EBU Score teams error on manually created movement
Hi. We use Bridgemates with EBU Score. We have generated a number of movements for our weak / strong random teams events. Recently I created a new movement for 6 tables. It seems to work, but partway through the session, EBU Score repeatedly gives me the error message Error 9 (Subscript out of range) in procedure SaveNSPlayers of FormfrmSeating.
At the end we can fix it by adding a few players names into EBU Score..
I have compared the new movement to the old working ones, and I can't see anything that appears to be incorrect. Is someone able to help me fix this, please?
The problem has occurred with the 6 table staggered movement, though I suspect that the non-staggered movement also has the same issue.
Thanks
Peter Burns
Chichester BC
Comments
I don't think it's the movement, which seems to be ok. You can always check a movement by saving it under a new name in EBUScore - it will protest if there is a problem there. Is your version of EBUScore up to date?
The Staggered movement is invalid. In Rd 1 you have 6 v 1, 4 v 3, 2 v 5, 3 v 6, 5 v 1, 2 v 4 so Team 2 is NS at both Table 3 and 6! The remaining movement has the same Team at Tables 3 and 6! I am not sure what was the intended, but it has confused the program with an impossible Lineup.
The other nonStaggered movement looks fine.
I am a bit surprised that the validity check allowed this, so that code needs tightening up to disallow saving this movement.
I have just worked out a fix which in fact showed that Team 1 is EW at Tables 1 and 5 in Round 1. All the other Chichester movements are valid.
Thank you for this. Are you are able to share the fix, please, or is the staggered movement just not possible?
These random teams sessions are well supported (better than any normal teams event) because we pair a weaker and a stronger pair together in a team, and the weak pairs only play against other weak pairs, and the strong against other strong pairs. So the movements are basically 2 Howell movements, with the polarity swapped in half of the tables. For 6 tables this means that on the last round all tables require boards 21-25 (2* 5 boards between 6 tables). To make this easier to manage, I just started tables 4-6 on round 3 of the Howell, meaning that they required boards 21-25 on round 3 rather than on round 5, and could share 10 boards between 3 tables.
The Fix was just to the Validity check which shows the movement is not feasible. You cannot have a Teams NS pair or Teams EW pair playing at two different tables in the same same direction in the same round! I think you can redesign the staggered movement such that it is reordered and thus feasible.
I think you've slightly missed the point though Jeff. In a normal teams movement each team would have one pair playing NS throughout and one pair playing EW throughout. Peter wants to do something different: each team should have one pair playing at tables 1-3 throughout and one pair playing at tables 4-6 throughout, but they might each switch directions. As Peter said, it's two Howells with pair N in the first Howell scoring up with pair N in the second Howell (arrow-switching the second Howell so that this makes sense).
For more than 3 tables, Peter's movement also means that each team has one pair playing NS and one playing EW in each round. So it all works fine in the scoring system, except that presumably the TD has to tell the pairs to ignore the movement shown in the scoring devices, and you wouldn't get cross-imps calculated correctly.
But for 3 tables he has rotated the rounds of one of the Howells in order to avoid excessive board sharing. That's ought to work out fine, but it means that now sometimes a team has both of their pairs playing NS at the same time, so it's not supported.
There must be a workaround for this. Is there a way to define two sections and score up between them? Maybe even set it up as a teams of 8 and have the TD mark half the boards as not played?
Failing that, could you artificially double the number of rounds? Have tables 1-3 playing in round 1 and tables 4-6 playing in round 2, with the other pairs "sitting out" (but in fact you're going to play both "rounds" simultaneously)?
You can set up a mirrored movement with two sections, simply by creating a 3-table Howell movement with NS & EW reversed and using that for one of the sections.
Hi, Further to this discussion, I have got involved at Chichester with trying to overcome the issues with these random teams movements. They are as follows, and I think PeterB has mentioned them, but we haven't been able to resolve the following:
1 wrong names in the bridgemates.
2 seating lineup - when and how to do it?
3 ximps not correct because of wrong pair being assigned assigned the scores. ie teammates' scores appearing in own place - one's scorecard is a mix of your own and your other pairs' scores
4 is this type of event valid for NGS, as the two halves of the room have a separate competition each so ximps isn't a fair balance across the room, as opposed to normal movements where stronger and weaker players are randomly moving around the room
5 there are a few minor issues, which may be resolved if these can be.
The random teams have two mirrored movements, as discussed previously, usually a howell movement, and seats are staggered, so teammates sit opposite one other across the room in opposite directions, ie NS at one and EW at another. Let's say lower half tables are the stronger pair of the team, and the higher half tables are the weaker.
So some stronger players start off NS in rd 1 and some EW, in their half of the room. In round 2, those who started rd 1 as EW and then switch to a NS seat in their half of the room; the bridgemates show their teammates' names rather than their own. This isn't an issue with the team's score, but it affects the ximps, and it isn't ideal as some pairs aren't aware of their partners' names, so creating a bit of confusion. It would be nice to sort this out . I have tried to do it with the seating lineup, but that doesn't seem to update the names in the bridgemates, nor change the scoring. Also, the pairs sitting NS get chosen as 1 and 2 in the seating lineup by default, ans are 1 and 2 in the team. It would be so much easier if seating lineup could do this by the team members names from round 1. Teamscorer appears to report from the NS viewpoint. eg at end of session, it showed 25 results for each team, allthough there were 50 ( 25 each seating direction). I can see why it is done this way, but there appears to be no way to tell the bridgemates which pairs should be there, and no way of reassigning the team members' names so that it is from the stronger pairs' point of view.
The Bridgemate problem is insoluable with the current Bridgemate software. In Kent for our teams-of-8 we have 3 movements of lines so that every pair on each time plays every other pair on every other team. This means that in the lat round pairs play against pairs who have been playing in their line up to now. This requires a Howell type movement. At teams the Bridgemate software will always expect a pair which starts NS to remain NS and a pair which starts EW to remain EW. therefore it gets the names wrong in this form of movemnent. For this stanza we therefore set the Bridgemates to only display names on the first round and not to display moving instructions and rely on players following the table cards.
We play this last round barometer style with a fair bit of board-sharing. The seating-line upworks fine for us and shows the correct pairs playing the correct boards against the correct opponents and calculates the X-imps correctly, although Bridgewebs has some difficulty displaying all this information correctly. However I thinkn this does not provide a solution for the problem David Collier identifies and I am not sure what does.
Thanks for that Paul. Much appreciated.
So I confirm - it is not possible with present software to resolve the issue no 1 without making eg a 4 section event, and making half of the tables sit-out in order to get the correct seating and boards played. At least this would generate the correct results list, and have the correct players' names in the Bridgemates. They could then be imported into a spreadsheet, and thus merged for the teams scores, as well as for ximps calculations based on own results. Is it a major task to then create the UMS file for masterpoints and ngs from such a spreadsheet?
Point 2 I'm still not clear as to how the seating lineup works if it doesn't send the movement as lined-up to the bridgemates?
I confirm it doesn't work if one updates movement from round 1 nor round 2 onwards once the bridgemates are connected. (these attempts have just left the bridgemate names the same, and the results sent to ebuscore were still mixed, ie some of own results, and some of teammates, ie no apparent effect on bridgemates). Does it work in ebuscore only, ie and only to be used after the event is completed, before creating final reports? And if so, how do you activate this change? Is it automatic as you change the values? I tried 12 34, and either left them or swopped to 34 12 according to the way actually played. for all teams and all rounds. It seemed to have no effect on the results.
Point 4: validity of ngs for such an event. The published methods I've seen of calculation of teams ngs is not very clear as to how exactly it is done. From the various descriptions I have seen, I believe it uses the ximps on each board, as well as the sopps of each round (session -ie the players against whom you played only). Is there somewhere to find the authoritative description of exactly how it is calculated for a teams event as opposed to pairs. If it IS based on the ximp score, this is derived from an unbalanced room, ie weaker pairs on one side and stronger pairs on the other. For a large number of tables, I think it becomes statistically more valid, but in eg a six table movement, there are only 2 other scores from the same side (Weak or Strong) and 3 from the other for calculation of ximps. For each side there is a 0 and 100 in mps, (as the movement in that half is a complete howell, (not sure if this affects anything)), if not scored across the room, and so the ximps when calculated across the room does not reflect the sopps of only the one side of the room.
My feeling is that we should scrap the NGS for such events, unless it is a truer reflection of the standard of opponents as played. As a 3 table howell there are only 2 other scores with which to compare with players of the sopp against whom you played, whereas ximps takes all 5 other scores across the room. Hope this makes sense - please correct me if wrong.
Page 14 of the Full Guide to the NGS describes how Multiple Teams and Swiss Teams events are re-scored as if it were a Cross-IMP Pairs. Swiss Teams events are graded match by match. Teams of Eight matches are possibly included now but only if results are submitted electronically from a scoring program, though I've never been aware of this ever having actually happened.
The matter of three table Howells and other events where there are few comparisons per board is being looked at.
Barrie Partridge - CTD for Bridge Club Live
Thanks Barrie,
That was as I have previously read, and thought there was a more comprehensive description somewhere, ie it uses ximps as generated in the scoring program, as I had thought from the descriptions I had read. Therefore the arguments as discussed above should hold and should not be used for NGS for a small number of teams in our mirrored events where the room is split by NGS grade for Stronger and Weaker halves, as the SOPPs calculated would be for your side of the room in which you played (Strong or Weak), but the ximps calculated across the room - therefore only becoming more valid as the number of teams increases,.