A duplicate bridge movement is a scheme used in a duplicate bridge session to arrange which competitors play which opponents when, and which boards they play. The arrangement has to satisfy various constraints which often conflict to some extent, requiring compromises. The resolution of these compromises is to a considerable extent a matter of taste, so players should be consulted as to their preferences if this is practicable.
Movements are categorized by the type of event—Individual, Pairs, or Teams.
Requirements for duplicate bridge movements
Three absolute requirements for a bridge movement are universal: No entrant (individual, pair, or team, depending upon the type of the event) may play the same deal more than once. The number of deals in each session must be appropriate to the level of competition and the circumstances. The movement must provide enough stationary or nearly stationary positions to accommodate the needs of all players who have disabilities and/or mobility impairments.
A sanctioning body may impose additional requirements as a condition of sanctioning or recognizing an event. For example, the English Bridge Union (EBU) now requires that each entrant be "scheduled to play" at least 70% of the deals in play in each session in an individual or pairs game,Movement Chart, English Bridge Union which, practically means that the number of rounds must be at least 70% of the number of groups of boards in play. The American Contract Bridge League (ACBL) likewise requires at least 2 1/2 tables for a sanctioned standard pairs game, even though there is a valid movement for such a game, while permitting two tables to play an individual game with a "One Winner" Movement or a team matches. ACBL Codification, Chapter VIB, Section 1.5
With respect to the second universal requirement, a typical "open" (unrestricted) session at a club or tournament in North America consists of about 25–28 boards (deals) except when 24 deals produces a complete movement (that is, with six, eight, or twelve tables playing some form of complete Mitchell movement as described below), but this may vary in other regions. Of course, a club may choose to play fewer deals in a session connected with a luncheon, party, or some other event or to ensure that an evening session ends at a reasonable hour for players who must go to work on the next day, and sessions on cruise ships also typically are shorter (about 18–20 deals). Sessions for less experienced players also typically consist of fewer deals and, conversely, sessions of some championship events may play more deals.
That said, there are several criteria that are desirable in bridge movements: It is desirable for each table to play a separate group of boards in each round, unless there are enough copies of the groups played simultaneously for each table to have its own copy. Nevertheless, some movements frequently chosen for reasons of fairness do require two adjacent tables to play the same group of boards in each round. In matchpoint scoring, all boards ideally should be played the same number of times so they all carry the same weight and influence in the final standing. Two entrants normally should not play against one another (as opponents the same table) or, in an individual event, come together as partners in more than one round, except when players in an individual event change directions and thus must play against each other twice (with different partners, and ideally on opposite sides—that is, once as right-hand opponent and once as left-hand opponent) to form a complete movement. The number of boards per round must be reasonable. In standard matchpoint games and in "board-a-match" team games, two boards per round is the practical minimum and four to five boards per round normally is the desirable maximum. In standard team games other than "board-a-match" competition, six boards per match is the normal minimum and twenty-four boards per match is the normal maximum for open events. If there is an odd number of entrants in an individual or pair game, such that entrants must take turns sitting out round due to lack of a full table, it is highly desirable to minimize the number boards per round, and thus the time that the affected entrants must wait for the start of the next round. When there is no sit-out, the decision to play more rounds of fewer boards or fewer rounds of more boards is solely a matter of preference. It is generally thought to be desirable that all entrants play all of the deals that are in play. The movement should be as fair as possible, as discussed below. If there is a break for lunch or any other reason in the middle of a session, it is desirable to arrange the movement so all tables play one subset of the deals before the break and another subset of the deals after the break to ensure that any discussion of deals during the break won't compromise the deals in play after the break. Subject to the other constraints, the movement should be as simple as possible to minimize the possibility of errors such as players going to the wrong table, players sitting in the wrong direction at a table, or the players at a table playing the wrong group of boards. This applies to the movement of boards as well as to the movement of players.
The preferences of a particular club or tournament organizer may dictate other constraints. Here are some examples of organizers' preferences that might affect the choice of movement.
Some organizers may prefer to divide the tables of a session into flights for players of different ranking or experience levels with the flights playing in separate sections and scored separately. This choice typically requires a different movement for each section to accommodate differences in size of the sections and possibly also in the number of boards that the various levels of players might play. Other organizers may prefer to have players of all rankings or abilities play together in the same section(s).
Some organizers may prefer to split larger events or flights into several sections that play fewer rounds with more boards per round, finding that the game progresses more quickly with fewer rotations. Other organizers may prefer to form larger sections that play more rounds with fewer boards per round, giving each entrant an opportunity to play against more of the other entrants.
In larger events or flights, some organizers may prefer to rank all entrants into one comparison field to produce a single winner while other organizers may prefer to have multiple comparison fields with a separate winner for each comparison field. In events or flights with more than one section, there are several variations of these options—a separate comparison field for each direction within each flight, a single comparison field in each direction across multiple sections, a single comparison field across both directions within each section with section, or a single comparison field encompassing all entrants in the entire event.
In the absence of explicit requirements dictated by the nature of a particular competition, these considerations are completely discretionary.
Fairness of bridge movements
There are several considerations in determining the fairness of a bridge movement.
A complete movement, in which each entrant plays against all of the other entrants or in which all entrants in each scoring field play against all of the same field of opponents, is inherently the fairest choice. The worst scenario is a movement that is one round short of complete: one entrant does not play against a very strong competitor, thus gaining a significant advantage, while another entrant does not play against a very weak competitor, thus incurring a significant disadvantage. When a movement is several rounds short of complete, averaging tends to take over—each competitor misses both weaker and stronger opponents, diminishing the effect. It is possible to diminish this impact further by "seeding" the field—that is, by dispersing entrants with higher and lower rankings or ability uniformly throughout the field.
Ideally, each entrant should have equal influence on the results of each of the other entrants, at least within each scoring field, taking into account the relative positions in which the entrants play each deal—direct opposition (at the same table), indirect opposition (in the same direction at different tables, thus generating direct comparisons), teaming (in opposing directions at different tables), and, in individual movements, partnership (opposite seats at the same table). Mathematical analysis of standard matchpoint scoring has derived the weight of opposition in the following table based upon the impact of a hypothetical entrant getting either extreme (clear top or clear bottom) scores on all deals compared to a hypothetical entrant getting average scores on all deals. Positive weights means that the entrants are adversaries (an abnormally good result by either hurts the other) and negative weights mean that the entrants are effectively teammates -- that is, an abnormally good result by either helps the other. The term "scoring unit" is that of Law 78A of the Laws of Duplicate Bridge.
Relative Position on Deal Weight of Opposition in Matchpoint Scoring Opponent at Same Table +1 Scoring Unit x Number of Comparisons per Board Same Direction at Different Tables +1 Scoring Unit Opposite Directions at Different Tables -1 Scoring Unit Partners (Individual Movement Only) -1 Scoring Unit x Number of Comparisons per Board
When there is more than one scoring field in an event or a flight, all scoring fields should have entrants of comparable ability or ranking so that all competitors face competition that's approximately the same strength. In individual and pair events, this extends to each direction of each section when using movements in which each competitor always sits in the same direction. This is usually achieved by "seeding" -- that is, distributing stronger and weaker pairs uniformly throughout the field. Vulnerability affects the fairness of a game scored either by Total Points or by International Matchpoints (IMPs) because a pair that is vulnerable receives more points (raw Bridge score) than a pair that's not vulnerable for making the same number of tricks in any game or slam contract while giving up more points for failing to make any contract. These higher scores for the same Bridge result also translate to more IMPs for the same pair of different Bridge results in IMP scoring. When scoring by either of these methods, all pairs should be vulnerable on an equal number of deals—a condition most easily met by playing four boards per round since all regular groups of four boards (1–4, 5–8, etc.) have one board with each combination of vulnerability (none, North–South only, East–West only, and both), ensuring that each partnership is vulnerable on two deals and not vulnerable on the other two deals. Some movements that do not meet this criterion have very bad imbalances of vulnerability.Tim Delaney, "How Fair Are Howell Movements?" on "bridgewinners.com" web site, 12 June 2013. This issue does not apply to standard Matchpoint scoring because all boards award an equal number of matchpoints in each direction regardless of vulnerability.
When scoring multiple sessions as a single event, these criteria apply to the aggregate of all sessions of the event rather than to individual sessions, with the caveat that they apply separately to rounds before and rounds after any elimination of entrants.
Barometer games and online bridge movements