Arrow-switching in teams of eight
In an inter-county teams-of-eight league yesterday, I was called to a table that had started board 9 at 90 degrees to the correct orientation. I told them to continue, and switched it at another table to maintain four comparisons. I entered "A" for arrow-switched into the traveller for that board in EBUScore at both tables, and I assumed all was now fine.
Another table in a different match played board 1 arrow-switched, and didn't notice until the end of the round. I entered "A" into the traveller for just that table, assuming it would do a sensible calculation with the remaining scores. It did, and the final cross-IMP score on the board agreed with my manual calculation according to WB3.7.2.2. However, something strange happened when I came to put the pair line-ups in. It wouldn't allow me to enter a pair from this team at one of the tables in that round. I had three pairs correctly seated, no one at the fourth table, the fourth pair on the "bench", but the player names and positions at the table were inactive, couldn't be clicked on, and couldn't be changed. (There weren't even empty player numbers at the fourth table.)
I thought maybe the programme doesn't like arrow-switching a single table, so I switched the pairs back and put an average in for the switched board. The cross-IMP score remained correct, and I was now able to enter the pair names.
The only thing that doesn't seem correct is the Butler ranks. Instead of all players playing 32 boards, it has half of the third team (where I switched two boards) playing either 25 or 39.
Does anyone know why it's doing this, and if there's anything I can do to fix it?
Comments
Well, perhaps we shouldn't ask how a table managed to play the first board of a set rotated through 90 degrees. & indeed how this should happen twice!
The fact that this was the first board might explain why EBUScore struggled to understand the seating line-ups. Presumably it looks at the traveller patterns, and tries to find two north-south pairs & two east-west pairs.
Have you tried setting the line-ups before applying the A/S to board 1?
I can't guarantee that this will work. If not, you might have to swap around the scores (at all tables) between boards 1 & 8 (same vulnerabilities, even if they don't match the hand records & dealership).
It wouldn't surprise me if XIMPs for the second board set (9-16?) are mis-assigned, for the two tables that played the first board in the wrong polarity.
Thanks, I'll try this when I'm next at the club (Wednesday) and report back. I suppose if necessary I could leave the scores unswitched and make a manual adjustment (in IMPs).
I put the scores back in, set the pair movements then put in the arrow-switches and averages. It all seemed to work fine, although for some reason the arrow-switches are not showing up on the website results. It may have been my mistake and I'll try again tonight. Thanks.
A related observation.
If there is an inadvertent arrow-switch in teams of eight and the TD fixes this by arrow-switching the board when it is played in the other section then EBUScore may not be able to score the board.
As long as there are two N/S scores, and two E/W scores, you can get a teams result by attributing those scores to the pairs scheduled to be playing in those polarities (even if in the wrong section). It just means that the XIMP reports might not be accurate.
I double-checked my adjustments because one of the players complained the website was showing one of his teammates playing his contract on the switched board. The scoring programme is showing that the teams have been arrow-switched so the scores are going to the correct teams. I suspect that when I tell EBUScore that on boards 9-16 in the white section Jim and Bob are playing EW for the home team and Tony and Sue NS, it sticks to this rigidly and assumes the pairs switched tables when they switched direction. Or it might be the way bridgewebs reads the output from EBUScore.
I'm not entirely convinced the Butler ranks will be correct, but I can live with that.
Well, it is probably self-evident that the "Seating Lineups" assume that whatever the lineup is for the first board of the round, it will be the same for the rest of the round. The issue is whether it reads the line-ups from their scheduled table on board one, or from the polarity in which they sit on board one (after switches have been applied).
It is for this reason that i suggested swapping the scores between boards 1 & 8 - so that at the very worst only one set of XIMPs will be wrong, rather than 7.
It doesn't surprise me that EBUScore can't work out what's going on, but that's only because I've used it so much and got to know its limitations. Had I not been so familiar with it it would have been self-evident to me that if you tell it which pairs are sitting at a table in which orientation for a round, and then tell it that they have played a board the wrong way round, it could work out which cards they were holding on that board and assign the scores to the correct teams and the correct players.