\succ }(b),\succ )\) and thus \(S(\overline{D}_{\succ ^{\prime }}(b),\succ ^{\prime })= S(\overline{D}_{\succ }(b),{\succ })\). This proves (3).

We proceed to show that . Assume for contradiction that this is not the case. We claim that this implies that

(4)

To see this, consider . We have to show that . Since, by assumption, , we have that \(a \!\notin \! S(\overline{D}_{\succ ^{\prime }}(c),\succ ^{\prime })\). We can thus apply (3) and get

which, together with the \(S\)-retentiveness of in \(T^{\prime }\), implies (4).

Having assumed that \(\mathcal R _S\) is pairwise intersecting, it follows from (4) that . Hence, , a contradiction. This shows that satisfies \(\mathsf{MON } \).

Finally assume that \(S\) satisfies \({\widehat{\gamma }}\) and SSP. We already know from the above that  satisfies SSP, so it remains to be shown that satisfies \({\widehat{\gamma }}\). Let \(T=(A,\succ )\) be a tournament, and consider two subsets \(B_1, B_2 \subseteq A\) such that . We have to show that . We will show that

$$\begin{aligned} S(\overline{D}_{B_1 \cup B_2}(c))=S(\overline{D}_{B_1}(c)) \quad \text{ for } \text{ all } c\in C. \end{aligned}$$
(5)

This implies that \(C\) is a minimal \(S\)-retentive set in \(T|_{B_1 \cup B_2}\). Since \(\mathcal R _S\) is assumed to be pairwise intersecting, we have .

To show (5), consider an arbitrary \(c \in C\). Since and are \(S\)-retentive in \(T|_{B_1}\) and \(T|_{B_2}\), respectively, we have \(S(\overline{D}_{B_i}(c))\subseteq C \subseteq B_1 \cap B_2\) for \(i \in \{1,2\}\). The fact that \(S\) satisfies SSP now implies \(S(\overline{D}_{B_1 \cap B_2}(c)) = S(\overline{D}_{B_1}(c))\) and \(S(\overline{D}_{B_1 \cap B_2}(c)) = S(\overline{D}_{B_2}(c))\), and thus \(S(\overline{D}_{B_1}(c)) = S(\overline{D}_{B_2}(c))\). Since \(S\) satisfies \({\widehat{\gamma }}\), we have \(S(\overline{D}_{B_1 \cup B_2}(c))=S(\overline{D}_{B_1}(c) \cup \overline{D}_{B_2}(c))=S(\overline{D}_{B_1}(c))\), as desired. \(\square \)

Rights and permissions

Reprints and permissions

About this article

Cite this article

Brandt, F., Brill, M., Fischer, F. et al. Minimal retentive sets in tournaments. Soc Choice Welf 42, 551–574 (2014). https://doi.org/10.1007/s00355-013-0740-4

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1007/s00355-013-0740-4

Navigation