[BudH]

7 table club game, 7x4 Mitchell. A player has a medical issue and cannot play the last two boards.

You award Average Plus - Average Minus on those last two boards. Hopefully everyone agrees with this.

But what if the medical issue happens much earlier forcing the pair to withdraw from the session? You don’t want to be giving Average Plus scores to a huge number of pairs. What do you do if the player is forced to miss:

1. One round (4 of 28 boards)
2. Three rounds (12 of 28 boards)
3. Five rounds (20 of 28 boards)

Feel free to emphasize the difference between “no play” and “average”, if you’d like.

[hrothgar]

BudH, on 2018-April-23, 16:41, said:

7 table club game, 7x4 Mitchell. A player has a medical issue and cannot play the last two boards.

You award Average Plus - Average Minus on those last two boards. Hopefully everyone agrees with this.

But what if the medical issue happens much earlier forcing the pair to withdraw from the session? You don’t want to be giving Average Plus scores to a huge number of pairs. What do you do if the player is forced to miss:

1. One round (4 of 28 boards)
2. Three rounds (12 of 28 boards)
3. Five rounds (20 of 28 boards)

Feel free to emphasize the difference between “no play” and “average”, if you’d like.

I have never understood the habit of awarding average plus boards in situations like this.

Personally, I think that there are two reasonable ways to go:

1. These boards don't count for ANYONE (Even those who played them)

2. Pairs who played the boards in question receive their scores / Pairs who did not play the boards have their score calculated with these treated as missing.

If the boards were played a small number of times, the former seems preferable.
If the boards were played a large number of times, the later seems preferable.

The trick of course is determining where the dividing line between large and small should be...

I think that the real issue here is whether or not some pairs might be artificially advantaged / disadvantaged because it is much easier to score tops / bottoms with small numbers of players competing on a board. In turn, this suggests that the dividing line should be determined dynamically after comparing the variance in the scores on the hands in question with the scores of the boards for the rest of the event.

If the variance is close, chose the latter. If it is far away go with the former.

[BudH]

The thing is, there should be official instructions or interpretations to follow when something like this happens ........

[hrothgar]

BudH, on 2018-April-23, 19:42, said:

The thing is, there should be official instructions or interpretations to follow when something like this happens ........

No doubt.

However, you are talking to the wrong audience to get that changed.

[blackshoe]

There should be, but there aren't.

Technically, the law says that if a board cannot be played at all, an artificial adjusted score should be given. Clearly the pair who is still there ready to play is in no way at fault for the board not being played, so they get average plus. The other pair, well, the options are "partly at fault" and "directly at fault". It may seem harsh, but in law they're directly at fault. So they would get average minus.

That's fine when you're only talking about a few boards. If you're talking about a lot of boards, that's not so good. Looking at the other end, if it's one or two rounds you can hopefully figure out a way to amend the movement, treating the missing pair as never having been there. If that's not possible, well, perhaps Hrothgar's option 2, although frankly I don't think that's legal.

[gordontd]

In the EBU we have a regulation that when a pair plays fewer than half the boards, all those boards are scrapped for them and their opponents. So that would be our solution for your last case. We also have a regulation for Swiss Pairs events, where matches are typically 6+ boards long, that you only get AV+ for half the boards missed and AV for the remainder. This wouldn't apply here, but is something to think about for rounds that are as long as four boards.

[BudH]

blackshoe, on 2018-April-23, 21:43, said:

There should be, but there aren't.

Technically, the law says that if a board cannot be played at all, an artificial adjusted score should be given. Clearly the pair who is still there ready to play is in no way at fault for the board not being played, so they get average plus. The other pair, well, the options are "partly at fault" and "directly at fault". It may seem harsh, but in law they're directly at fault. So they would get average minus.

That's fine when you're only talking about a few boards. If you're talking about a lot of boards, that's not so good. Looking at the other end, if it's one or two rounds you can hopefully figure out a way to amend the movement, treating the missing pair as never having been there. If that's not possible, well, perhaps Hrothgar's option 2, although frankly I don't think that's legal.

I gave a hypothetical case. In real life, it was 7 1/2 tables and my most hated movement, the ACBL H8ROVER external was being used with a Pair 15 rover. Stationary pair at Table 4 had the medical issue and had already been bumped by the roving pair for two boards early in the session. 9 boards out of 28 were missed.

[hrothgar]

BudH, on 2018-April-24, 05:59, said:

I gave a hypothetical case. In real life, it was 7 1/2 tables and my most hated movement, the ACBL H8ROVER external was being used with a Pair 15 rover. Stationary pair at Table 4 had the medical issue and had already been bumped by the roving pair for two boards early in the session. 9 boards out of 28 were missed.

What is the variance of the the scores on the 9 boards?

What is the variance of the scores on the remaining boards?

[barmar]

hrothgar, on 2018-April-23, 17:19, said:

If the boards were played a small number of times, the former seems preferable.
If the boards were played a large number of times, the later seems preferable.

Not sure what you mean by this. All the affected boards will be played N-1 times, where N is the original number of times they were scheduled to be played.

The difference between withdrawing early or late is in how many boards are affected, not how many times the boards get played.

There's another solution you might be able to use. If your club uses BBO robots to fill in a half-table movement, the bots could take the place of the pair that takes ill.

[hrothgar]

barmar, on 2018-April-24, 08:37, said:

Not sure what you mean by this. All the affected boards will be played N-1 times, where N is the original number of times they were scheduled to be played.

The difference between withdrawing early or late is in how many boards are affected, not how many times the boards get played.

D'oh

thanks. (Haven't been sleeping well recently and I fear it is starting to show)