What do you need to know about what a bid means?

[Brenda Egeland]

When defining a bidding sequence in a convention, what do you need to know (or might need to know) about the bid?

1. The sequence of bids that preceded it, e.g., in Stayman, 2C preceded by 1NT - P is Stayman.

2. Whether the opening bid in the sequences was 1st, 2nd, 3rd, 4th seat

3. HCP: min and/or max

4. Shape: should this just be text, or can this be standardized without having to provide a zillion possibilities?

5. Announceable, and announce text

6. Alertable, and alert text

7. Vulnerability conditions

8. Forcing: - 1 round, to game

9. Additional text detail

[Adam Parrish]

On Tue, Nov 10, 2020 at 11:28 PM Brenda Egeland <bre...@rsvpbridge.com> wrote:

When defining a bidding sequence in a convention, what do you need to know (or might need to know) about the bid?

1. The sequence of bids that preceded it, e.g., in Stayman, 2C preceded by 1NT - P is Stayman.

It might be useful to also define the State of the auction. eg, game forcing, major-suit fit, etc. Or the Type of auction: 1NT opening, NMF situation, Opponents double, etc. So rather than have to spell out every possible preliminary sequence, we say, "This is a Fourth-Suit-Forcing situation" 

2. Whether the opening bid in the sequences was 1st, 2nd, 3rd, 4th seat

A subsection here might be "Is partner a passed hand." Many bids are virtually identical in 1/2 or 3/4th seat and partner's PH status is what's relevant. 

3. HCP: min and/or max

This one is tricky. Is it a range? How much do we allow for judgment and/or extra considerations. For example, I would say my range for an opening 1-bid is 12-24, but I open many 11-counts with a 5+ card suit and 10s with extra shape based partly on the rule of 20. At the other end, there are 23-counts I open with a 1-bid and 19-counts I open 2C.  

4. Shape: should this just be text, or can this be standardized without having to provide a zillion possibilities?

Some categories might help, like balanced, single suited, two-suited, etc. Possibly filling in exact pattern as shape becomes known. For example, say I open 1NT and then show five hearts. My shape might be shown as ?5??  Or I open 1S and after Jacoby 2NT show diamond shortness. 5+?0-1?   

5. Announceable, and announce text

I would lump this in with alert 

6. Alertable, and alert text

This is jurisdiction dependent. You might need fields for which jurisdictions require the alert. For example, Jacoby 2NT. You have the definition text for an alert, then say it is alertable in ACBL, USBF, Australia, Germany, but don't include areas where it's not, like England (I'm just making this up, I have no idea about the various alert procedures. But I know what's alertable here isn't necessarily alertable there and vice versa.) 

7. Vulnerability conditions

8. Forcing: - 1 round, to game

9. Additional text detail

--
You received this message because you are subscribed to the Google Groups "Conventions Index subGroup" group.
To unsubscribe from this group and stop receiving emails from it, send an email to conventions-in...@bswg.org.
To view this discussion on the web, visit https://groups.google.com/a/bswg.org/d/msgid/conventions-index/c47cbfc7-df35-4d05-b025-58e65099db54n%40bswg.org.

[Thorvald Aagaard]

I think that when we define a convention we should not include any bidding up to the activation of the convention. That is a hard statement, but what I am trying to say is that the information derived from the bidding is needed, not the bidding as we will not try to parse the sequence :-)

Just think about this bidding

4D - (P) - 4N = RKCB-1430

but 4D can be preempt in D or Namyats (Good 4S) and that is not part of the convention

So the definition of RKCB-1430 starts at 4N, and the responses are defined based on that. I know we can use RKCB-1430 with 4 different trump suits, and ideally it is the same convention, but I could also see it as 4 different conventions (As you might need different bids to ask for trump queen)

But RKC-1430 is not always started in 4N but can also be started when suit agreement, so we could either define that as a new convention or include that in the input for the convention.

So a convention is in fact a small application and that can take a lot of input, like seat, vulnerability, offensive or defensive, defined range (So the convention can use min/max), suit length, voids,  etc.

The current bidding could be needed for input but it is the derived information, that is needed

And for me we have Blackwood, RKCB-0314, RKCB-1430, MinorWood, LackWood, VoidWood and many more.

One other issue with conventions is that we as players often make small changes because we like that better, so perhaps we should have a way to Copy a convention to our system (Mark as modified)

Currently I have defined more that 100 sequences that starts RKCB-1430, but I cant make it generic as sometimes we play 4N Quantitative. But all this is part of my system and not the convention.

Best regards

Thorvald

--
You received this message because you are subscribed to the Google Groups "Conventions Index subGroup" group.
To unsubscribe from this group and stop receiving emails from it, send an email to conventions-in...@bswg.org.
To view this discussion on the web, visit https://groups.google.com/a/bswg.org/d/msgid/conventions-index/c47cbfc7-df35-4d05-b025-58e65099db54n%40bswg.org.

Thorvald Aagaard
Krydservej 33
4040 Jyllinge
Mobil : 22 99 55 25
http://www.netbridge.dk
http://www.thorvald.dk

[Kiat Huang]

That's a pretty comprehensive list.

On Wed, 11 Nov 2020 at 04:28, Brenda Egeland <bre...@rsvpbridge.com> wrote:

When defining a bidding sequence in a convention, what do you need to know (or might need to know) about the bid?

Although not part of the convention itself, a proposed index (later ratified by a body) that provides a 1-1 correspondence to the precise definition of that convention, e.g. AB023

I think at a fundamental level we need to agree that each convention has an index that identifies it.  It will be useful as a shortcut in System Notes and Convention Cards to select the index which then enables a lookup to populate the required fields.

1. The sequence of bids that preceded it, e.g., in Stayman, 2C preceded by 1NT - P is Stayman.

2. Whether the opening bid in the sequences was 1st, 2nd, 3rd, 4th seat

3. HCP: min and/or max

4. Shape: should this just be text, or can this be standardized without having to provide a zillion possibilities?

5. Announceable, and announce text

6. Alertable, and alert text

7. Vulnerability conditions

8. Forcing: - 1 round, to game

9. Additional text detail

[Thorvald Aagaard]

Ok. What about starting with a definition of Blackwood. Lets agree on the definition of the convention, then we can look at how we want do formalize the description

From:  https://en.wikipedia.org/wiki/Blackwood_convention

So how should our definition of the convention be?

Do we include the 5N / 6 bid?

I suggest we start with a Blackwood, where there are no trump agreed.

This is just a sample of how the definition could be (Not all bids yet included)

Thorvald

To view this discussion on the web, visit https://groups.google.com/a/bswg.org/d/msgid/conventions-index/CACnBAZfzbvEMOxh0R08gY1bMibiUN7vHHiLhgTzuYqpvjC0y4w%40mail.gmail.com.

[ais523]

On Wed, 2020-11-11 at 18:29 +0100, 'Thorvald Aagaard' via Conventions

Index subGroup wrote:
> Ok. What about starting with a definition of Blackwood. Lets agree on
> the definition of the convention, then we can look at how we want do
> formalize the description
>
> From: https://en.wikipedia.org/wiki/Blackwood_convention
>

> So how should our definition of the convention be?

I think Blackwood is a good example of how defining a convention can be
difficult!

If we're talking about simple four-Ace Blackwood (probably the easiest
variant to define), the bid that triggers Blackwood is always 4NT.
However, 4NT isn't always Blackwood. For example, most partnerships
play 1NT, 4NT as natural (and even if they play 1NT, 4NT as Blackwood,
then a followup 5NT has to be natural rather than a King ask). So a lot
of thought needs to go into defining the *context* of the convention:
with what bidding sequences is a given bid Blackwood?

We also need to consider what the convention shows. What hand types use
Blackwood? It's not like it shows a particular HCP range or a
particular length in a suit. In most cases, you have to define it in
more qualitative terms, e.g. "we use this bid when we want to force
slam but are worried that we might be missing two Aces".

For any convention that has coded replies, there are also potential
lateral-thinking uses of the convention. Two examples with Blackwood:

1. A player has all four Aces, and wants to play 5C. However, they're
worried that 5C would be interpreted as invitational or as artificial
(maybe the opponents opened a "could be short" club suit), and thinks
that 6C won't make. The solution is to use Blackwood (forcing a 5C
response), and then passing the response.

2. A player wants to force slam, but isn't sure whether there's a heart
fit, and wants to play 6H if there is or 6NT if there isn't. However,
notrumps need to be played from their side due to a directional stopper, and nobody has bid notrumps yet. The solution is to bid Blackwood as an artificial forcing notrump bid, so that if partner subsequently corrects 6H to 6NT, the previous 4NT bid ends up right-siding the hand. (I took this example from Jeff Goldsmith's website, <http://www.jeff-goldsmith.org/html/useful.html>.)

So we need to decide whether these possible uses are explained in a
format that captures the essence of a convention. I don't think it's
possible to list all the possibilities of this nature, because there
are probably plentry that we haven't thought of.

Instead, I strongly suspect that the best approach for asking bids like
Blackwood is to define them by what the replies are (and by what can be
shown subsequently in the bidding). That makes it possible to deduce
whether the bid would be worth using or not. (For example: say someone
uses 4NT as four-Ace Blackwood when the agreed suit is clubs; they
probably have at least two Aces, because they will be forced into slam
if partner has only one Ace. This isn't really part of what Blackwood
"shows", it's simply an inference based on what the replies are.)

Bridge calls are something of a mix of "telling" and "asking". In some
systems, such as relay systems, these are neatly separated, and they
act differently enough from each other that it would almost make sense
to use different formats for each. However, there are also systems
where it's much more mixed; for example, I've designed systems where
bids could be either telling or asking and it isn't clarified until
future calls are made (they start off as asking, so that the possible
responses are known, but later on in the bidding specific inferences
can be drawn from the choice of asking bid, so a particular asking bid
sequence can be used to show something). Another good example is the
control bid (e.g. in 1H, 3H; 4C, many players play 4C as showing
control of clubs and denying control of spades); here, the 4C is saying
something about the hand ("I have club control, I don't have diamond
control, I have a slam try"), but is also, in effect, asking something
("Do you have a spade control and/or a diamond control?").

====

I'm not sure whether it's better to start a system for defining
conventions with the easiest cases, or the hardest cases (things like
Kickback are definitely very hard to define); focusing on the latter
will at least help to prevent us from backing ourselves into a corner.

I'm also not even sure what the easiest cases are. My first thought
about easy cases (among conventions that are generally known) were
specific Stayman variants, but that can lead to a lot of complexity
about what the followups were.

My second thought was McKendrick (1NT, 2S! initially asking whether
opener is maximum or not, but maybe planning to play 3 of a minor), but
then I realised that many players are unaware of the potential 1NT,
2S!; 3C!, 6NT sequence, and I was wondering whether this was part of
the convention or just a lateral-thinking addendum. (When I looked up a
definition of McKendrick online
<https://www.bridgewebs.com/monmouth/The%20Mckendrick%20Convention.pdf>
it turned out that this particular sequence is listed, so at least one
person has decided that lateral-thinking sequences should be listed as
part of a convention definition. However, they didn't list the very
similar 1NT, 2S!, 3C!, 7NT sequence, and therefore are missing a hand
type that might end up using the convention.)

All this probably implies, though, that we need to get a better
understanding of what the problem is before trying to solve it.

--
ais523

[Brenda Egeland]

Adam, I think something about state of the auction could be useful. What are some examples where it wouldn't be known from the bidding sequence? For example, a 1NT opening or NMF would be part of the sequence.

I've made the convention library in Bridgodex public so I can show examples. Here are some to look at: (better urls will come eventually...)

RK1430: https://bridgodex.com/convention/20

Stayman: https://bridgodex.com/convention/6

Jacoby Transfers: https://bridgodex.com/convention/7

New Minor Forcing: https://bridgodex.com/convention/76

Mind you ... these are examples of how this could all work. I'm not a bridge expert so the examples above are not meant to be definitive!

[Brenda Egeland]

I'm in favor of showing complete bidding sequences as much as we can. I think it will be clearer to a user, and clearer to any developer trying to interpret the convention for robots, auto-alerts, etc.

It could be that a bid has more than one potential meaning, so a sequence could exist more than once in someone's card. For example, 1c - x could mean shortness in clubs and an opening hand, or it could mean a big 17+pt hand.

It could also be that someone makes a bid that isn't defined on their card.

[ais523]

On Wed, 2020-11-11 at 12:29 -0800, Brenda Egeland wrote:
> I'm in favor of showing complete bidding sequences as much as we can. I
> think it will be clearer to a user, and clearer to any developer trying to
> interpret the convention for robots, auto-alerts, etc.
>
> It could be that a bid has more than one potential meaning, so a sequence
> could exist more than once in someone's card. For example, 1c - x could
> mean shortness in clubs and an opening hand, or it could mean a big 17+pt
> hand.
>
> It could also be that someone makes a bid that isn't defined on their card.

One of the complexities here is that you can't show a complete bidding
sequence simply by listing the bids. Compare:

(1C), P, (1D), 1H

(1C), P, (1D[A]), 1H
[A]: Shows hearts

I (and most likely many other pairs) play the first of these sequences
as natural with hearts, and the second sequence as at least 4 spades
and 3 cards in each minor. But the bidding up to this point is
identical.

As a consequence, I think the context information for a convention has
to be based on what the bids so far *mean*, not on what the bids so far
*are*. A good example of this is that many partnerships will have a
structure for "responding to Stayman", so what you care about is not
really the bidding so far, but whether partner's call is defined as
Stayman or not. In many cases, this structure will be adapted according
to what level the bidding has reached, rather than being fixed at a
particular point.

Here's another example. When working on a system of my own, I tried to
make it as objective as possible whether any given convention was "on"
or not. For one convention, the rule to decide whether it was active
was "the most recent bid by either side was at the 2 or 3 level, we
have not agreed a major suit, and partner's most recent call neither
asked for a stopper nor denied a stopper" (the system had an explicit
method of agreeing major suits, so it was always clear whether a major
suit had been agreed or not). Trying to convert something like that
into a list of complete bidding sequences is almost impossible
(especially because the list of bids that ask for stoppers depends on
which suits the opponents have shown, and they might well have shown
them artificially).

--
ais523

[Brenda Egeland]

If it weren't for interference, this would all be quite easy!

The two sequences you listed have a subtle difference, one has the 1D natural, and one has the 1D alerted. We could capture that in a sequence, say 1c - p - 1d - 1h vs 1c - p - 1d! - 1h. But there could be a whole slew of meanings for that alert. 

Is the problem just what the opponent's bid means? We can certainly capture the meanings of our own bids in sequences. So in the given sequence, we need something like 1c - p - (shows 4+hearts) - 1h

[Brenda Egeland]

Another question...

In the sequence (1c) - p - (1d!) - 1h, where the 1d is showing 4 hearts, is our 1h response a convention, or a defense? Do we need to differentiate?

So perhaps if we have defined a convention for the meaning of the alerted 1d bid, we would then have defenses for that convention, which would include our 1h call?

[Thorvald Aagaard]

I have the following definitions

• Undisturbed bidding (We open and they pass all the time)
  
• Disturbed bidding (We open, and they try to disturb us by bidding)
  
• Defensive bidding (They open the bidding)

We could add a fourth

Competitive bidding

but that is a subset of Disturbed and defensive bidding, depending on strength shown

Thorvald

On 12-11-2020 14:48, Brenda Egeland wrote:

Another question...

In the sequence (1c) - p - (1d!) - 1h, where the 1d is showing 4 hearts, is our 1h response a convention, or a defense? Do we need to differentiate?

So perhaps if we have defined a convention for the meaning of the alerted 1d bid, we would then have defenses for that convention, which would include our 1h call?

Thorvald Aagaard

[richard willey]

Here is an attempt to characterize a variety of different types of opening bids.  Some of these dimensions are orthogonal to one another.  (Note that I don’t have a category for bids that promise strength, in ALL opening bids have some definition around strength even if its “absolutely anything”)

 

Take it for what it’s worth

 

First:  Bids that have a known “anchor” suit (they promise specific length in a known suit) versus bids that don’t promise a known anchor suit

 

Example: 

 

·      A one spade opening playing 5 card majors (typically) promises 5+ Spades

·      A one heart opening playing MOSCITO promises 4+ Spades

·      A multi 2D opening does not have an anchor suit

·      A strong artificial and forcing 2C opening does not have a known anchor suit

 

This can be further broken down into cases in which the anchor suit is the suit opened (The 1S example) and cases in which the anchor suit is not the suit opened.

 

Second:  Bids that promise balanced or semi balanced hands

 

Example

 

·      “Natural” 1N openings

·      “Natural” 2NT openings

 

Third:  Bids that Promise Shortness

 

Examples:

 

·      A mini Roman 2D opening

·      Most of the opening bids playing “DELTA”

 

Fourth:  Bids that promise length in two or three (known suits)

 

Examples:

 

·      Ekrens 2D (promises length in Hearts and Spades)

·      Flannary 2H (promises length in Hearts and Spades)

·      Precision 2D (Promises length in Hearts, Spades and Clubs)

 

Fifth:  Hands that incorporate deliberate randomization

 

Sixth:  “Chimera”  Bids that are best described as some combination of other categories

 

·      A multi 2D that is either

o   A weak hands with 6+ spades

o   A weak hand with 6+ hearts

o   A strong balanced hand

o   A strong three suited hand

·      A Polish club opening that is either

o   A balanced hand with 12 – 14 HCP

o   A single suited hand with 6+ clubs and 15 – 17 HCP

o   An unbalanced hand with 4+ Clubs and shortness and 12+ HCPs

o   A strong hand with 18+ HCP