Movatterモバイル変換


[0]ホーム

URL:


Jump to content
WikipediaThe Free Encyclopedia
Search

Tournament (graph theory)

From Wikipedia, the free encyclopedia
(Redirected fromTournament (mathematics))
Directed graph where each vertex pair has one arc
Tournament
A tournament on 4 vertices
Verticesn{\displaystyle n}
Edges(n2){\displaystyle {\binom {n}{2}}}
Table of graphs and parameters

Ingraph theory, atournament is adirected graph with exactly one edge between each twovertices, in one of the two possible directions. Equivalently, a tournament is anorientation of anundirected complete graph. (However, as directed graphs, tournaments are not complete: complete directed graphs have two edges, in both directions, between each two vertices.[1]) Equivalently, a tournament is acompleteasymmetricrelation.[2][3]

The nametournament comes from interpreting the graph as the outcome of around-robin tournament, a game where each player is paired against every other exactly once. In a tournament, the vertices represent the players, and the edges between players point from the winner to the loser.

Many of the important properties of tournaments were investigated byH. G. Landau in 1953 to model dominance relations in flocks of chickens.[4] Tournaments are also heavily studied invoting theory, where they can represent partial information about voter preferences among multiple candidates, and are central to the definition ofCondorcet methods.

If every player beats the same number of other players (indegree − outdegree = 0) the tournament is calledregular. The number of unlabeled regular tournaments with 2n+1 vertices goes:

1, 1, 1, 3, 15, 1223, 1495297, 18400989629, 2406183070160597,... (sequenceA096368 in theOEIS)

Paths and cycles

[edit]
a is inserted betweenv2 andv3.

Any tournament on afinite numbern{\displaystyle n} of vertices contains aHamiltonian path, i.e., directed path on alln{\displaystyle n} vertices (Rédei 1934).

This is easily shown byinduction onn{\displaystyle n}: suppose that the statement holds forn{\displaystyle n}, and consider any tournamentT{\displaystyle T} onn+1{\displaystyle n+1} vertices. Choose a vertexv0{\displaystyle v_{0}} ofT{\displaystyle T} and consider a directed pathv1,v2,,vn{\displaystyle v_{1},v_{2},\ldots ,v_{n}} inT{v0}{\displaystyle T\smallsetminus \{v_{0}\}}. There is somei{0,,n}{\displaystyle i\in \{0,\ldots ,n\}} such that(i=0viv0)(v0vi+1i=n){\displaystyle (i=0\vee v_{i}\rightarrow v_{0})\wedge (v_{0}\rightarrow v_{i+1}\vee i=n)}. (One possibility is to leti{0,,n}{\displaystyle i\in \{0,\ldots ,n\}} be maximal such that for everyji,vjv0{\displaystyle j\leq i,v_{j}\rightarrow v_{0}}. Alternatively, leti{\displaystyle i} be minimal such thatj>i,v0vj{\displaystyle \forall j>i,v_{0}\rightarrow v_{j}}.)v1,,vi,v0,vi+1,,vn{\displaystyle v_{1},\ldots ,v_{i},v_{0},v_{i+1},\ldots ,v_{n}}is a directed path as desired. This argument also gives an algorithm for finding the Hamiltonian path. More efficient algorithms, that require examining onlyO(nlogn){\displaystyle O(n\log n)} of the edges, are known. The Hamiltonian paths are in one-to-one correspondence with the minimalfeedback arc sets of the tournament.[5] Rédei's theorem is the special case for complete graphs of theGallai–Hasse–Roy–Vitaver theorem, relating the lengths of paths in orientations of graphs to thechromatic number of these graphs.[6]

Another basic result on tournaments is that everystrongly connected tournament has aHamiltonian cycle.[7] More strongly, every strongly connected tournament isvertex pancyclic: for each vertexv{\displaystyle v}, and eachk{\displaystyle k} in the range from three to the number of vertices in the tournament, there is a cycle of lengthk{\displaystyle k} containingv{\displaystyle v}.[8] A tournamentT{\displaystyle T} isk{\displaystyle k}-strongly connected if for every setU{\displaystyle U} ofk1{\displaystyle k-1} vertices ofT{\displaystyle T},TU{\displaystyle T-U} is strongly connected. If the tournament is 4‑strongly connected, then each pair of vertices can be connected with a Hamiltonian path.[9] For every setB{\displaystyle B} of at mostk1{\displaystyle k-1} arcs of ak{\displaystyle k}-strongly connected tournamentT{\displaystyle T}, we have thatTB{\displaystyle T-B} has a Hamiltonian cycle.[10] This result was extended byBang-Jensen, Gutin & Yeo (1997).[11]

Transitivity

[edit]
A transitive tournament on 8 vertices.

A tournament in which((ab){\displaystyle ((a\rightarrow b)} and(bc)){\displaystyle (b\rightarrow c))}{\displaystyle \Rightarrow }(ac){\displaystyle (a\rightarrow c)} is calledtransitive. In other words, in a transitive tournament, the vertices may be (strictly)totally ordered by the edge relation, and the edge relation is the same asreachability.

Equivalent conditions

[edit]

The following statements are equivalent for a tournamentT{\displaystyle T} onn{\displaystyle n} vertices:

  1. T{\displaystyle T} is transitive.
  2. T{\displaystyle T} is a strict total ordering.
  3. T{\displaystyle T} isacyclic.
  4. T{\displaystyle T} does not contain a cycle of length 3.
  5. The score sequence (set of outdegrees) ofT{\displaystyle T} is{0,1,2,,n1}{\displaystyle \{0,1,2,\ldots ,n-1\}}.
  6. T{\displaystyle T} has exactly one Hamiltonian path.

Ramsey theory

[edit]

Transitive tournaments play a role inRamsey theory analogous to that ofcliques in undirected graphs. In particular, every tournament onn{\displaystyle n} vertices contains a transitive subtournament on1+log2n{\displaystyle 1+\lfloor \log _{2}n\rfloor } vertices. The proof is simple: choose any one vertexv{\displaystyle v} to be part of this subtournament, and form the rest of the subtournament recursively on either the set of incoming neighbors ofv{\displaystyle v} or the set of outgoing neighbors ofv{\displaystyle v}, whichever is larger. For instance, every tournament on seven vertices contains a three-vertex transitive subtournament; thePaley tournament on seven vertices shows that this is the most that can be guaranteed.[12] However,Reid & Parker (1970) showed that this bound is not tight for some larger values of n{\displaystyle n}.[13]

Erdős & Moser (1964) proved that there are tournaments onn{\displaystyle n} vertices without a transitive subtournament of size2+2log2n{\displaystyle 2+2\lfloor \log _{2}n\rfloor } Their proof uses acounting argument: the number of ways that ak{\displaystyle k}-element transitive tournament can occur as a subtournament of a larger tournament onn{\displaystyle n} labeled vertices is(nk)k!2(n2)(k2),{\displaystyle {\binom {n}{k}}k!2^{{\binom {n}{2}}-{\binom {k}{2}}},}and whenk{\displaystyle k} is larger than2+2log2n{\displaystyle 2+2\lfloor \log _{2}n\rfloor }, this number is too small to allow for an occurrence of a transitive tournament within each of the2(n2){\displaystyle 2^{\binom {n}{2}}} different tournaments on the same set ofn{\displaystyle n} labeled vertices.[12]

Paradoxical tournaments

[edit]

A player who wins all games would naturally be the tournament's winner. However, as the existence of non-transitive tournaments shows, there may not be such a player. A tournament for which every player loses at least one game is called a 1-paradoxical tournament. More generally, a tournamentT=(V,E){\displaystyle T=(V,E)} is calledk{\displaystyle k}-paradoxical if for everyk{\displaystyle k}-element subsetS{\displaystyle S} ofV{\displaystyle V} there is a vertexv0{\displaystyle v_{0}} inVS{\displaystyle V\setminus S} such thatv0v{\displaystyle v_{0}\rightarrow v} for allvS{\displaystyle v\in S}. By means of theprobabilistic method,Paul Erdős showed that for any fixed value ofk{\displaystyle k}, if|V|k22kln(2+o(1)){\displaystyle |V|\geq k^{2}2^{k}\ln(2+o(1))}, then almost every tournament onV{\displaystyle V} isk{\displaystyle k}-paradoxical.[14] On the other hand, an easy argument shows that anyk{\displaystyle k}-paradoxical tournament must have at least2k+11{\displaystyle 2^{k+1}-1} players, which was improved to(k+2)2k11{\displaystyle (k+2)2^{k-1}-1} byEsther andGeorge Szekeres in 1965.[15] There is an explicit construction ofk{\displaystyle k}-paradoxical tournaments withk24k1(1+o(1)){\displaystyle k^{2}4^{k-1}(1+o(1))} players byGraham and Spencer (1971) namely thePaley tournament.

Condensation

[edit]

Thecondensation of any tournament is itself a transitive tournament. Thus, even for tournaments that are not transitive, the strongly connected components of the tournament may be totally ordered.[16]

Score sequences and score sets

[edit]

The score sequence of a tournament is the nondecreasing sequence of outdegrees of the vertices of a tournament. The score set of a tournament is the set of integers that are the outdegrees of vertices in that tournament.

Landau's Theorem (1953) A nondecreasing sequence of integers(s1,s2,,sn){\displaystyle (s_{1},s_{2},\ldots ,s_{n})} is a score sequence if and only if:[4]

  1. 0s1s2sn{\displaystyle 0\leq s_{1}\leq s_{2}\leq \cdots \leq s_{n}}
  2. s1+s2++si(i2), for i=1,2,,n1{\displaystyle s_{1}+s_{2}+\cdots +s_{i}\geq {i \choose 2},{\text{ for }}i=1,2,\ldots ,n-1}
  3. s1+s2++sn=(n2).{\displaystyle s_{1}+s_{2}+\cdots +s_{n}={n \choose 2}.}

Lets(n){\displaystyle s(n)} be the number of different score sequences of sizen{\displaystyle n}. The sequences(n){\displaystyle s(n)} (sequenceA000571 in theOEIS) starts as:

1, 1, 1, 2, 4, 9, 22, 59, 167, 490, 1486, 4639, 14805, 48107, ...

Winston and Kleitman proved that for sufficiently largen:

s(n)>c14nn5/2,{\displaystyle s(n)>c_{1}4^{n}n^{-5/2},}

wherec1=0.049.{\displaystyle c_{1}=0.049.}Takács later showed, using some reasonable but unproven assumptions, that

s(n)<c24nn5/2,{\displaystyle s(n)<c_{2}4^{n}n^{-5/2},}

wherec2<4.858.{\displaystyle c_{2}<4.858.}[17]

Together these provide evidence that:

s(n)Θ(4nn5/2).{\displaystyle s(n)\in \Theta (4^{n}n^{-5/2}).}

HereΘ{\displaystyle \Theta } signifies anasymptotically tight bound.

Yao showed that every nonempty set of nonnegative integers is the score set for some tournament.[18]

Majority relations

[edit]

Insocial choice theory, tournaments naturally arise as majority relations of preference profiles.[19] LetA{\displaystyle A} be a finite set of alternatives, and consider a listP=(1,,n){\displaystyle P=(\succ _{1},\dots ,\succ _{n})} oflinear orders overA{\displaystyle A}. We interpret each orderi{\displaystyle \succ _{i}} as thepreference ranking of a voteri{\displaystyle i}. The (strict) majority relationmaj{\displaystyle \succ _{\text{maj}}} ofP{\displaystyle P} overA{\displaystyle A} is then defined so thatamajb{\displaystyle a\succ _{\text{maj}}b} if and only if a majority of the voters prefera{\displaystyle a} tob{\displaystyle b}, that is|{i[n]:aib}|>|{i[n]:bia}|{\displaystyle |\{i\in [n]:a\succ _{i}b\}|>|\{i\in [n]:b\succ _{i}a\}|}. If the numbern{\displaystyle n} of voters is odd, then the majority relation forms the dominance relation of a tournament on vertex setA{\displaystyle A}.

By a lemma of McGarvey, every tournament onm{\displaystyle m} vertices can be obtained as the majority relation of at mostm(m1){\displaystyle m(m-1)} voters.[20] Results byStearns and Erdős & Moser later established thatΘ(m/logm){\displaystyle \Theta (m/\log m)} voters are needed to induce every tournament onm{\displaystyle m} vertices.[21]

Laslier (1997) studies in what sense a set of vertices can be called the set of "winners" of a tournament.[22] This revealed to be useful inpolitical science to study, in formal models ofpolitical economy, what can be the outcome of a democratic process.[23]

See also

[edit]

Notes

[edit]
  1. ^Weisstein, Eric W.,"Tournament",MathWorld
  2. ^Moulin, Hervé (1986)."Choosing from a tournament".Social Choice and Welfare.3 (4):271–291.doi:10.1007/BF00292732. RetrievedJanuary 19, 2025.A tournament is any complete asymmetric relation over a finite set A of outcomes describing pairwise comparisons.
  3. ^Laffond, Gilbert; Laslier, Jean-Francois; Le Breton, Michel (January 1993)."The Bipartisan Set of a Tournament Game".Games and Economic Behavior.5 (1):182–201.doi:10.1006/game.1993.1010. RetrievedJanuary 19, 2025.A tournament is a complete asymmetric binary relation U over a finite set X of outcomes.
  4. ^abLandau (1953).
  5. ^Bar-Noy & Naor (1990).
  6. ^Havet (2013).
  7. ^Camion (1959).
  8. ^Moon (1966), Theorem 1.
  9. ^Thomassen (1980).
  10. ^Fraisse & Thomassen (1987).
  11. ^Bang-Jensen, Gutin & Yeo (1997).
  12. ^abErdős & Moser (1964).
  13. ^Reid & Parker (1970).
  14. ^Erdős (1963)
  15. ^Szekeres & Szekeres (1965).
  16. ^Harary & Moser (1966), Corollary 5b.
  17. ^Takács (1991).
  18. ^Yao (1989).
  19. ^Brandt, Brill & Harrenstein (2016).
  20. ^McGarvey (1953);Brandt, Brill & Harrenstein (2016)
  21. ^Stearns (1959);Erdős & Moser (1964)
  22. ^Laslier (1997).
  23. ^Austen-Smith & Banks (1999).

References

[edit]

This article incorporates material from tournament onPlanetMath, which is licensed under theCreative Commons Attribution/Share-Alike License.

International
National
Other
Retrieved from "https://en.wikipedia.org/w/index.php?title=Tournament_(graph_theory)&oldid=1326142021"
Category:
Hidden categories:

[8]ページ先頭

©2009-2026 Movatter.jp