[pescetom]

You are running a pairs tournament and decide to run a single session splitting the field into two sections each playing the same boards in a Mitchell and then merge the results.
If the organiser desires a single winner, is it fair to let the scoring program decide this in the merge?
Would it help if both sections were arrow switched?

[jillybean]

I've fwd this to our Master of Movements. but I doubt that he will like being called that.

[McBruce]

jillybean, on 2026-May-18, 09:47, said:

I've fwd this to our Master of Movements. but I doubt that he will like being called that.

Wait, what? That's me you're referring to?

In ACBL this is now the norm for any matchpoint game that expands to more than one section; we call it scoring across the field, and most scoring programs will do it for you (or allow you to score and/or rank within each section). Few ACBL tournaments care too much for overall awards to be given out based on an arrow switch balancing the NS and EW field, comparisons-wise, but that's just ACBL, and if your players are used to it, it has fairness benefits.

Few tournaments here have huge pair games anymore, but I recall two decades ago some of the opening night Penticton Regional pair games used to have as many as seven or eight sections, and the TDs to my surprise would combine them for scoring but would have two groups of 3-4 sections for some unknown reason. This made it possible that a score of +980 would be worth more matchpoints in Section A than in Section G, but it might have been that in the early days of BridgeMates we were unsure how many a computer could handle at once!

[jillybean]

Thanks Bruce,

I forgot to mention, pescetom is with the FIGB but it sounds like same will apply.

[mycroft]

Yes, score across both sections. I don't mind the 25 table webs, but two 12s playing 24 is quite reasonable.

Obviously, in the top events, they will score across all sections there. Day 2 of the National Mixed Pairs I was in, I was surprised to see a 64 top with 5 sections.

The issue is that after about 25 comparisons, the differences almost never are significant. In sections A and B, +980 will be maybe 20/25; in C D and E, +980 will be something like 30/38. It'll be off by < 5% on all but one or two boards in a session. Almost always, the differences balance out over 26 boards to near 1%. Now, yes, that *could* determine the winner. But so could randomly hitting the crazy grand against the relay Precision pair that can find it, or hitting the one pair that can execute the pentagonal squeeze (or are good enough to be able to destroy it before it happens!)

The downside of scoring all the way across the field (once computers came in - hand-matchpointing a 64 top seems like an Lovecraftian Horror) is that one mistake/adjusted score/reported 90-degrees/appeal means you're reprinting *every recap*. (No comment about "with everything on Live, I've stopped printing recaps unless the tournament requests it" here, we still do it in Penticton) You can go through 100 pages of 11x17 fanfold in a single session really quickly that way.

Obviously, if you live anywhere but the ACBL, and you want a single winner, you run a one-winner movement.

If you're running a two-session event, you can arrange for the crossover between sections to be nice, in a way that a crossover within a single section - isn't.

One thing that is a debate in the ACBL is "rank across sections" for the purposes of section (not overall) awards. There is a distinct argument for "no", that being "well, your 'section' is the pairs that played the same opponents as you". There's a distinct argument for "yes", that being "your section is the pairs that you score against." (there's a distinct argument for "4 section tops" (which means you don't rank across), especially in a Gold Rush event, because section tops are Gold points. There's an equivalent argument for "two Gold awards each direction, rank across" in that the awards are higher - but it could be that both awards go to the same section and it "looks bad".) I'm sure none of these are relevant for FIGB (but, probably, replaced by considerations of similar importance).

[pescetom]

Thanks to both.

Mycroft, you say "you can arrange for the crossover between (the two) sections to be nice", but you don't spell out how
Are you saying that the scoring program will probably do it nicely with two sections, or that arrow switching both sections is useful and sufficient, or that Director should do something manually to obtain a fairer crossover than the program may offer?

[mycroft]

We do manual-ish crossovers, but here's "standard" in ACBL:

For A and B:

• A E-W sits A N-S at same table in second session.
  
• A N-S goes to B E-W.
  
• Similarly, B E-W to B N-S and B N-S to A E-W.
  
• There's a toggle on the screen to swap table "m" N-S and E-W's assignments; do that for all the "need to be double N-S" pairs.

Notice that:

• most pairs play against the other section's pairs straight up, so obviously new people.
  
• the ones that get "double Northed" end up playing against "the people they scored against in the first session" - again, new people.

And that most pairs get one sit-session and one move-session.

For three sections, same thing, but A->B->C->A.

There are issues when there are two sections and they are not the same size (and usually, they're off by one, which is worst), but you can usually work things out. With more than two sections, you can resolve the issue cleanly. Apart from that issue, it's basically mindless.

With one section, best you can do (barring double or combined Howells, shhh) is move E-W up enough tables that they hit all the first-session misses for the first few rounds, and then swap about half the pairs so that even when they hit the played line again, it's only half that are playbacks.

One large issue (may or may not be a downside) is that you "score" against (most of) the same pairs each time - which matters if you're not ranking across sections, more if you're not scoring across sections. One slightly less large issue (unless you're "that pair", then it's incredibly annoying) is that the pair you're following in the first session is the pair passing boards to you in the second session (or vice versa) - and if they're the ones who play every round except the first and the one after the break in 15 minutes (but the others in 20)...

This all is independent of how the movement in each session is planned and "how many winner movement" we have in each session.

[McBruce]

mycroft, on 2026-May-18, 19:00, said:

The downside of scoring all the way across the field (once computers came in - hand-matchpointing a 64 top seems like an Lovecraftian Horror) is that one mistake/adjusted score/reported 90-degrees/appeal means you're reprinting *every recap*. (No comment about "with everything on Live, I've stopped printing recaps unless the tournament requests it" here, we still do it in Penticton) You can go through 100 pages of 11x17 fanfold in a single session really quickly that way.

Aha! That makes perfect sense; been wondering about that since it happened in like 2003 before I was even directing at clubs and just the daily bulletin guy wondering why ABC and DEFG were scored together in two groups rather than just all seven. "That's just the way we do it," was the answer I got at the time.

As for printing recaps, it took me one tournament of printing (on laser), pasting them up, walking to the wall to post them, then noticing I was the only one in the room and had to collect and box four sets of boards in perhaps ten minutes before the next sale was due to start, before I decided it was futile. Now the process is "yes, you can have a summary sheet if you insist, but I have to get all sections printed first. Write your section and pair number and direction down here, it'll be about ten minutes or so." And I grab the press without recaps and run them through a word processor to 14-inch paper and a custom monospaced font, so I can get 15-16pt print for the results, rather than the ACBLscore laser printer standard of 6 or 7 or 8pt which nobody can read...

[McBruce]

Another important issue with multiple sections, single session, one winner desired is that the entry seller(s) must make some effort to ensure that the two or more sections have similar numbers of good players, excellent players, and random sources of free matchpoints. I get more complaints about the top players being all one way or all within one of the sections than I do about the ones who are good for three or four random 1400s a game if you can double them at the right time. Both have effects on the scoring, especially if they are not fairly evenly distributed, both into all sections and directions, and not as consecutive pairs within those sections. You may have the perfect number of "A" pairs in each section, but if they are all at tables 4, 5, 6, 7, 8, and 9, the E-W pairs will move from table 3 and see an uphill climb for the next two hours that might not be a lot of fun. To make this happen in ACBL we still sell paper entries prepared well in advance with some scheme by which tables are pre-assigned to different levels.

Up against this in 2026 is the ever-growing number of pairs who have played forever and need a North-South, which sometimes appears to distort the balance, except that a lot of these players have the masterpoints but no longer the level of play they indicate, so it's usually less of an issue than it appears.

An infamous horror story on this theme: a local club in the 1980s was about to run one of the first World-Wide Pairs (each board came with a special scoreslip with possible scores pre-listed along with an Instant Matchpoint score of 0-100, so it didn't matter what others in your section did, it mattered what the list thought was a good result on the board), then called the Epson Pairs. With Omar Sharif involved as the booklet editor, signup was so good they needed extra space and an extra seller/director. A room down the hall used on weekdays by (I kid you not) the local Psychic Society was rented for the Saturday Afternoon, and became the non-smoking section, with a fine local player as the seller and playing TD. He later said "based on the weird scoring, I thought it best to seat the experienced players North-South, reserving table 1 for myself and partner, and those many that I simply did not know as East-West. Normally, this would be suicide, since I would be comparing scores against all the other experienced pairs for regular matchpoints, but with Instant Matchpoints, facing a steady stream of excited new East-West players who could be counted on to overbid and underplay, I was surprised none of us rolled a world-beater!" I was one of those unknowns and went home happy to have just broken average against the superstar North-South field. No credit in the world-wide placings for this achievement though....