ACBL Sections

[Dealing_Don]

Can anyone explain the reasoning for scoring ACBL tournaments across all players and then dividing up into sections for awarding points?

 

This results in some sections averaging 40% and others 60%. I saw one tournament where one B player had a 41% and came in second in their section while another had 59% and came in 4th in another section.

 

It would seem that either 1) scores should only be compared within a section resulting in a 50% average for N/S and E/W within the section or 2) scores should be compared across all teams and the tournament should be scored as one section.

 

It is different in land tournaments, where you are actually competing against different players and so sections make sense. But here, you only play three to six rounds, so there is really not a "section as such, so a different form of scoring would seem to make sense.

 

Comments?

[uday]

I believe that this is how "real tourneys" are scored in acbl land as well.

 

Perhaps the theory is that your field (your direction) may be tougher than the field in other sections. The same is probably true for the opps? Maybe it is about the opps more than your direction.

[TylerE]

No. Practically all (As in, 99%+ including sectional and regional events) are NOT matchpointed across sections.

[Dealing_Don]

If our opponents are tougher, than while it is true that we end up with a likely lower score than our opps, the reality is that we only played 4 or 5 of those opps (maybe not even any of the "tougher" ones).

 

Since with a short format, you cannot play the majority of opponents in your section and the assignment of opps is random, why not just score across all sections and assign points as if in one big section?

 

The reason this is not done in ACBL land is twofold:

 

1) The sections and number of boards are usually devised so that you play the majority of opps in your section. This makes it fair as you and those playing your direction have the same quality of opponent.

 

2) The number of duplicate boards is fixed and to have a section of 60 would not work logistically

 

I am sure that number 1 reason is the primary driver for ACBL tourneys.

 

But on BBO, playing only a few of the opps in your directions makes it almost meaningless to have a section unless the sections consisted of only the number of pairs that you have rounds, but that would limit you to one winner per secction, so it comes back to the only real fair way is to matchpoint and assign points across all sections and all pairs.

[hotShot]

I don't know about the ACBL, but if we have different groups playing the same boards (Howell-Type movement or scrambled Mitchel) and all scores are used, they are scored in as one group.

[barmar]

No. Practically all (As in, 99%+ including sectional and regional events) are NOT matchpointed across sections.

And when they are matchpointed across sections, the places and awards are based on the combined sections.

[mycroft]

By and large they are matchpointed across sections (which happens "all the time" in events in my district) unless the sponsor says otherwise.

 

The reason the sponsor says otherwise is because they think their players are whiners or can't understand the scoring (and by and large, in my experience, they underestimate the ability of their players to learn). Well, not really, but "why is my 1100 only a 22 when their 1100 is a 25? They're both the best results I can see" is a common complaint, as is "But I got put in a hard section, so my results are worse than that section over there where they got an easy draw" - well, that's why we run two-session events (although it's strange how the people who say that never seem to get put in the "easy" draw. Must be those evil directors seeding the event just to annoy them).

 

There are studies that show that the difference between "totally across the field" and "across sets of 3 sections, doing remainder one as 2 and 2" is washable, and people can understand top of 25 or 38, but after that it tends to warp their minds, so that's what we do (except in events like the second day of the Blue ribbons or the like). Add to that the amount of work a fouled/misdealt board/scoring error is in RL bridge, which is magnified by the number of sections that have to be rebuilt, this also weights in favour of the 3 and 3 over across the field.

 

Of course with online games, where the boards can't get fouled, and the scoring is automatic, and the board results are in percentages rather than MPs, it's easier just to score across the field.

 

Unless you have a "all-play-all", there are always going to be unfairnesses, and there will always be someone who is or feels hard done by by the system. Even with all-play-all, it's not fair if you hit the precision/relay pair that can find 7C trivially on board 21 in an IMP pair game, rather than the two flat games on 23 and 24, isn't it?

[mycroft]

Having said that, mostly due to the feeling of unfairness, when the ACBL scores across sections, section awards are still scored within the section, rather than within "E/W in A,B,and C".

[Dealing_Don]

Interesting discussion about R/L bridge and players perception of fairness.

 

So it begs the question, what is the downside to matchpointing and awarding points across all pairs in a 50 pair event?

[barmar]

Interesting discussion about R/L bridge and players perception of fairness.

 

So it begs the question, what is the downside to matchpointing and awarding points across all pairs in a 50 pair event?

When you matchpoint within a section, the pairs your scores are being compared with have mostly played against the same set of opponents. If there's a beginner pair sitting the opposite direction, giving out gifts, just everyone you're competing with (or almost everyone, if it's not a compete movement) has a chance to play against them and get the gift. Pairs in other sections don't have this opportunity, so it seems unfair to rank the sections together.

 

It's harder to justify not matchpointing across sections, though. Before computers, the excuse was probably that matchpointing on a 25 or 38 top was difficult for most directors (some of the legendary top directors of yore were able to do it flawlessly, though). In RL tournaments, it adds a little more time to getting the results out, because they have to wait until the results for all the sections are entered, but this is obviously not an issue for online tourneys.

[TimG]

Interesting discussion about R/L bridge and players perception of fairness.

 

So it begs the question, what is the downside to matchpointing and awarding points across all pairs in a 50 pair event?

The "downside" is that fewer masterpoints are awarded, or fewer pairs win masterpoints, I forget which. But, the way this tends to be done in ACBL events is in large part due to masterpoint awards.

[Dealing_Don]

Uday, can you shed some light on what the effect on distribution of masterpoints would be with awarding points as a single section versus the current method of awarding by multiple sections?

[uday]

The ACBL formula is complicated and not symmetrical. I dont know what we mean by "the current method of awarding by multiple sections".

 

I believe I score across all sections, award within a section. But stratification and overalls and movement also affect this.

[ArtK78]

Having said that, mostly due to the feeling of unfairness, when the ACBL scores across sections, section awards are still scored within the section, rather than within "E/W in A,B,and C".

This is not true.

 

When ACBL events are scored over several sections, or over the entire field, the session awards are scored over the several sections, or over the entire field, in one direction.

 

So, for example, in the qualifying session of a major pair event, the pairs may be scored across groups of 3 sections (A/B/C and D/E/F and G/H/I). If there are thirteen rounds, then each board will be played 39 times within a "triple-section," resulting in a 38 top. All of the North-South pairs in A/B/C will be scored against each other as one group for purposes of the session awards, as will all of the NS pairs in D/E/F and G/H/I, as well as all of the East-West pairs in each of the "triple-sections."

 

Overall qualifying is also done within a triple section. So, for example, if there are 16 tables in each section, each triple section will have 48 tables. If 50% of the field qualifies for the second day of the event, the top 24 pairs in the triple section A/B/C North-South will qualify, as will the top 24 pairs in the triple section D/E/F and so on. This sometimes results in a non-qualifier in one triple-section having a higher score than a qualifier in another triple-section.

 

In one-session pair games at Sectional tournaments, which tend to be no more than 3 sections, it is not uncommon for the entire event to be scored across the field. In that case, the session awards are for all of the North-South pairs and for all of the East-West pairs, with the overall awards being given to the top pairs regardless of the direction that they were sitting.

 

There is a significant benefit on BBO to breaking up the game into sections rather than awarding masterpoints across the field. By breaking the game up into sections, there are multiple 1st place awards, multiple 2nd place awards, and so on. These awards are far larger than the awards that would be given out if the field were ranked from top to bottom (as was done on e-bridge). For example, suppose there are 100 pairs in a BBO ACBL pair game. The filed will be broken down into 3 or 4 sections - assume 4 sections of 12, 12, 13 and 13 tables. Each of the eight first place pairs - North-South and East-West - will get the award for first place in their respective sections. That award would be 0.96 or 1.04 or 1.12 - I don't recall what the exact award is for first in a 12 or 13 table section. The maximum section award is 1.20. The second place pair gets 75% of the award for 1st place, and the third place pair gets 75% of the award for second place, and so on, down to the maximum number of places that are given masterpoint awards (40% of the field, I believe). The section award is equal to 75% of the award for the immediately higher position, or the reciprocal of the order of finish times the first place award (sixth place gets 75% of the fifth place award or 1/6 of the first place award, whichever is higher). In addition, there are overall awards - the pair that is first overall gets 2.00, the pair that is second overall gets 1.50, the pair that is third overall gets the higher of their section award or 1.13. I don't know if there is a fourth place overall award, as it would be 0.85, and the section award is likely to be higher. Each pair is entitled to the greater of the award for their overall finish or for their finish in their section.

 

If one were to rank the field from top to bottom, and award 2.00 to the first place pair, the awards would quickly shrink into miniscule amounts. For example, the award for 10th place in the event would be 0.20, and the award for 20th place in the event would be 0.10. As you can see by looking at the top 12 finishers in any BBO ACBL pair game, the actual awards for the top 12 finishers in the event are much higher than that. The reason is that the section awards that are given out by breaking up the event into smaller sections are much higher than the corresponding awards if the event were ranked across the field.

[TimG]

When ACBL events are scored over several sections, or over the entire field, the session awards are scored over the several sections, or over the entire field, in one direction.

I don't believe that is always the case. I remember this issue being raised in New England and being told by the directors that there was a difference in total masterpoints awarded (and/or number of pairs who win masterpoints) depending upon which method was used. This may have been 10 years ago, so maybe it has since changed and is done the same throughout ACBL, but I doubt it.

[ArtK78]

In sectional and higher rated events, it is my understanding that the number of pairs that receive session awards used to be 30% (truncated). So, in a 15 table section, there will be 4 session awards in each direction. That number may have been increased. I seem to recall more places being awarded in pair games.

 

In club games (and the ACBL pair games on BBO are considered club games) the number of pairs that receive session awards is 40% (rounded up). So, in a 15 table section, each of the North-South and East-West pairs in the top 6 places will receive session awards.

 

To the best of my knowledge, the same scheme is used for determining the number of pairs that receive session awards when the scoring is done over multiple sections. If there are 45 pairs in a triple-section (3 sections of 15 tables each) and those three sections are scored together, the session awards are awarded to the top 13 pairs in the triple-section. Of course, the awards for the last 4 or 5 places are negligible.

 

The method of awarding masterpoints is revised from time to time. As of January 1, 2008, the award scheme for knockout team events was changed. It used to be true that 2nd place in a KO was worth 75% of 1st place. Now, it is worth about 50%. But the award for first place was increased. Not only is the award for first place based on the total number of teams in the event, but also for the "quality of the field" based on total masterpoints. I was personally responsible for increasing the masterpoint award at the last local sectional KO event by getting Dave Treadwell (25,000 masterpoints) as my partner. It was also useful that we won the event, as it was worth over 21 points.

 

The old method created a peculiar result of the semi final match of many small KO events being worth more masterpoints than the final match. For example, in a typical bracketed KO event held at a sectional tournament, the field will be divided into 3 brackets of about 6 or 7 teams each (no more than 8 - it is a one day event). Surviving the first round gets you to the round of 4 - the semi-finals. You get match awards for surviving the first round. In the top bracket at a sectional tournament, the first place award prior to 1/1/08 was about 13 points, and the second place award was about 9.75 points (75% of first). So, if your match award was 1.20 for surviving the first round, you were playing for 8.55 points in your semi-final match. If you won your semi-final match, you were playing for an additional 3.25 points in the finals (and all of the glory that goes with winning the KO).

 

Now, with the 2nd place award equal to about 50% of the award for 1st place, winning the final match is worth more than winning the semi-final match (but not by a lot). Of course, with the increased point awards, both the semi-final match and the final match are worth more than any single match was worth before the change in the award structure.

[A2003]

If the scores are entered in ACBL program to find out the section winners and overall winners, will it produce the same winners and placements?

Not sure.

I picked up one result and one section:

I tried to do manually for one section and assign round numbers.

example:

11 teams : 10 for top.

1 for each team lower score, 1/2 pt for tie, 0 for each team higher score.

Results did not come with the same percentage as announced in BBO.

Again, It is complex, tiring, and fractions make some difference may be for multiple sections.

 

For only one section, results comes out ok.

 

In regular ACBL tournaments:

For 8 teams in one section:

The scores are announced as 7 for top and so on.

Since, it is not announced as 38 or 45 for top, and so, it might not be compared against all teams. Just my guess.

[mycroft]

If you see on the recap top of 9,8,12... it's scored in section. Better hope your direction isn't stacked. If the recap says top is 17, 25, 38 - scored across sections. Now you have to hope that your opponents weren't the stacked line.

[tony stack]

If the scores are entered in ACBL program to find out the section winners and overall winners, will it produce the same winners and placements?

Not sure.

I picked up one result and one section:

I tried to do manually for one section and assign round numbers.

example:

11 teams : 10 for top.

1 for each team lower score, 1/2 pt for tie, 0 for each team higher score.

Results did not come with the same percentage as announced in BBO.

Again, It is complex, tiring, and fractions make some difference may be for multiple sections.

 

For only one section, results comes out ok.

 

In regular ACBL tournaments:

For 8 teams in one section:

The scores are announced as 7 for top and so on.

Since, it is not announced as 38 or 45 for top, and so, it might not be compared against all teams. Just my guess.

This is the correct behavior. In ALL bbo tournaments, results for each board are obtained by comparing the result of ALL tables that played the board. So, if you had say 2 sections of 11 tables, it will be scored as a 22 table game, a top of 21, not 2 tops of 10 (10% for beating a pair)). Thus for perfect scoring, would need 21 tables (top 20, or 5% for beating a pair). I would be cool if bbo allowed TD's to specify thiier section sizes, and determine if any of the sections should be combined. As is allowed in ACBL land live tournies/club games :).

[TylerE]

Had an occurrence last night that I think is indicative of some of the things that people get frustrated about. Had a pretty strong game, 59.9%. Finished 7th of 13 in our section. No masterpoints. 12th overall in a game with 4 sections (104 pairs total). A total of 9 A pairs won masterpoints while having lower percentage games. Ten pairs won masterpoints despire being below 50%, 2 of them A pairs. Love to have someone explain the logic of how one A pair can place with a 46.1% game, and another A pair in the same field with a 59.9% does not.

[Dealing_Don]

Tyler,

 

This is a result of the sections that BBO sets up. I argue that sections do not make much sense for BBO tournaments which are typically only rounds. If there are 15 pair in each section, you would only play 27% of those pairs in the entire tournament. Hardly a valid reason to be in such an arbitrary grouping.

 

BUT...and this is a big BUT.... the way ACBL allows masterpoints to be awarded encourages sections--be they fair or not. Under the current rules if the tournament were scored as one large sedtion, there would not be many more awards in that large section as there are in each current section and the size would be approximately the same. So what BBO is doing is using the system to provide players with the maximum points allowed under the ACBL rules.

 

This is an ACBL issue of changing the award "tables" to keep up with technology that makes super large sections possible.