These are variations on a standard Mitchell and allow 24 boards to be played in movements where there is a large number of tables. Gordon Rainsford explains more. In the last issue of Club Focus, John Pain identified the problem we have when we don't have enough tables for the number of rounds we would like to play, and he offered the solution of the Hesitation Mitchell. Sometimes, however, the problem is the opposite - we have too many tables for the number of rounds we would like to play.
Most clubs deal with this by putting out extra board-sets, and this is acceptable if there are not too many extra. However, at the most extreme, I’ve known of a club playing a Mitchell in a single section of 26 tables, with 52 boards in play of which each pair only played 24. Obviously, this is undesirable, since pairs will only have a few boards in common with most other competitors, and all pairs will be competing against a few other pairs with whom they have not a single board in common!
You could, of course, run two separate sections and combine the results afterwards, but unless you have exactly the right number of tables you are still likely to have spare board-sets in circulation - and then there's the problem of uneven section strengths, the possibility of a three-board sit-out if there's a half-table, and it's simply a better competition if you play against more pairs.
There are some movements that are specifically designed for having two tables more than the number of board-sets - "Bowman" movements, also known as "1½ table Appendix Mitchells" to be found in the EBU movements manual (Manning) - but they have historically been unpopular because one of the tables shares boards with a different table each round, playing boards in descending order - unless there is a half table (hence the movement's alternative name). With the advent of board duplication and wireless scoring, those problems have largely been removed, and a Bowman is a sensible way to play 24 boards for 10 tables, 27 boards for 11 tables, or 26 boards for 15 tables. However, when the number of tables is more than two greater than the number of board-sets, the usefulness of a Bowman movement diminishes. In particular, the very large sections we talked about earlier would have only slightly less of a problem with a Bowman than with a regular Mitchell.
With this in mind, a chance mention in a post by Marvin French on the Bridge Laws Mailing List (BLML) drew my attention to Web movements, which were invented in the 1970s by an American TD, John "Spider" Harris (hence "Web"), but had never really gained popularity because they required two sets of boards - and America was a long way behind Europe in adopting pre-dealt boards. These movements aren't listed in the EBU movements manual (Manning) or in the Swedish Jannersten one (though I later found out that they are mentioned in the Australian McKinnon directing manual of 1979). However, with some further research online, I managed to identify the basic structure of these movements, and found that they allow large single sections with a fixed number of boards (Appendix Mitchell movements do this too, but those have the major disadvantage of needing 26 moving pairs).
In its simplest form a Web movement has an even number of tables and an odd number of board-sets, and all pairs get to play all the boards. It's ideal for playing 26 boards (13 board-sets) with 14, 16, 18, 20, 22 or 24 tables, or for playing 27 boards (9 board-sets) with 10 tables. The way it works is that you put out your first complete set of boards in the normal way to the lower half of the room (e.g. for a 16-table movement playing 26 boards put out board-sets 1 to 8 on tables 1 to 8 with the remainder on a relay table between tables 8 and 1). Then, with your second set of boards, put out the highest numbered board-set on the highest numbered table (set 13 on table 16), board-set 1 on the next table down (table 15), then board-set 2 and so on until you have board-set 7 on table 9 (with the remainder on a relay table between tables 9 and 16). It's not really complicated and, of course, once you have the movement in your scoring program (using the matrices referred to at the end of this article), the scoring devices will keep everyone straight. There are plans for the Web movements to be included among the standard movements in future versions of ScoreBridge.
The movement of pairs is completely straightforward - NS pairs are stationary, and EW move up a table each round - no skip move is required. If you want a one winner movement you can arrow-switch the last two rounds in the usual way. The boards also move down one table in the usual way, but each set of boards remains in its own half of the movement. So boards that have been played at table 1 go to a relay and then get fed back in at the half-way table (table 8 in our example 16-table movement). Boards also go to a relay after being played at the bottom table of the top half of the movement (table 9) and then get fed back in at the top table (table 16). You will note that NS pairs play the boards in ascending order in the bottom half of the movement, but in descending order in the top half. EW pairs will not play boards in any consistent pattern.
There is a variation to this basic Web movement that allows it to be played with an odd number of tables, although it does involve some sharing unless there is a third complete set of boards. The way it works is that the entire first set of boards is put out on the lower-numbered tables (e.g. board-sets 1 to 13 on tables 1 to 13), and the remaining tables (which will always be an even number) are treated like an even-numbered Web movement for the purpose of boards. So, for 19 tables, table 14 would start with board-set 1, table 15 board-set 2, table 16 board-set 3, table 17 board-set 2, table 18 board-set 1, and table 19 board-set 13 (i.e. tables in this section have to share boards unless there is a third complete set of boards). As always, EW pairs move up the entire section, and the boards move down within their sub-sections. Movement mavens will notice that in its simplest variant (a 15-table movement with 13 board-sets), the Web movement is identical to the Bowman movement we mentioned above. In fact a Bowman movement is simply one specific instance of a Web movement!
Having drawn Max Bavin's attention to these movements, he used them to great effect at the EBU's Summer Meeting in Brighton this August. They were ideal for the "Play with the Experts" Pairs, which requires everyone to play the same boards, and were used in many of the Open Pairs events, as well as in the Seniors Pairs semi-finals. Then Max took the basic idea and, together with Ian Mitchell at the EBU, extended it for use running the Pivot Teams - an event which is run with two mirrored sections of teams, where once again everyone needs to play exactly the same boards.
Finally, Jim Proctor, another EBU National TD, noticed that the basic Web movement, with an even number of tables, can also be used with an even number of board-sets, simply by employing the familiar technique of having a skip-move halfway through the movement. In fact the Australian Ian McKinnon (who has a new movements publication out soon) had already noted this. Many clubs like to play 24 boards in a session, so with 16 tables you would put out board-sets 1 to 8 on tables 1 to 8. Then, with your second set of boards, put out board-set 12 on table 16, board-set 1 on table 15, then board-set 2 and so on until you have board-set 7 on table 9. You then have a skip move after 6 rounds. Doubtless there are still other variations yet to be found, but with this basic idea most clubs that are equipped with wireless scoring and pre-duplicated boards can ensure that all their players usually play exactly the same boards as everyone else in the event, which certainly reduces the element of luck in the results - as well as making for more interesting discussions in the bar after play.