Average Plus and Average Minus

[bluejak]

What I can't understand is why the same algorithm isn't used in the two cases.  In both situations the team has been denied its opportunity to play the match through no fault of its own, and hence I suggest that they be made the same in the final version of the 2010 White Book .

Too late, I fear. While I do have wide latitude in my editing, it does not extend to changing established regulations, especially regulations created before I was on the L&EC! Plus, the timing is poor: the White book is to be finalised in 24 hours and nine minutes, and I am not at home, and when I get home I expect just to have time to look at a long screed form a certain J Allerton - and nothing more!!! ;)

[bluejak]

I hope I don't need to say why any comparison to matchpoint scoring makes no sense.

You do to convince me. I do not care whether it is a full match, part of a match, or a single board: whether it is imps, BAM, MPs, aggregate or anything else: whether it is teams of four, teams of more than four, pairs or an individual. In no case do I feel there is any reason whatever for giving a contestant 100% for any board not played through no fault of their own.

[Mbodell]

I have answered:

- Teams sit the same way

We locally (ACBL northern california) are stricter than your -1 IMP/board for this. If teams sit the wrong way in a swiss or round robin match scored by VP then the VP score is 0 to 0 and neither team gets any VP. This happened in a bracketed round robin sectional tournament recently where two of the best teams in the top bracket played the same direction in the first match which lead to a much tighter final result.

[McBruce]

I have answered:

- Teams sit the same way

We locally (ACBL northern california) are stricter than your -1 IMP/board for this. If teams sit the wrong way in a swiss or round robin match scored by VP then the VP score is 0 to 0 and neither team gets any VP. This happened in a bracketed round robin sectional tournament recently where two of the best teams in the top bracket played the same direction in the first match which lead to a much tighter final result.

"we sat the same way at both tables so we will both get 0vp for this match"

"let's report it as a dead tie"

[jdonn]

I hope I don't need to say why any comparison to matchpoint scoring makes no sense.

You do to convince me. I do not care whether it is a full match, part of a match, or a single board: whether it is imps, BAM, MPs, aggregate or anything else: whether it is teams of four, teams of more than four, pairs or an individual. In no case do I feel there is any reason whatever for giving a contestant 100% for any board not played through no fault of their own.

There are multiple differences.

- Any form of imp scoring is inherently much higher variance than matchpoint scoring, and score adjustments should reflect that a team not at fault as lost a much larger chance at a maximum score (do you think many pairs score 100% of the matchpoints over the course of the first 8 boards? In a round of a swiss it happens all the time of course.)

- Matchpoint scoring makes adjustments in terms of a percentage (not total matchpoints), but swiss team scoring adjusts in terms of a total. The comparison is apples and oranges.

- I feel it is wrong (and inherently misleading) to even consider an adjustment like this in terms of victory points. It's putting the cart before the horse since the adjustment is individual imps for each board. If the proper adjustment for the opponents causing me to be unable to play one board is 3 imps then that is the proper adjustment for any number of such boards and what happens after a conversion to victory points is luck of the draw.

 

Anyway I could probably settle for something like 22 (which is 3 imps a board in an 8 board round). But 18 is almost nothing. It's the equivalent of 1 imp a board + 1. Or one game swing and 7 flat boards.

 

Last point, don't only consider fairness for the team not at fault, what about the team at fault? You want to give them 12 for causing a round to be unable to be played, which is nearly a tied round?

[paulg]

the complement of their opponent's average VP match score

That seems a rather sloppily drafted regulation. Shouldn't it be the average of the complements of the opponent's VP scores?

 

If the opponents have played three matches, scoring 0, 0 and 11, my expectation is

  mean(25,25,19) = 23

but the complement of mean(0,0,11) is 25.

Thanks, you understand what we intended.

 

Already fixed in the next set of regulations.

 

Paul

[jallerton]

What I can't understand is why the same algorithm isn't used in the two cases.  In both situations the team has been denied its opportunity to play the match through no fault of its own, and hence I suggest that they be made the same in the final version of the 2010 White Book .

Too late, I fear. While I do have wide latitude in my editing, it does not extend to changing established regulations, especially regulations created before I was on the L&EC! Plus, the timing is poor: the White book is to be finalised in 24 hours and nine minutes, and I am not at home, and when I get home I expect just to have time to look at a long screed form a certain J Allerton - and nothing more!!! :unsure:

Sorry about the "timing" but I did point out this anomaly two weeks ago to the Editor of The White Book and the Secretary of the L&EC.

 

Notwithstanding the official deadline, I would have thought that, even at the cost of being a day or two late, it more is important to ensure that all known apparent anomalies are investigated and corrected if necessary.

 

As I understand it, the 2010 White Book is not being printed as a book (which is a shame, as it is often a lot easier to read in this format), so there are no deadlines with printers to be met.

[Mbodell]

I have answered:

- Teams sit the same way

We locally (ACBL northern california) are stricter than your -1 IMP/board for this. If teams sit the wrong way in a swiss or round robin match scored by VP then the VP score is 0 to 0 and neither team gets any VP. This happened in a bracketed round robin sectional tournament recently where two of the best teams in the top bracket played the same direction in the first match which lead to a much tighter final result.

"we sat the same way at both tables so we will both get 0vp for this match"

"let's report it as a dead tie"

I think the motivation is to make sure a team can't intentionally clinch an overall win (or a better score than they'd expect) by sitting the wrong way on purpose. If they knew they'd get a few VP, or worse a 10-10 tie, then they could intentionally do it. And as long as at least one of the teams contains an ethical player then you ought not get a false report.

[bluejak]

The EBU automatically gives 0-0, plus a report to the L&EC, to any team that agrees anything, whether a tie or sitting the wrong way. But the vast majority of infractions are based on stupidity, so we do not treat it as malice.

[bluejak]

Sorry about the "timing" but I did point out this anomaly two weeks ago to the Editor of The White Book and the Secretary of the L&EC.

You make it sound as though that should be adequate timing. Not only were approximately 1000 items pointed out to me in the last two weeks, but to change a regulation it needed to be before the last L&EC meeting, not two weeks ago.

 

What did you expect me [or the Secretary] to do? Decide that a decision made in maybe 1980 or so by the L&EC of the day was wrong and change it without authority? And why is it wrong? Because one person says so? Sorry, Jeffrey, I have a high regard for you, but not that high! ;)

[jallerton]

The 2010 White Book was not sent to the proof-readers until after the last L&EC meeting, so it would have been tough for any of them to have pointed out any anomalies before that meeting!

 

I was under the impression that the L&EC members do discuss some business by e-mail, but if that is not the case, then we just have to accept that England is a lot slower at changing anything than some other countries.

 

(gnasher @ Feb 27 2010, 10:46 PM)

QUOTE (cardsharp @ Feb 27 2010, 01:54 PM)

the complement of their opponent's average VP match score

 

 

That seems a rather sloppily drafted regulation. Shouldn't it be the average of the complements of the opponent's VP scores?

 

If the opponents have played three matches, scoring 0, 0 and 11, my expectation is

  mean(25,25,19) = 23

but the complement of mean(0,0,11) is 25. 

 

 

Thanks, you understand what we intended.

 

Already fixed in the next set of regulations.

 

Paul

 

In Scotland, regulations can be updated within 12 hours of someone suggesting an improvement!

[bluejak]

In Scotland, regulations can be updated within 12 hours of someone suggesting an improvement!

Or suggesting something considerably worse, like not alerting doubles because WBF regulations for events they never run say so.

 

I am sorry if you consider a rushed judgement without proper consultation an advantage. I think it an awful approach. I am pleased that the EBU does not go in for such things. The EBU has enough problems without taking away the consultative process just because one person thinks he has a solution to a problem no-one else has seen.