And now for something completely different

[newroad]

Hi All,

 

This is really a slow burn idea for anyone interested.

 

I have for some time been a fan of the idea behind the 1970's shortage-showing Polish methods. It has been a source of both amusement and bewilderment to me that the most successful of the Polish methods of this era, Regres, was at the other end of this spectrum - it in essence had five mini-NT style openings ;)

 

Delta, on the other hand, epitomised the shortness showing idea, but was somewhere between rarely and never used for openings due to its low "mean* opening" (of 1.66). However, we've had 40 years or so to improve things since then, so I have been slowly constructing a method based on this idea: Delta Heart Pass (DHP).

 

The openings are as follows (with shortage defined as 0-1 length in a suit)

 

Pass: ART, 14+ hcp, 15+ hcp if BAL

1♣: 9-14 hcp, BAL or semi-BAL (the latter defined as 5422 or 6m322)

1♦: 9-13 hcp, short in one or both m's

1♥: ART, 0-8 hcp, unsuitable for a PRE

1♠: 9-13 hcp, short ♥

1NT: 9-13 hcp, short ♠, <5♥ unless 0=5=4=4

2♣: 9-13 hcp, short ♠, 5+♥

2♦: 9-13 hcp, 6♥322 or 7♥222

2♥: 9-13 hcp, 6♠322 or 7♠222

2♠: 9-13 hcp, any 2 or 3 shortages (e.g. 6-5-1-1 through to 10-1-1-1 or similar)

2NT: 9-13 hcp, 7m222

3♣+: PRE's to taste

 

If nothing else, I hope it opens the mind to consider the possibility of primary shortness showing - it tends to be immensely valuable in taking a broad (fitting rather than fit) based view of a deal. It also addresses some of Delta's weaker points, by giving a (long) major suit focus to the shortage showing openings as well as both raising the mean opening significantly and addressing competitive concerns (by taking the 6M322's/7222's out of 1♣ and raising the Fert from 1♦ to 1♥).

 

Speaking of 40 years extra advancement, I have a symmetric-style structure to underpin the relay responses, rather than the Polish methods of old.

 

Happy to field any interest.

 

Regards, Newroad

 

* As an aside, I think most historic mean opening calculations have been flawed, as Pass is counted as 0 and 1♣ as 1, whereas they are in fact pretty much equal from a pre-emptive perspective: Pass allows 1♣ whereas 1♣ allows DBL.I would count them both as 1 (or 0).

[dake50]

Wow counter culture.

Now develop the frequency of payoffs tables for showing each singleton/shape.

The double-dummy databases exist. Just need some bridge player/programmer

to code regres, delta, delta heart pass, and others revealed from that charting.

I'd be interested in those chart results.

[newroad]

Hi Dake50,

 

Alas, the problem is evaluation is always going to be very anecdotal - there's not enough real world material to work with (only double-dummy analysis).

 

The best one can do is determine which design principles normally yield good methods and try and optimise them. Even what consitutes these design principles is very anecdotal, but people tend to know that the Balanced Hand Principle is a priori a good idea, and similarly splinters. It is this type of thing that DHP tries to enshrine whilst dealing with the weaknesses (all systems have them).

 

Regards, Newroad

[straube]

My first thought was that I want to show what I have rather than what I don't have, but it's an interesting idea. When you open 1C partner knows you have tolerance for any 5-cd suit, 1D for either 5-cd major . When you open 1S or 1N you preempt the opponents' (presumably) major, let partner know what degree of fit they might have and show tolerance for any other 5-cd suit. Can you tell us more about the continuations?

[awm]

It seems like you could get much better preemption by using 2M as natural 9-13 hcp 6-7 cards no shortage and 2D as weak multi. This only loses the rare "multiple shortage" hands which can probably be put elsewhere.

[newroad]

Hi Straube,

 

That's a large question to answer.

 

Let's take the trivial case - relays after a 1♠[=short hearts] opening (which could/should be viewed as the archetypal UNBAL resolution).

 

1♠ 1NT R

2♣: long high (so in this case, 5+♠)

2♦: Long middle (so in this case, 5+♦, <5♠

2♥: 5+ middle/5+ low or 3/S (so in this case 5+♦/5+♣ or 3/S)

2♠: Long low/4 high (so in this case, 5+♣/4♠)

2NT: Long low (so in this case, 6+♣ single suited)

then resolve for long low/4 middle ...

3♣: 3=1=4=5

3♦: 2=1=4=6

3♥: 3=0=4=6 etc

 

1♠ 1NT

2NT 3♣ R

3♦: 3=1=3=6

3♥: 3=1=2=7

3♠: 2=1=3=7 (this step could be compressed with the one above, if that is your style) etc

 

1♠ 1NT 2♠ 2NT 3♣+ as per 1♠ 1NT 3♣+

 

1♠ 1NT

2♥ 2♠ R

2NT: 5+♦/5+♣

3♣: 4=1=4=4

3♦: 5=0=4=4

3♥: 4=0=5=4 etc

 

1♠ 1NT

2♥ 2♠

2NT 3♣ R

3♦: 2=1=5=5

3♥: 2=0=6=5

3♠+: 2=0=5=6

 

1♠ 1NT 2♦ 2♥ 2♠+ as per 1♠ 1NT 2♠+

 

1♠ 1NT

2♣ 2♦ R

2♥: 5+♠/4♦

2♠: 6+♠

2NT: 5+♠/5+♣

3♣+: 5+♠/4♣ as per 1♠ 1NT 3♣+

 

1♠ 1NT 2♣ 2♦ 2NT 3♣ 3♦+ as per 1♠ 1NT 2♥ 2♠ 2NT 3♣ 3♦+ (i.e. 5/5 resolution)

 

1♠ 1NT 2♣ 2♦ 2♥ 2♠ 2NT+ as per 1♠ 1NT 2♣ 2♦ 2NT+

 

Apologies in advance for the likely typo or two above, but anyone who has a rudimentary knowledge or better of Symmetric Relay should be able to recognise the basic patterns and hence intent.

 

Expanding, the 1NT and 2♣ openings are the equivalent of the 1♠ 1NT 2♦+ and 1♠ 1NT 2♣ sequences.

 

Similarly, the 1♦ opening splits as follows after the 1♥ R

1♠: short clubs, otherwise as per the 1♠ opening

1NT: short both minors (resolution for another time)

2♣+: short diamonds, otherwise as per 1♠ 1NT 2♣+

 

Whet your appetite or turn you off?

 

Regards, Newroad

[newroad]

Hi Adam,

 

It seems like you could get much better preemption by using 2M as natural 9-13 hcp 6-7 cards no shortage and 2D as weak multi. This only loses the rare "multiple shortage" hands which can probably be put elsewhere.

 

Three points

 

• I'm personally not a fan of ANY Multi/Wagner style 2♦ - I regard it as inferior against organised opposition to straight WK/NAT 2's
  
• The 2R openings are sufficiently constructive that (a) the transfer effect is worth using for right-siding purposes, and (b) it allows auto-super accepts with a prime maximum, where opener shows the three card fragment or bids 2NT with the 7222
  
• All six combinations of two-shortage hands are not as rare as you might imagine (do the maths) and even if they were, look at the space you need to unscramble them. Putting them elsewhere either screws up the relays or gives you too little space to explore them. Ideally, I would have liked to switch the 2♠ and 2NT openings around, but this, in effect, would somewhat give up on the two shortage hands. That said, I can see the possibility of doing this after a period of at the table experience.
  

 

Regards, Newroad

[straube]

 

Whet your appetite or turn you off?

 

Not sure. I'm sure the relays are all right. It's what happens when responder has a constructive or GI hand that makes me wonder. Say the bidding starts 1S (short hearts). Responder is well-placed when he has clubs because opener won't bid too high unless he also has clubs. So 1S-1N could be GF relay or hands with clubs. 1S-2C might be hands with diamonds and opener would rebid 2D without great support. 1S-2D hands with hearts? And these preliminary guesses are probably all wrong.

 

Anyway, relay breaks and immediate bids other than the relay would seem to require a lot of optimization and I've really no idea if it would be successful or not. I think someone would have to do a lot of math before we could tell.

 

In fact, you really can't separate how the relays are unwound from what happens when responder decides not to relay. It's all inter-related.

[newroad]

Hi Straube,

 

The intent is to R with GI+ unless there is an alternate system approach that suits better.

 

I'm sure after some time at the table, some optimisations/refinements will suggest themselves. These would then be subject to a cost benefit analysis (memory glitch * cost versus non-optimisations * cost). In reality, it's often better to keep the pattern.

 

Yes, there is some risk of getting too high with a misfit and only GI values. How much of an issue that will be in practice, who knows? In similar more normal symmetric systems, I have rarely seen this be a major issue (and here, arguably, you are better placed, because whilst you don't know about fit, you do know how well you are fitting by virtue of the knowing the shortage at the start).

 

Regards, Newroad

 

PS That said, I don't normally think about optimising the same way you just did (with some apparent merit) - I'll digest it some more and maybe comment on it again in due course