[pescetom]

I will be directing regional pairs (F2F) this weekend and it looks likely there will be 28-30 pairs.
I need to identify a single winning pair but the indications are to not make it too long.
My first thought was a Mitchell of 13 rounds x 2 boards with an arrow switch before last two rounds.
Does anyone see anything wrong with that, or a better alternative for some reason?

[jillybean]

web?

I've never run one, but here's a movement from Patrick Cater, NZ Bridge. I can put you in touch with Patrick if you want more info

Hi Kathryn,

16-22 tables playing 7 rounds you would need 4 sets of boards

3 sets of boards would be enough for 16, 17, 19 and 21 tables

If 4 sets is too many play 20 boards with 2 boards per table for the others

All of the movements you will need to set up a scoring file

I will just give you the 7 round movements

Let me know if you need me to detail the 10 round movements for 20 boards

15 Tables 7 Rounds

One set around tables 1-7 in normal layout

One set at tables 8-11 with feed ins in sequence coming in to table 11

Boards can flow right through tables 1-11 feeding out at table 1

Table 12 Boards 7-9

Table 13 Boards 4-6

Table 14 Boards 1-3

Table 15 Boards 19-21

Feed-ins at table 15.... 16-18.... 13-15.... 10-12

This set is flowing from 15 and feeding out at table 12

The players in this part of the movement play the sets in reverse order

[McBruce]

Considerations for pairs events to take into account:

How many sessions/boards? Here it sounds like we want one short-ish session of 20-24 boards
How many expected pairs? The estimate here is 28-30. Seldom in ACBL-land can we estimate that closely, but we'll go with it.
How many sets of boards can we have made, and is it enough for the movement we want to run? Not given, we'll assume more than one is possible and give options for only one.
What movements can the scoring software handle? Also not given, we'll assume that someone asking for movement advice can handle anything with whatever scoring software is in use.

What we want here is a one-winner movement that is as fair as practical. With one session and 28-30 pairs you can't have anything close to an all-play-all, but you should at least try to have each pair play as many different opponents as practical. I would certainly go with a movement with two-board rounds. Any sort of 'regional pairs' championship should not object to short rounds as some local club games, where people would rather move six times than eleven or twelve, seem to.

The advantage of a Web movement is that every board is in play the maximum number of times for a large top. If you have an even number of pairs, a Web movement will see each board played once at every table, instead of having [#tables x 2] boards that only go part-way through the room before the game ends, resulting in the post-mortem at the restaurant leaving a portion of the table out of the conversation on any single board. I usually explain Web movements' advantages by describing the post-game bar scene at a local final-day NAP qualifying game many years ago where the director was surprised at the attendance and ran a 44-board Mitchell with the help of masking tape over the board numbers in a second set: players spent more time drinking than discussing deals and needed cabs to take them home.

Arrow switching is, I am told, common in Europe and reviled in much of the ACBL, where players who have not sat East-West since 1957 will launch angry campaigns to have the Director hung, drawn and quartered if they don't get to play a North-South slam because of it. The purpose of arrow-switching is to intermix the matchpoint comparisons so that all comparisons are not between two groups, the N-S players, and the E-W players, making for a fairer result and reducing the advantage that can happen when the assignments are hastily made as the majority arrive with ten minutes to spare, and one of the groups is stronger than the other.

The combination of arrow-switching and Web movements is not as well-documented online as arrow-switching Mitchell movements. Since Webs have multiple sets of boards in play at one time, arrow-switching certain rounds in a Web makes it possible to overswitch some of the boards and underswitch others. An extreme example, in a 19-table, 12-round Web, there are 24 boards in play and in round 12, a round that is often arrow-switched, there are four copies of boards 3-4 in play and only one of boards 7-22.

The best solution for a game of this type is to see if you can get a list or an estimate of the pairs participating in advance and assign starting positions that:
--prevent one direction from having a clear majority of the pairs in any of the classes. It is just as important to distribute the best pairs as it is to distribute the worst!
--prevent one 'street' (consecutive tables in one direction) from having a 'murderer's row' of good pairs to be faced in consecutive rounds, or a 'power play' of the less experienced to be beat up on in consecutive rounds: since some will play against the whole 'street' and some will not.
--if you can successfully distribute the pairs in this way, there is less reason to worry about whether an arrow-switch is needed and you can concentrate on a good Web movement with a high top.

This is what we try to do in ACBL tournaments, with entries pre-marked for strat A, B, and C. My plan for most open pairs is to place the A+ pairs at 3 and 9 (by tradition, apparently originally set so that the seeded pairs would not skip one another after seven rounds in a 14-table Mitchell), and in a larger Web, also at 15, 21, 27. Put the rest of the A pairs at other odd-numbered tables, put the C pairs at tables divisible by 4, and the B pairs at the other even-numbered tables.

It doesn't always work out perfectly. Sometimes you end up erasing the strats pre-marked on your entries (hopefully in pairs) to change one number to another strat. Often the Cs and Bs arrive early and the As and A+s arrive late (or not at all) and you need to scramble. Sometimes the need for stationary pairs intervenes with the system and solutions need to be found (there's always someone who walks the room and then announces that they will only accept table 12 for some reason and the strat never fits...). But starting with a plan and minimizing the number of adjustments usually results in a well-balanced field, which is equally as important as choosing a good movement.

I think if you have the capability to make two full sets of boards, an 11-round Web for 15 tables is excellent. There may be a few rounds (usually in the middle and at the beginning and end) where you will need three copies of one or two sets, but these can easily be shared, or made from higher numbered-boards as needed (any board number 16 higher will work for dealer and vulnerability). An 11-round web for 14 tables is even easier, with tables 1-7 getting one set and 8-14 getting a second. If the players want to play a longer session or don't want to play only 20 boards if they are sitting out, there is a complication: Webs of odd-numbers of tables and even numbers of rounds do not work, so 15 tables and 12 rounds would be complicated (14 tables and 12 rounds are fine.) There are workarounds for these problems but they are either quite complicated, or imperfect, or both.

My link for Web movements is here:

https://www.bridgewe...d=display_page5

There is a pdf explaining how Web movements work as well as pre-sets for Web movements of all types for ACBLscore.

If only one board set is available, a Mitchell of 14 or 15 tables (possibly with a sitout) will be fine. Pairs will miss only 2-4 boards each and an arrow switch will create a fair single winner. Do what you can to balance the field; don't rely on the arrow switch to do it for you.

Good luck!

[mycroft]

Very nice McBruce!

I would state that (all signs to the contrary noted and ignored :-) if you do have two sessions, you're in great shape. You can put all the N-S into a 7 table Howell (unless you require more than one absolute stationary and one "could move once"), and all the E-W into another one, and play another 13 rounds. With 28 pairs, you play 26 of them (you can do the Web or the Skip Mitchell the first session; Web gets you a matching top and no complaints about "missing the slam we obviously would have found"); pretty much the best result possible.

With 30, you have either a sitout in both Howells or the Interwoven Howell (which ACBLscor has, not sure about anywhere else), where the two "sitouts" play each other. With 15 tables and 13 rounds of the Mitchell, you can arrange (if you're *very* careful, but if you get it right the first round, the rest will follow) to have that "playback" be one of the two they didn't play the first time.

With 29, the goal (of course) is to find a pair you can stick in the hole, whether it's one session or two :-).

One more thing that is possible; if 28 (or 30!) boards are acceptable in a 1-session event (a serious enough 1-session event might be okay with this), you can play the share-relay 14 (two sets of boards best here) or the full straight Mitchell 15. Yeah, yeah, "indications say 'not too long'". If that's the prime criterion, do two sections, 7x3 Mitchell, arrow-switch the last round and call it a day :-).

[pescetom]

Thanks very much to all. A web movement sounds the way to go, but it's not in our software and I wouldn't like to risk a first attempt in an important tournament anyway. Sometime later in my club.

I know the number of pairs fairly precisely, as this is the regional final and 30 pairs were qualified of whom 27 (currently) are turning up.

What I didn't know until this afternoon is that there is also a Final B to be run at the same time with another 13 pairs, and obviously they have only one PC available and one BM Server

So fancy movements go out the window this time, I will already be at my limit running two tournaments one of them without BM (and without experience with BM in this context, although I did handle 140 of them with different SW for WBF).
Mitchell with arrow switch (and well balanced lines to start with) it is (and if NS beef about missing a slam because the movement had them playing EW that round, I will ask if they ever complained about a Howell).
Maybe I could handle the final B with a 13 x 2 board Howell and thus run both tournaments with the same start, finish and round changes... think that through tomorrow morning as I have a tournament to play now