Construction of balanced sports schedules using partitions into subleagues

https://doi.org/10.1016/j.orl.2007.09.007Get rights and content

Abstract

Direct constructions for balanced tournaments known so far solve problems with 2n teams if 2nmod31 or 2n=2p,p3 or n is odd. Our construction uses an arbitrary partition of the league into subleagues. It solves more than half of the missing cases and provides structured solutions for some known cases.

Introduction

We consider one of the most extensively studied sports scheduling problems: balanced tournament design (BTD) on multiple venues (also known as Prob026 from CSPLib). A balanced tournament design for 2n teams, denoted by BTD(n), is defined by conditions (1)–(4).

  • (1)

    There are 2n teams and n different stadiums.

  • (2)

    Every team plays one game against every other team.

  • (3)

    Each team plays exactly one game each day of the tournament.

  • (4)

    The stadiums where a team plays are distributed as equally as possible, i.e., each team plays twice in each stadium except one.

It has been shown in [8] that BTD(n) exists for all 2n>4. Nevertheless, obtaining concrete solutions is another challenge. Several promising methods have been developed using tabu search [1], [5], constraint programming [11], [12] and integer programming [9]. Although these methods achieved great success in recent years, algorithmic approaches (for instance based on integer programming formulation and therefore on implicit enumeration procedures) suffer from a prohibitive computational time as the size of the problem increases (starting at 2n=16 for basic methods and 2n=40 for more involved ones). An alternative to this approach is to provide some direct constructions; [8] suggests solutions if n is odd, [7] if 2nmod3=0 or 2 (this case is also solved in [6] by a linear time algorithm) and [3] if 2n is a power of 2. Notice that [3] gives more precisely solutions for a variant of BTD, denoted by BTD*, where each team plays exactly twice in each stadium. This is possible when each team plays once against every other team except one against which it plays twice; this is called a duplicated game. Here we have the additional constraint that duplicated games, involving the same teams, must not be played in the same stadium. However, it turns out that the given construction provides a schedule where all the duplicated games are played in the same day and therefore it is sufficient to remove that day in order to obtain a solution for BTD. As far as we know, the above cases include all direct constructions for BTD given in the literature so far. To summarize, the only case for which no direct construction is known is 2nmod12=4 and 2n2p for some integer p.

In this note, we give a direct construction for BTD(n) for specific values of n. In the construction method, we will arbitrarily partition the league of 2n teams into several subleagues of equal size (i.e., number of teams). This will allow us to derive a procedure exploiting this structure. Notice that this partition is simply a tool of the construction which allows us to have a procedure which produces a structured solution to BTD(n). In some cases however a partition into subleagues may be given from the beginning (based for instance on geographical requirements: each subleague may consist of teams which are located close to each other; see for example [2]). A well-known example of this is Major League Baseball in the United States (see [4]). Then, some requirements on the internal (respectively external) games (i.e., games played between teams of the same subleague (respectively, of different subleagues)) may have to be considered; for instance having in any day only internal games or only external games (see [2], [3]). Our procedure, while providing solutions to the basic problem BTD(n), allows us to take also into account requirements of the above type when a partition into subleagues is given beforehand.

To our knowledge, the method described in [3] is so far the only direct construction based on a partition into subleagues: the authors give a recursive method for 2n=2p with subleagues of size 2i where i may vary between 3 and p1. Our method, although not recursive, is another construction using subleagues that may contain several even numbers of teams and not only a number 2i. Moreover, the solutions obtained have the property that the games played each day are either all internal or all external. We provide solutions for more than half of the cases not covered by the known constructions. Another advantage of such a direct construction is that solutions can be constructed directly for arbitrarily large values of 2n. Last but not least, it gives very structured solutions using subleagues; even though another solution is already known for some values of 2n, having new types of schedules may be useful in order to satisfy various additional requirements, which may arise in different contexts. It is therefore important to have a wide sample of solutions, among which to choose the best one.

Section snippets

Balanced tournament design using subleagues

BTD(n) can be alternatively formulated as an arrangement of the n(2n1) distinct pairs of teams in the set T of 2n teams into a (2n1)×n array in such a way that every element of T occurs exactly once in each row (day) and at most twice in each column (stadium).

Let us recall that a Graeco-Latin square of order n is an n×n array where the entries contain all distinct ordered pairs of elements from two sets of size n in such a way that each element appears once in each row and once in each column.

Results and discussion

Clearly, Table 7 can be extended to infinity while leaving some holes, i.e. values of 2n for which no direct construction is known so far. Programming approaches (integer programming or constraint programming) leave no such holes but they are practically limited to relatively small values of n (see for instance [5], [11]).

As can be observed in Table 7, our construction gives BTD(mt) for all m odd and for all values of t such that a Graeco-Latin square of order t is known and BTD(t) can be

Acknowledgement

The research of Tınaz Ekim is supported by Grant 200020-113405 of the Swiss National Science Foundation.

References (12)

There are more references available in the full text version of this article.

Cited by (0)

View full text