[Vampyr]

Trinidad, on 2013-February-28, 04:57, said:

You don't have 6 spades. Your spade length is about 4.5.

I think that the "6" referred to the actual number of cards, not the number of tricks the suit can be expected to produce.

[gordontd]

Vampyr, on 2013-February-28, 09:26, said:

When I was first given this hand, I didn't look too deeply into it. I felt that I had told my story, and passed. I might well do this at the table too, though after reading this thread, I believe that it would have been the wrong thing to do.

I would not do this, however, if I were given the hand in a poll or as a member of an AC. I would give the matter more thought. It may be the case that polls and analysis by AC members does not accurately reflect what would have happened at the table.

This is a problem with polling - which is not to suggest that we should not do it. Players and AC members usually know the outcome by the time they consider it, and that's one reason why it can be a good start to ask TDs instead, if they are peers of the players in question, because they haven't already played the hand.

Best of all is when there are two events going on using different boards; then we can get players without too many preconceptions. However, it will always be the case that those polled are aware of a reason behind the question, and some find it hard to disregard that. Personally I don't find it hard - I can usually answer the question without needing to guess the back-story, but I know lots of others can't seem to do that. Sometimes though they will guess wrong - they'll think it's a UI case when it's an MI case for example, and directors can encourage that by asking questions about earlier parts of the auction than those they are interested in.

[lalldonn]

rhm, on 2013-February-28, 09:11, said:

I do not consider a yarborough 3=4=3=3 support and if this is all my partner needs I expect him to bid the game himself, in particular when the opponents bidding have indciated the hand is no misfit.
I know it is easy to contract for 12 tricks with a yarborough when looking at all 4 hands.
I would contract for 4 tricks on defense.

Rainer Herrmann

Then to answer your ATB question, 100% to RHM

There is a reason my last post got four quick upvotes, and it's not because I said anything clever or funny...

[cherdano]

rhm, on 2013-February-28, 09:11, said:

I do not consider a yarborough 3=4=3=3 support and if this is all my partner needs I expect him to bid the game himself, in particular when the opponents bidding have indciated the hand is no misfit.

Pro tip: don't defend 4H doubled when you have a double-fit including spades.

[jallerton]

lamford, on 2013-February-28, 09:02, said:

The arguments on this thread convince me that double is better than 4S. However, both will lead to partner correcting to the cold 5♦ as he has four-card diamond support. There was a BIT by South over 4♥ and the TD consulted several people and decided Pass was an LA and awarded an adjusted score of +420/+450 for EW. The AC overruled the poll and were of the opinion that North would always bid 4S. The L&E then criticised the AC for overruling the TD, as they felt greater weight should have been given to the poll that was conducted. I disagree with this view, as no appeal in UI cases would ever succeed if the AC always woodenly accepted the views of those polled. It was their duty to decide whether Pass was an LA for this particular North, and the poll is only a guide.

Wrong. The AC's job is to review a TD's ruling, starting with the presumption that the TD's ruling is correct. Where a poll has been conducted by a TD, the AC should check the validity of the poll: did the TD ask the right question(s), was the right peer group of people polled, how many people were asked and what were the answers? If the poll is valid, the AC should not ignore its results.

[lesh]

lamford, on 2013-February-27, 11:05, said:

---

West's 2C was precision style. What do you do now, and what other bids do you seriously consider?

I will bid either DBL and spades or 4♦ Leaping michales. If I have bidded like that I will be 4♠. Even if I go down in 4♠ I am pretty sure they would have a game. It could be a double swing as well or go completely wrong.

[FrancesHinden]

lamford, on 2013-February-28, 09:02, said:

The arguments on this thread convince me that double is better than 4S. However, both will lead to partner correcting to the cold 5♦ as he has four-card diamond support. There was a BIT by South over 4♥ and the TD consulted several people and decided Pass was an LA and awarded an adjusted score of +420/+450 for EW. The AC overruled the poll and were of the opinion that North would always bid 4S. The L&E then criticised the AC for overruling the TD, as they felt greater weight should have been given to the poll that was conducted. I disagree with this view, as no appeal in UI cases would ever succeed if the AC always woodenly accepted the views of those polled. It was their duty to decide whether Pass was an LA for this particular North, and the poll is only a guide.The full hand is on page 6 of http://www.ebu.co.uk...Jan%20draft.pdf for those that are interested.

the L&E felt greater weight should have been given to the poll. I don't see that as saying the AC must "woodenly accept" the results of a poll.

jallerton, on 2013-February-28, 12:42, said:

Wrong. The AC's job is to review a TD's ruling, starting with the presumption that the TD's ruling is correct. Where a poll has been conducted by a TD, the AC should check the validity of the poll: did the TD ask the right question(s), was the right peer group of people polled, how many people were asked and what were the answers? If the poll is valid, the AC should not ignore its results.

Other reasons the AC might change a UI ruling made with the aid of a poll, in addition to those mentioned by Jallerton:
- they learn more about the methods of the pair in question that the TD hasn't discovered, making the poll invalid
- they might disagree about which calls are demonstrably suggested by the UI
- they might agree with the results of the poll and a player's LAs, but might still think that the subsequent auction would have been different to that ruled by the TD
- the basis of the appeal might be that there is no UI, or that the play would have gone differently in the ruled substitute contract

[PeterAlan]

jallerton, on 2013-February-28, 12:42, said:

Wrong. The AC's job is to review a TD's ruling, starting with the presumption that the TD's ruling is correct. Where a poll has been conducted by a TD, the AC should check the validity of the poll: did the TD ask the right question(s), was the right peer group of people polled, how many people were asked and what were the answers? If the poll is valid, the AC should not ignore its results.

FrancesHinden, on 2013-March-01, 01:36, said:

the L&E felt greater weight should have been given to the poll. I don't see that as saying the AC must "woodenly accept" the results of a poll.

L&E Draft Minutes linked to by lamford said:

TD: ... My discussion/polling among my colleagues established that pass was a logical alternative to North’s 4♠ bid

This wording would suggest to me that the poll in this case was limited to 2 or 3 TD colleagues consulted. Was the extent of the poll established?

[lamford]

jallerton, on 2013-February-28, 12:42, said:

"was the right peer group of people polled"

If the AC decided that they would not pass, and they decided that the particular player (I have no idea who he or she was) was more a peer of themselves than of those polled, then they would correctly ignore the result of the poll.

[PhilKing]

I think the AC dropped the ball on this one. Whether or not they all believe 4♠ is obvious is almost certainly irrelevant (unless they were 2♠ bidders). How can it possibly be obvious to someone who overcalled only 2♠?

Now I would double 4♥, which also leads to 5♦, but I would not have overcalled 2♠, so my view of what to do now is immaterial.

[Trinidad]

I read the original writeup from the link. I think we have to be very careful with the difference between a poll and a consultation.

In a consultation, you freely discuss the question whether a certain action would be an LA. You do this with a group of experienced players or TDs (and you should do it with the whole group together). You will get arguments, discussion leading to new arguments and at the end the arguments are weighted to reach some form of concensus: "We judge that action so and so is / isn't an LA."

In a poll, you don't discuss whether the action was an LA. You investigate it. You go around and ask peers of the player two questions: "What actions do you seriously consider?" and "What action do you chose?". No discussion, you just write down what the peer replied to the questions. If you get a sufficiently large amount of replies from the player's peers you will be able to judge on a statistical basis whether the action was an LA. The problem with polls is that the "sufficiently large amount" can be quite a large amount and that may be unpractical.

In this case, there was no poll. A few TDs were consulted. They discussed the matter. (I hope they did this together, but I suspect that these were a few 1 on 1 discussions.) The conclusion of the discussion was that pass was an LA.

There was an appeal. The AC discussed the matter again. In their discussion they reached the conclusion that pass was not an LA. Now, you may disagree with their conclusion (I do) but I really don't see any reason why the discussion between the TDs should be given any weight at all (other than that it was a decision with some foundation which requires good arguments to be overturned). It was not a poll and an AC simply outranks the TD in judgement decisions.

_________

I did find something inconsistent in the write up draft. The draft describes first that the TDs in consultation decided that pass would be an LA ("My discussion/polling among my colleagues established that pass was a logical alternative"). Then the L&EC writes in their comment that several people would pass at the table ("The TD had carried out a consultation and determined that several people not only felt that pass was a logical alternative, but that they would actually pass at the table.".

The L&EC version is significantly different from the TD's contribution.

Rik

[Trinidad]

PhilKing, on 2013-March-05, 07:13, said:

I think the AC dropped the ball on this one. Whether or not they all believe 4♠ is obvious is almost certainly irrelevant (unless they were 2♠ bidders). How can it possibly be obvious to someone who overcalled only 2♠?

Now I would double 4♥, which also leads to 5♦, but I would not have overcalled 2♠, so my view of what to do now is immaterial.

If you don't have specialized methods to show twosuiters, it is certainly legitimate (as well as quite common) to show these hands by bidding one of the suits hoping/expecting that you will get a chance to show the second one. The alternative would be to make a takeout double first, then force to game and then show your two suits. By that time you may be bidding at the six level.

In this case, it is reasonable to expect to get a second bid.

Rik

[gordontd]

Trinidad, on 2013-March-05, 09:28, said:

I read the original writeup from the link. I think we have to be very careful with the difference between a poll and a consultation.

In a consultation, you freely discuss the question whether a certain action would be an LA. You do this with a group of experienced players or TDs (and you should do it with the whole group together). You will get arguments, discussion leading to new arguments and at the end the arguments are weighted to reach some form of concensus: "We judge that action so and so is / isn't an LA."

In a poll, you don't discuss whether the action was an LA. You investigate it. You go around and ask peers of the player two questions: "What actions do you seriously consider?" and "What action do you chose?". No discussion, you just write down what the peer replied to the questions. If you get a sufficiently large amount of replies from the player's peers you will be able to judge on a statistical basis whether the action was an LA. The problem with polls is that the "sufficiently large amount" can be quite a large amount and that may be unpractical.

In this case, there was no poll.

You make some unwarranted assumptions.

There is no reason why TDs can't be polled if they are peers of the player, and I've commented before that they can often be very good people to poll because they haven't played the hand. Having polled them, there is no reason not to then show them the whole hand and have a discussion with them about it. Hence they can be polled and also be part of a discussion about the hand.

That's what happened in this case and that's why I worded it like that. It happened at the end of the evening when there were few tables still in play and by the time the North player (who was very disruptive and difficult) had allowed me to get on with the job of making a ruling, there were no players around I could ask, but there were some TDs who were peers of the player, in the sense that they were about the same standard as him (which I think I am too) and were prepared to accept the auction as given.

Sometimes the "sufficiently large amount" can be quite small: if it seems clear to me as a TD that pass is a logical alternative, and I poll two people, who are peers of the player, and they both say that they would pass, that is enough to establish that pass is a logical alternative. My memory of what happened in this case is that it was something like that, though there may have been more of us by the time the discussion happened.

Many more polls & consultations have taken place since then, and they often start with showing a colleague a single hand and giving him/her an auction. As soon as there's any doubt or difference of opinion, then we're off to find players, sometimes to poll, sometimes to consult with. It does depend on the nature of the case which questions should be asked or what discussions instituted.

[Trinidad]

gordontd, on 2013-March-05, 10:55, said:

There is no reason why TDs can't be polled if they are peers of the player, and I've commented before that they can often be very good people to poll because they haven't played the hand. Having polled them, there is no reason not to then show them the whole hand and have a discussion with them about it. Hence they can be polled and also be part of a discussion about the hand.

That may seem correct, but it is not pure and it introduces bias in your poll. Below it will be clear why.

gordontd, on 2013-March-05, 10:55, said:

That's what happened in this case and that's why I worded it like that.

I had not understood it that way from the write up, but that is probably my reading.

gordontd, on 2013-March-05, 10:55, said:

Sometimes the "sufficiently large amount" can be quite small: if it seems clear to me as a TD that pass is a logical alternative, and I poll two people, who are peers of the player, and they both say that they would pass, that is enough to establish that pass is a logical alternative.

Clear, and it seems true. But... this is not the full story.

What would have happened if the two first TDs would have said that they would double or bid 4♠? You would have gone on to find more and more peers. They weren't there and because of that, you will not be able to finish the poll. You will discard it, thinking something like "at least I tried" and you will move on to base your decision on the discussion with your colleagues. This is why I wrote above that you are introducing bias in your poll: If the result of the poll tends towards pass being an LA, you will use the poll. If the result of the poll tends towards pass not being an LA, you will discard it because you will not be able to get a good poll. To exaggerate a little: 100% of the successful polls will state that pass is an LA, even if it isn't.

To prevent this bias you will have to make sure that you will be able to finish the poll whatever the result will be. Therefore, if you will not be able to conduct a poll to get both possible answers, you should not start it.

As far as I can see (but you can see that better), this is a case where the poll should not have started, since it was impossible to obtain a result of "pass is not an LA". That means that effectively there was no valid poll, only a discussion between TDs.

And if there only was a discussion between TDs, there is no reason why a discussion between AC members should not be able to overrule the result of the TD discussion.

Rik