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:
Any tournament on afinite number of vertices contains aHamiltonian path, i.e., directed path on all vertices (Rédei 1934).
This is easily shown byinduction on: suppose that the statement holds for, and consider any tournament on vertices. Choose a vertex of and consider a directed path in. There is some such that. (One possibility is to let be maximal such that for every. Alternatively, let be minimal such that.)is a directed path as desired. This argument also gives an algorithm for finding the Hamiltonian path. More efficient algorithms, that require examining only 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 vertex, and each in the range from three to the number of vertices in the tournament, there is a cycle of length containing.[8] A tournament is-strongly connected if for every set of vertices of, 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 set of at most arcs of a-strongly connected tournament, we have that has a Hamiltonian cycle.[10] This result was extended byBang-Jensen, Gutin & Yeo (1997).[11]
A tournament in which and 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.
Transitive tournaments play a role inRamsey theory analogous to that ofcliques in undirected graphs. In particular, every tournament on vertices contains a transitive subtournament on vertices. The proof is simple: choose any one vertex to be part of this subtournament, and form the rest of the subtournament recursively on either the set of incoming neighbors of or the set of outgoing neighbors of, 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 .[13]
Erdős & Moser (1964) proved that there are tournaments on vertices without a transitive subtournament of size Their proof uses acounting argument: the number of ways that a-element transitive tournament can occur as a subtournament of a larger tournament on labeled vertices isand when is larger than, this number is too small to allow for an occurrence of a transitive tournament within each of the different tournaments on the same set of labeled vertices.[12]
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 tournament is called-paradoxical if for every-element subset of there is a vertex in such that for all. By means of theprobabilistic method,Paul Erdős showed that for any fixed value of, if, then almost every tournament on is-paradoxical.[14] On the other hand, an easy argument shows that any-paradoxical tournament must have at least players, which was improved to byEsther andGeorge Szekeres in 1965.[15] There is an explicit construction of-paradoxical tournaments with players byGraham and Spencer (1971) namely thePaley tournament.
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]
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 is a score sequence if and only if:[4]
Let be the number of different score sequences of size. The sequence (sequenceA000571 in theOEIS) starts as:
Insocial choice theory, tournaments naturally arise as majority relations of preference profiles.[19] Let be a finite set of alternatives, and consider a list oflinear orders over. We interpret each order as thepreference ranking of a voter. The (strict) majority relation of over is then defined so that if and only if a majority of the voters prefer to, that is. If the number of voters is odd, then the majority relation forms the dominance relation of a tournament on vertex set.
By a lemma of McGarvey, every tournament on vertices can be obtained as the majority relation of at most voters.[20] Results byStearns and Erdős & Moser later established that voters are needed to induce every tournament on 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]
^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.
^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.
Austen-Smith, D.; Banks, J. (1999),Positive Political theory, University of Michigan Press
Bang-Jensen, J.;Gutin, G.; Yeo, A. (1997), "Hamiltonian Cycles Avoiding Prescribed Arcs in Tournaments",Combinatorics, Probability and Computing,6 (3):255–261,doi:10.1017/S0963548397003027
Takács, Lajos (1991), "A Bernoulli Excursion and Its Various Applications",Advances in Applied Probability,23 (3), Applied Probability Trust:557–585,doi:10.2307/1427622,JSTOR1427622.