Ruling in Pula - the rest of the story

[Herman De Wael]

I'd say that if one out of seven chose an alternative, it's clearly
within the limit of being an LA.
And it seems like you didn't ask the polled players if they considered
any other alternative than the chosen.
You might have found that several of the 4S bidders considered passing,
and that it perhaps was a close decision for some of them.

How close the decision is, does not matter for the sake of L16B1b. And I
don't need a poll to rule that a significant proportion would consider
passing.
What I'm concerned about is when we should "judge some might select it".
Considering those words at face value, we should rule against even if no
one in the poll does select it, because "some might".

If only 1/10 chose an alternative, and none of the others considered it,
it's below the threshold for an LA. IMO.

1/10? Do you really want to be polling 10 players for each case? Do you
know how difficult it is to find 10 players to poll?
And surely everyone considers alternatives. The word in the laws is
"seriously consider" it. What does that mean. Don't you want your
pollees to seriously consider each alternative?

Herman.

[Harald Berre Skjæran]

If I and all other polled thought 4S was absolutely clear, none of us would
consider pass at all. And pass wouldn't be an LA at all.

If I and a few other polled thought this was a somewhat close decision, say
70-30, but all of us, except one who passed, chose 4S, we all seriously
considered pass. And pass has to be an LA.

Thus, I'd say it's not enough to poll some number of players what they
would bid. You also have to ask them if they had any serious alternatives,
and how close the decision was for them. If not, I'd say the poll result is
not good enough to base a ruling upon.

Post by Herman De Wael

I'd say that if one out of seven chose an alternative, it's clearly
within the limit of being an LA.
And it seems like you didn't ask the polled players if they considered
any other alternative than the chosen.
You might have found that several of the 4S bidders considered passing,
and that it perhaps was a close decision for some of them.

How close the decision is, does not matter for the sake of L16B1b. And I
don't need a poll to rule that a significant proportion would consider
passing.
What I'm concerned about is when we should "judge some might select it".
Considering those words at face value, we should rule against even if no
one in the poll does select it, because "some might".

If only 1/10 chose an alternative, and none of the others considered it,
it's below the threshold for an LA. IMO.

1/10? Do you really want to be polling 10 players for each case? Do you
know how difficult it is to find 10 players to poll?
And surely everyone considers alternatives. The word in the laws is
"seriously consider" it. What does that mean. Don't you want your
pollees to seriously consider each alternative?
Herman.
_______________________________________________
Blml mailing list
http://lists.rtflb.org/mailman/listinfo/blml

[Eric Landau]

If I and all other polled thought 4S was absolutely clear, none of us would consider pass at all. And pass wouldn't be an LA at all.
If I and a few other polled thought this was a somewhat close decision, say 70-30, but all of us, except one who passed, chose 4S, we all seriously considered pass. And pass has to be an LA.
Thus, I'd say it's not enough to poll some number of players what they would bid. You also have to ask them if they had any serious alternatives, and how close the decision was for them. If not, I'd say the poll result is not good enough to base a ruling upon.

Well put.

I fear the term "poll" is misleading. "Poll" suggests determining a "winner" by one vote per pollee, like an election. Which is why we give results like "6 of 7 bid this, 1 of 7 bid that", which tells us essentially nothing. "What would you bid?" is the wrong question; it should be "What might you bid?" or "What would you consider bidding?" An action is a logical alternative if a significant number of "peers" would seriously consider it, even if it would get no "first choice" votes. We should think of what we're doing as "consulting" rather than "polling".


Eric Landau
Silver Spring MD
New York NY

[Wayne Burrows]

I'd say that if one out of seven chose an alternative, it's clearly within
the limit of being an LA.
And it seems like you didn't ask the polled players if they considered any
other alternative than the chosen.
You might have found that several of the 4S bidders considered passing,
and that it perhaps was a close decision for some of them.
If only 1/10 chose an alternative, and none of the others considered it,
it's below the threshold for an LA. IMO.

Certainly the way the laws are now written the definition of logical
alternative has a very low threshold for the number of players that would
choose an action.

There are also significant problems in polling and basing decisions on
polls with very small samples. The margin of error is great. One in only
seven could in reality mean a much lower or considerably higher proportion.
And some polls are conducted on significantly smaller samples.

Herman's point that it depends on partner is of course valid. I contend
often that in an established partnership especially one is likely to know
well, at least in the long run, what partner is most likely to be thinking
about in any given situation.

Here I don't have any general problem with the ruling as the poll indicates
4S, for a random player I think the hesitation is more likely to be an
awkward rather than a strong hand, perhaps the weaker the player the more
likely they would be to have values, and the poll concurs with my thoughts.
With a small sample it is always reasonable to weigh the poll against your
own thoughts if there is a difference to see if there is reason for the
discrepancy. That is the poll should only be a guide.

Of course in the actual situation we are again dealing with a small sample,
1. So this could be the time that a strong player hesitates with values or
whatever other anomaly might turn up. This makes rulings very hard to be
fair and to be seen to be fair.

You have all discussed this ruling at length.
Indeed, most of you have guessed correctly, the 3Sp bid did not come as
quickly as could be.
3Sp bidder had KJ106 J7 9875 653.
And therein lies a first problem. What did you expect partner could
possibly have as a maximum? Surely we must allow a player to know where
the maximum his partner can have, would lie? Not so as to be sure of
what hesitating partner has, but so as to be able to average the
probabilities. Someone calculated the probability of making against a
near yarborough at near 50%. Surely the probabilities go up with the
number of points partner can hold. If partner can hold as much as 5
points, where the probability of making 4Sp will be 90% or so, then the
average probability reaches 70%, more than enough to make the game
biddable.
And yet no-one (not in my poll, not in another one, not on blml) asks
about the aggressivity of partner.
So let's continue the actual story.
Although all the boards in Pula were pre-dealt and duplicated, this
actual board was dealt at the table. Someone (in fact the possible
offenders on this hand, but that is not important) had sat at a wrong
table, and started playing board 9. When the were transported to the
correct table, this left two matches (four tabled) without a board 9. So
I hand-dealt a replacement, and this was it
For completeness, the full hand
N: KJ106 J7 9875 653.
W: Q74 K10965 106 1094
E: 93 AQ832 KQ32 J8
S: A852 4 AJ4 AKQ72
Since no one apart from 16 players had seen the board, I could do a
large poll. I used three of the directors (all Polish), two players
(both Czech) and two friends (Belgians). The result of my poll: 6-1.
(space left blank for creation of tension)
6-1 in favour of 4Sp.
The sole outcast was the ageing Belgian, who would have bid 4Sp if
vulnerable but not when not vulnerable.
Considering that the actual pair were young Poles I decided to allow the
4Sp bid.
The opponent decided to appeal. He told be he had asked three players,
all of whom passed.
Between this time and the appeal, I asked by partner (4Sp) and Slawek
Latala texted his group of top Poles (who bid 4Sp at 6-1 also).
The Appeal Committee consisted of 3 players. One of them stated he would
also bid 4Sp, and the others did not object. But the Committee did its
job and asked for the definition of LA to be read.
And then they decided, albeit by 2-1, to consider Pass a LA and to
return the score to 3Sp+1.
Which was the reason for my question. If a poll produces a result of
6-1, should that not count as a minority too small to make it a LA?
Should we not be allowed to use judgment as to who we consider peers. We
ask someone, and he's in a minority of one. He is totally unlike the
player in age and bridge tradition (French <> Polish, you cannot be more
different).
If the polling system is going to work, we need a treshold below which
we might accept a minority opinion not to be a LA.
Otherwise, we can just keep on asking and we'll be certain to find
someone. That sounds too much like 'if it hesitates, shoot it', to me.
Comments?
Herman.
_______________________________________________
Blml mailing list
http://lists.rtflb.org/mailman/listinfo/blml

--
Kind regards,
Harald Berre SkjÊran
_______________________________________________
Blml mailing list
http://lists.rtflb.org/mailman/listinfo/blml

[Steve Willner]

Post by Wayne Burrows
Certainly the way the laws are now written the definition of logical
alternative has a very low threshold for the number of players that
would choose an action.
There are also significant problems in polling and basing decisions on
polls with very small samples. The margin of error is great.

I agree with both of Wayne's points. The way I usually explain "LA" is
that you don't have to bid hearts when you really have spades, and you
don't have to pass a forcing bid, but anything that's at all reasonable
in context is a LA. We don't have percentages any more, but it probably
corresponds to an action that 5-10% of equivalent players would actually
choose.

Reliably estimating 5-10% from a poll would require hundreds of pollees.
That's obviously impractical. The purpose of polling is not to reach a
numerical decision but rather to make sure the TD hasn't overlooked some
line of reasoning for or against some action.

Post by Wayne Burrows
not only the number of players who made the alternative bid,
but also the number of those who considered it, should be taken into
account. Here, among us, it is "most". Whence pass is a LA.

This is another reason polls are tricky. As Konrad wrote, one needs to
be sure the pollees are peers of the players concerned. On the surface,
a 6-1 vote tends to suggest both options are LAs. Even 7-0 might do so
if many of the 7 are considering the other option.

In this case, I have no doubt that both pass and 4S are LAs.

Post by Wayne Burrows
But that's not, not at all, the key to the problem. The key is that no
LA has been suggested strongly enough by the tempo.

I agree that "what is suggested?" is the key to the problem, but I don't
think it's so clear that neither alternative is suggested. As I wrote
earlier, I thought the tempo suggested passing, but the actual hand at
the table was suggesting 4S. Maybe this is a case where one has to know
one's partner. I'd want to ask the player who bid 4S why he bid it
before ruling, but I have a lot of sympathy for the player at the other
table who passed because he thought partner's tempo suggested bidding on.

[David]

The problem that several of us are having with this ruling is not
whether 3S is an LA

rather

is 4S DEMONSTRABLY suggested over Pass.

We believe the converse to be true.




**********************************
***@lineone.net
**********************************



-----Original Message-----
From: Herman De Wael
Sent: Tuesday, September 16, 2014 9:24 AM
To: Bridge Laws Mailing List
Subject: Re: [BLML] Ruling in Pula - the rest of the story

I'd say that if one out of seven chose an alternative, it's clearly
within the limit of being an LA.
And it seems like you didn't ask the polled players if they considered
any other alternative than the chosen.
You might have found that several of the 4S bidders considered passing,
and that it perhaps was a close decision for some of them.

How close the decision is, does not matter for the sake of L16B1b. And I
don't need a poll to rule that a significant proportion would consider
passing.
What I'm concerned about is when we should "judge some might select it".
Considering those words at face value, we should rule against even if no
one in the poll does select it, because "some might".

If only 1/10 chose an alternative, and none of the others considered it,
it's below the threshold for an LA. IMO.

1/10? Do you really want to be polling 10 players for each case? Do you
know how difficult it is to find 10 players to poll?
And surely everyone considers alternatives. The word in the laws is
"seriously consider" it. What does that mean. Don't you want your
pollees to seriously consider each alternative?

Herman.

[Nigel Guthrie]

{Herman]
You have all discussed this ruling at length. Indeed, most of you have
guessed correctly, the 3Sp bid did not come as quickly as could be. 3Sp
bidder had KJ106 J7 9875 653.
And therein lies a first problem. What did you expect partner could possibly
have as a maximum? Surely we must allow a player to know where the maximum
his partner can have, would lie? Not so as to be sure of what hesitating
partner has, but so as to be able to average the probabilities. Someone
calculated the probability of making against a near yarborough at near 50%.
Surely the probabilities go up with the number of points partner can hold.
If partner can hold as much as 5 points, where the probability of making 4Sp
will be 90% or so, then the average probability reaches 70%, more than
enough to make the game biddable. And yet no-one (not in my poll, not in
another one, not on blml) asks about the aggressivity of partner. So let's
continue the actual story.

[Nige1]
Pollees must rely on the director to present the full facts, including any
matters of style, of which he is aware.

[Herman]
[SNIP] Since no one apart from 16 players had seen the board, I could do a
large poll. I used three of the directors (all Polish), two players (both
Czech) and two friends (Belgians). The result of my poll: 6-1.
(space left blank for creation of tension)
6-1 in favour of 4Sp. The sole outcast was the ageing Belgian, who would
have bid 4Sp if vulnerable but not when not vulnerable. Considering that the
actual pair were young Poles I decided to allow the 4Sp bid. The opponent
decided to appeal. He told be he had asked three players, all of whom
passed. Between this time and the appeal, I asked by partner (4Sp) and
Slawek Latala texted his group of top Poles (who bid 4Sp at 6-1 also).
The Appeal Committee consisted of 3 players. One of them stated he would
also bid 4Sp, and the others did not object. But the Committee did its job
and asked for the definition of LA to be read. And then they decided, albeit
by 2-1, to consider Pass a LA and to return the score to 3Sp+1. Which was
the reason for my question. If a poll produces a result of 6-1, should that
not count as a minority too small to make it a LA? Should we not be allowed
to use judgment as to who we consider peers. We ask someone, and he's in a
minority of one. He is totally unlike the player in age and bridge tradition
(French <> Polish, you cannot be more different). If the polling system is
going to work, we need a treshold below which we might accept a minority
opinion not to be a LA. Otherwise, we can just keep on asking and we'll be
certain to find someone. That sounds too much like 'if it hesitates, shoot
it', to me. Comments?

{Nige2]
When an extensive poll of the player's peers reaches an overwhelming
conclusion, I agree with Herman that should be decisive. Anyway, an appeals
committee should be reluctant to second-guess the director on matters of
judgement, especially when no new facts or arguments emerge.

[Alain Gottcheiner]

Which was the reason for my question. If a poll produces a result of
6-1, should that not count as a minority too small to make it a LA?

AG : IIRC, not only the number of players who made the alternative bid,
but also the number of those who considered it, should be taken into
account. Here, among us, it is "most". >Whence pass is a LA.

But that's not, not at all, the key to the problem. The key is that no
LA has been suggested strongly enough by the tempo. Whence neither 4S
nor pass can be disallowed. (not even 4C IMO)

Best regards


Alain

[Konrad Ciborowski]

And I don't need a poll to rule that a significant proportion would
consider
passing.

Beware of this approach, Herman.
As they say in Hollywood "nobody knows anything".

First of all don't forget that the LA depends on the class of player
involved.

I've run myself or watched taken several polls in my life.
The results were astonishing to me at times.
You would never believe what bids poeple could consider or not consider
making.
I remember a poll where everyone in my circles could think of no other
call than pass (there was no other option even considered) & yet when people
with sklills similar to the players involved were polled they
would all bid 5C and 6C (total brain cancer to me)
and not even contemplate passing.

So for the class of players involved passing wasn't an LA at all.

Always make a poll. Don't be lazy.

Best regards,
Konrad Ciborowski
Kraków, Poland

[Herman De Wael]

Post by Konrad Ciborowski

And I don't need a poll to rule that a significant proportion would
consider
passing.

Beware of this approach, Herman.
As they say in Hollywood "nobody knows anything".
First of all don't forget that the LA depends on the class of player
involved.
I've run myself or watched taken several polls in my life.
The results were astonishing to me at times.
You would never believe what bids poeple could consider or not consider
making.
I remember a poll where everyone in my circles could think of no other
call than pass (there was no other option even considered) & yet when people
with sklills similar to the players involved were polled they
would all bid 5C and 6C (total brain cancer to me)
and not even contemplate passing.
So for the class of players involved passing wasn't an LA at all.
Always make a poll. Don't be lazy.

Well, if you give the hand, and the reply does not come in 2 seconds,
that means they have considered it, no?
So take it for read that many polled players considered passing. That
part of the definition was OK.

Herman.

Post by Konrad Ciborowski
Best regards,
Konrad Ciborowski
Kraków, Poland

[Olivix]

Hello,

May bey ou should have asked two questions (it's what i do usually)
1) for the director
What do yhou consider bidding?
2) for the players
What do you bid?

The second one is of no use for ruling, but it helps a lot to have a fair
answer to 1) witch is the way to know if "pass" is a LA
Sometimes (...) everybody consider passing or bidding (or balance between
two bids) and after everybody do the same thing,
Since 2007, that means that the 0% vote-bid IS a LA

If hou have, say,
1) 40% balancing between "pass" & "4S"
2) 100% biding "4S"
So "pass" is a LA and you rule for 170

Best regards,
Hope to see you in a championship, i am now available to direct more often
:)
Olivier,

-----Message d'origine-----
De : blml-***@rtflb.org [mailto:blml-***@rtflb.org] De la part de
Herman De Wael
Envoyé : mardi 16 septembre 2014 09:46
À : blml
Objet : [BLML] Ruling in Pula - the rest of the story

You have all discussed this ruling at length.

Indeed, most of you have guessed correctly, the 3Sp bid did not come as
quickly as could be.

3Sp bidder had KJ106 J7 9875 653.

And therein lies a first problem. What did you expect partner could possibly
have as a maximum? Surely we must allow a player to know where the maximum
his partner can have, would lie? Not so as to be sure of what hesitating
partner has, but so as to be able to average the probabilities. Someone
calculated the probability of making against a near yarborough at near 50%.
Surely the probabilities go up with the number of points partner can hold.
If partner can hold as much as 5 points, where the probability of making 4Sp
will be 90% or so, then the average probability reaches 70%, more than
enough to make the game biddable.

And yet no-one (not in my poll, not in another one, not on blml) asks about
the aggressivity of partner.

So let's continue the actual story.

Although all the boards in Pula were pre-dealt and duplicated, this actual
board was dealt at the table. Someone (in fact the possible offenders on
this hand, but that is not important) had sat at a wrong table, and started
playing board 9. When the were transported to the correct table, this left
two matches (four tabled) without a board 9. So I hand-dealt a replacement,
and this was it For completeness, the full hand
N: KJ106 J7 9875 653.
W: Q74 K10965 106 1094
E: 93 AQ832 KQ32 J8
S: A852 4 AJ4 AKQ72

Since no one apart from 16 players had seen the board, I could do a large
poll. I used three of the directors (all Polish), two players (both Czech)
and two friends (Belgians). The result of my poll: 6-1.



(space left blank for creation of tension)






6-1 in favour of 4Sp.
The sole outcast was the ageing Belgian, who would have bid 4Sp if
vulnerable but not when not vulnerable.
Considering that the actual pair were young Poles I decided to allow the 4Sp
bid.

The opponent decided to appeal. He told be he had asked three players, all
of whom passed.

Between this time and the appeal, I asked by partner (4Sp) and Slawek Latala
texted his group of top Poles (who bid 4Sp at 6-1 also).

The Appeal Committee consisted of 3 players. One of them stated he would
also bid 4Sp, and the others did not object. But the Committee did its job
and asked for the definition of LA to be read.

And then they decided, albeit by 2-1, to consider Pass a LA and to return
the score to 3Sp+1.

Which was the reason for my question. If a poll produces a result of 6-1,
should that not count as a minority too small to make it a LA?
Should we not be allowed to use judgment as to who we consider peers. We ask
someone, and he's in a minority of one. He is totally unlike the player in
age and bridge tradition (French <> Polish, you cannot be more different).

If the polling system is going to work, we need a treshold below which we
might accept a minority opinion not to be a LA.

Otherwise, we can just keep on asking and we'll be certain to find someone.
That sounds too much like 'if it hesitates, shoot it', to me.

Comments?

Herman.