[pescetom]

I've seen several references here and on BW to the idea of stratifying pairs as A, B or C level in order to optimise seating positions or other aspects of a tournament. Is this as common in the US (perhaps elsewhere) as the writers seem to assume? It would be unusual here, to say the least.

How does it work? Who decides the stratification level and with what criteria (combined RA points for the two players? Some specific RA ranking for pairs? Personal judgement of Director?)? Is it specific to the local club? Is there a public list or is it just known to Director? Is there no problem of people who are disgruntled by their assigned level?

[barmar]

ACBL club games and tournaments are almost always stratified this way, as this allows maximizing the number of pairs that win masterpoints. And if you're using a two-winner movement (e.g. Mitchell), you generally try to balance the distribution between NS and EW, so that pairs will have to play against opponents of different levels; it would be unfair if all the C pairs had to play against all the A pairs, and also to be able to award points better (you need to have at least 2 pairs in a stratum to be able to award points for that stratum).

Strats are usually assigned based on combined masterpoints, for lack of any better objective criteria. Tournaments and some special club games have fixed masterpoint limits for each stratum (e.g. C = 0-499, B = 500-1499, A = 1500+), but regular club games can set them however they like, often trying to arrange for roughly equal numbers of players in each stratum.

[pescetom]

 barmar, on 2019-September-12, 11:47, said:

ACBL club games and tournaments are almost always stratified this way, as this allows maximizing the number of pairs that win masterpoints. And if you're using a two-winner movement (e.g. Mitchell), you generally try to balance the distribution between NS and EW, so that pairs will have to play against opponents of different levels; it would be unfair if all the C pairs had to play against all the A pairs, and also to be able to award points better (you need to have at least 2 pairs in a stratum to be able to award points for that stratum).

Strats are usually assigned based on combined masterpoints, for lack of any better objective criteria. Tournaments and some special club games have fixed masterpoint limits for each stratum (e.g. C = 0-499, B = 500-1499, A = 1500+), but regular club games can set them however they like, often trying to arrange for roughly equal numbers of players in each stratum.

Thanks. Is all this widely accepted? Here it would be difficult. There is already resistance if the Director tries to balance the distribution between NS and EW by moving a few pairs according to his judgement. I suspect an explicit stratification would encounter rebellion, even if based upon RA points. Many players do not even know (or pretend not to know) their points and even those who earnestly collect them admit that they do not really reflect ability. And the national culture is suspicious of stratification and has a strong sense of justice. Maybe if there was a reak ranking based upon result quality like in UK it would be different.

[barmar]

 pescetom, on 2019-September-12, 12:39, said:

Thanks. Is all this widely accepted? Here it would be difficult. There is already resistance if the Director tries to balance the distribution between NS and EW by moving a few pairs according to his judgement. I suspect an explicit stratification would encounter rebellion, even if based upon RA points. Many players do not even know (or pretend not to know) their points and even those who earnestly collect them admit that they do not really reflect ability. And the national culture is suspicious of stratification and has a strong sense of justice. Maybe if there was a reak ranking based upon result quality like in UK it would be different.

Yes, it's widely accepted, I'm not sure why it would be resisted. If the field isn't balanced, a poor pair could end up being matchpointed against an excessive number of advanced pairs, and they'll have little chance. Or they'll face lots of advanced opponents, which will also disadvantage them.

Club directors generally know the level of all the regulars.

The main resistence I've heard about are players that need (or just demand) a stationary table. If there are too many of these that have to be accomodated, this can skew the field.

[pescetom]

 barmar, on 2019-September-13, 09:36, said:

Yes, it's widely accepted, I'm not sure why it would be resisted. If the field isn't balanced, a poor pair could end up being matchpointed against an excessive number of advanced pairs, and they'll have little chance. Or they'll face lots of advanced opponents, which will also disadvantage them.

Club directors generally know the level of all the regulars.

The main resistence I've heard about are players that need (or just demand) a stationary table. If there are too many of these that have to be accomodated, this can skew the field.

Ok, I guess it's just a cultural difference. Yes the club director knows the level of all the regulars, that's the problem - when he makes his evaluation public then many of the B and C pairs are going to resent it and some contest it. Probably they would more easily accept a stratification based upon the club's internal classification rather than RA points, although that's even more a measure of attendance rather than ability.

Yes we too have the problem of players that need (or just demand) a stationary table and it tends to skew the field. Short of demanding a medical certificate, the only remedy I can see to that is to always assign sitouts to the stationary line, although they grumble about that too.

[blackshoe]

I doubt very much that it's possible to arrange a game of duplicate in such a way that no one ever complains.

[barmar]

 pescetom, on 2019-September-13, 10:06, said:

Ok, I guess it's just a cultural difference. Yes the club director knows the level of all the regulars, that's the problem - when he makes his evaluation public then many of the B and C pairs are going to resent it and some contest it.

It's not just the TD who knows it -- all the regulars know it. The results are posted on the wall, on the web site, etc. It's no secret who is regularly at the bottom of the results.

And it's not like the TD announces "You sit at table 2 because you suck". In games where the TD assigns seats as people arrive, you can't easily tell what criteria is being used. Our club lets players choose their own seats. On the rare occasions when we have enough for a Mitchell, the TD looks around the room just to make sure we don't have too many A pairs all sitting the same direction (again, it's no secret who they are ), and makes one or two switch directions.

No one is being ostracized.

[blackshoe]

In my experience, over the past few years the only reason pairs are asked to move to a different starting spot is to allow some other pair to be stationary. I don’t think I have seen any attempt to balance the field in at least the last five years and it’s probably been ten or more.

[pran]

 blackshoe, on 2019-September-15, 01:26, said:

In my experience, over the past few years the only reason pairs are asked to move to a different starting spot is to allow some other pair to be stationary. I don’t think I have seen any attempt to balance the field in at least the last five years and it’s probably been ten or more.

I have the same experience, but I must add that we normally use barometer movements which give the best balance possible (as has been shown mathematically). So we just do not worry about balance.

[barmar]

 blackshoe, on 2019-September-15, 01:26, said:

In my experience, over the past few years the only reason pairs are asked to move to a different starting spot is to allow some other pair to be stationary. I don’t think I have seen any attempt to balance the field in at least the last five years and it’s probably been ten or more.

It tends to be needed only in tiny games, like 7 or 8 tables. With fewer tables you'll probably run a Howell movement, so there's only one field, and with more tables you're less likely to have a concentration in one direction.

And at some clubs the director assigns seats rather than letting the players choose their seats, and they'll (hopefully) balance the field as they do this.

At my regular club the players seat themselves, but the players have learned to balance the field themselves. The good pairs tend to sit opposite each other (we don't worry too much about B vs. C, as we don't even know where the dividing line will be until we finish entering names into the computer). But we rarely have games large enough to use a Mitchell movement, so it's not that important.

[pescetom]

 barmar, on 2019-September-16, 09:08, said:

It tends to be needed only in tiny games, like 7 or 8 tables. With fewer tables you'll probably run a Howell movement, so there's only one field, and with more tables you're less likely to have a concentration in one direction.

And at some clubs the director assigns seats rather than letting the players choose their seats, and they'll (hopefully) balance the field as they do this.

At my regular club the players seat themselves, but the players have learned to balance the field themselves. The good pairs tend to sit opposite each other (we don't worry too much about B vs. C, as we don't even know where the dividing line will be until we finish entering names into the computer). But we rarely have games large enough to use a Mitchell movement, so it's not that important.

We almost always run a Mitchell movement and unfortunately the players make little attempt to balance the field themselves. The stronger players do their best to sit NS and the weaker players are not distressed to sit EW where it is easier to place top of line.

I still think you fail to to understand the cultural difference underlying the ABC grading issue. How many of your club voted communist in the last ten years or think that universities have no right to set admission criteria? In my club a majority.

[blackshoe]

The last time the Communist Party of the USA fielded a presidential candidate was in 1984 (Angela Davis). I haven't seen them on any ballot this century.

[pran]

 pescetom, on 2019-September-18, 14:10, said:

We almost always run a Mitchell movement and unfortunately the players make little attempt to balance the field themselves. The stronger players do their best to sit NS and the weaker players are not distressed to sit EW where it is easier to place top of line.

This is OK provided you have no arrow-switching and entirely separate score results for NS and EW.
(Any attempt to merge Mitchell results into one common list by arrow-switching or similar is deemed to be unbalanced.)

[pescetom]

 pran, on 2019-September-19, 01:20, said:

This is OK provided you have no arrow-switching and entirely separate score results for NS and EW.
(Any attempt to merge Mitchell results into one common list by arrow-switching or similar is deemed to be unbalanced.)

It's not OK because the local classification usually gets merged into a national classification and our national level placings will be skewed in a way that is unfair to pairs in other clubs.

[barmar]

You do understand that this classification is basically only used for determining who wins ACBL masterpoints, right? And ensuring a good distribution generally means that more players win masterpoints in the game. It's purely a marketing gimmick.

[pran]

 pran, on 2019-September-19, 01:20, said:

This is OK provided you have no arrow-switching and entirely separate score results for NS and EW.
(Any attempt to merge Mitchell results into one common list by arrow-switching or similar is deemed to be unbalanced.)

 pescetom, on 2019-September-19, 02:49, said:

It's not OK because the local classification usually gets merged into a national classification and our national level placings will be skewed in a way that is unfair to pairs in other clubs.

NW and EW should play in different groups with individual scoring for each group.

I didn't emphasize the obvious fact that if two or more groups at an event are of significantly different strength they must of course be ranked accordingly (A-level and B-level or whatever nomination you use).

My main point was that there is no way a Mitchell movement can be fairly balanced across the entire field of contestants however much arrow-switching is attempted. The resulting balance will always be mediocre or worse.

[pescetom]

 barmar, on 2019-September-19, 09:42, said:

You do understand that this classification is basically only used for determining who wins ACBL masterpoints, right? And ensuring a good distribution generally means that more players win masterpoints in the game. It's purely a marketing gimmick.

I'm not sure what you mean by that. Are you saying that if we do not balance the local field and as a result one of our NS takes first place in the national classification (thanks to the many tops obtained playing always against weaker pairs) I should not worry because there are only masterpoints at stake? If so, I find it excessively cynical. Not that many of our players care about the masterpoints, but they all care about their position in classification. And people in other clubs have a right to a fair playing field.

[pescetom]

 pran, on 2019-September-19, 09:43, said:

NW and EW should play in different groups with individual scoring for each group.

I didn't emphasize the obvious fact that if two or more groups at an event are of significantly different strength they must of course be ranked accordingly (A-level and B-level or whatever nomination you use).

That's the only thing I was insisting on as not OK, that we are not currently ranking or any equivalent measure to distribute skill.

I agree with all you are saying.

[barmar]

 pescetom, on 2019-September-19, 09:55, said:

I'm not sure what you mean by that. Are you saying that if we do not balance the local field and as a result one of our NS takes first place in the national classification

I'm not sure what "national classification" you're talking about. I'm talking about ordinary club games.

[pran]

 barmar, on 2019-September-19, 10:01, said:

I'm not sure what "national classification" you're talking about. I'm talking about ordinary club games.

FWIW:
Our national regulation for Norwegian masterpoints requires clubs which run more than one "group" of contestants at an event and the groups are of different strength, to apply different scales according to the strength of a group for awarding/reporting masterpoints.