This blog provides supplementary thoughts and ideas to the site. If you haven't seen the main site, there is a lot there including the Martel and Rodwell interviews, photos, and articles. This blog is focused on advancing bridge theory by discussing the application of new ideas. All original content is copyright 2009 Glen Ashton.

Friday, April 27, 2007

The NFL draft is tomorrow so here's a coaching story. Tee Walsh (perhaps a relation of Bill Walsh) had invented the South Coast offence, an integrated running/passing approach that took the run n' gun style to a new level. However when Walsh became head coach of the LA Terminators, he inherited a starting quarterback who was a low percentage passer with a big guaranteed contract, so the South Coast could not work there. So he installed a new Hollywood offence, taking advantage of his world class speed smurf receivers and an offensive line right out of Sumo wrestling school. This got Walsh's team to the Super Bowl, where they played the Toronto Beavers. Now the Toronto team used the South Coast offence, with some northern tweaks, and they proceeded to run up a 30 point lead on Walsh's team. At this point, sideline reporter Avril Lavigne stuck a microphone into Walsh's face and said "how cool aint it that you invented their South Coast offence, eh?"

Sadly, there was no microphone malfunction at this point, so that was the end of sideline interviews in NFL games.

Wednesday, April 25, 2007

Let's go for another design walk to look at the considerations for system designers. Most systems start with a base premise, and here it will be that balanced hands show themselves quickly, with a bid that means only balanced. This seems like a good core concept, but how to deliver it?

So 1NT could be 12-14, and if 18+ balanced we will open on two level, say Mexican 2D for 18-19, 2NT for 20-21, 2C for 22+, so 2C will be like standard, also handling big unbalanced hands.

So where do we park 15-17 balanced? This type of concern is always key for system designers - where to place a set of hands so it has sufficient bidding space available for it, yet it conserves bidding options as much as possible for the rest of the system. We could open it 1S, but now responder plays 1NT if wants to signoff. 1C and 1D could work, but that loses a nice low opening bid. 1H could be just right - responder can pass to sign off in Hs - cool! Responder can bid 1S to relay to 1NT, to play there, or follow-up by showing hand type. Responder can bid 1NT to show 4Ss but signoff value - this might get the contract played by the wrong side, but finds the 4-4 S fit if it is there. The rest of a 1NT structure can be played directly over 1H.

So we have, at the one level:
1H: 15-17 balanced
1NT: 12-14 balanced
1C, 1D, 1S: available, must handle all four suits with less than 2C opening strength and unbalanced.

So we wonder, is this the time for the little major approach, such as:
1C: 4+Hs
1D: 4+Ss
1S: Unbalanced with no 4 card major

What are the awkward and frequent hands now parked in 1S? This is a design question one always needs to ask: are there frequent hands stuck in an opening that will make it hard to find the best spot. Here these frequent and awkward hands would be the 5-4-3-1, having 5-4/4-5 in the minors, and a 3 card major. These need to explore for a 5-3 major fit, but they no longer have the bidding space if they open 1S. And these will come up a lot. So opening these 1S will not work, and thus we need to open unbalanced with a minor by opening either 1C or 1D (or both) to allow for major suit fit investigation. This leaves the 1S opening available, and it might as well show 5Ss: a standard type opening, which we know works. So we have, at the one level:

1H: 15-17 balanced
1NT: 12-14 balanced
1S: 5+Ss, less than 2C opening strength.
1C, 1D: available, must handle unbalanced with Cs, Ds, or Hs

So one of 1C or 1D will do double duty here, handling more than one primary suit. How should we split things up here?

We could do a strength split:
1C: 15+ unbalanced, with Ds, Hs, Cs (and perhaps even Ss, so the 1S opening is limited)
1D: 10-14, unbalanced, not 5+Ss.

Over 1D we seem to have a rebid problem: if it goes 1D-1S--?, opener has to rebid 2H to show Hs, which may be too high without a fit. What about if we flip-flop the major suit responses, so we have:
1D-1H = Ss
1D-1S = Hs

Now after 1D-1H (showing Ss), without a S fit opener bids: 1S=5+Hs, 1NT=Cs, 2C=Ds, 2D=both minors. Or perhaps after 1D-1H (showing Ss), without a S fit opener bids: 1S(a minor, 1NT asks), 1NT (both minors), 2x=natural with Hs. That works.

What if the opponents compete over the 1D bid? They will often do this in Ss. So 1D(10-14)-2S(overcall)-Pass-Pass--? Now opener does not have any strength to come in here on most hands, even though shapely. How about we use an intermediate and weak/strong split:
1C: Unbalanced, not 5+Ss, 10-12 or 17+
1D: 13-16, unbalanced, not 5+Ss.

Now after 1D(13-16)-2S(overcall)-Pass-Pass--?, opener has the strength to compete on a bunch of hands, and also responder can make a value showing double, to get opener to describe hand, after 1D(13-16)-2S(overcall)-?

After the 1C opening (10-12 or 17+), responder assumes 10-12, and opener, if 17+ will show strength later.

At this point, we could now look at a whole bunch of world championship hands and try to guess how this scheme might have worked on each one. Say we do that, and we find that not showing the Hs immediately is costing on a bunch of hands. So we have auctions in the world championships where it goes 1H-4H, but using our new system we open 1D or 1C, responder bids low to find out what opener has, and the opponents come in with Ss. So we reconsider our approach and decide that one of 1C or 1D should show Hs, and the other either minor. We have another question too - should the opening that shows Hs promise 5+Hs or just 4Hs. The advantage of showing just 4Hs, is now we can take these hand types out of the other opening. The disadvantage is that responder does not know about the 5th H. However if responder bids a heart raise to 2H on just 3, and ends up in a 4-3, this is not necessarily bad - we have seen the four card major opening crew (such as the Australians who play transfer major suit openings) use 2 of a major raises with good success - it prevents the opponents from balancing as much since an 8 card fit is not promised, and so the assumed fit (they can assume a fit if we have a fit) is not as pronounced.

Also in looking at it closer, we like some of the collateral fallout if the either minor opening denies 4+Hs. Now if the opponents compete, responder can double to show general values, knowing that opener may rebid in Ss, but not naturally in Hs. In addition, the cheapest H bid by opener can be artificial, showing extras, which will keep the bidding low, and give responder a chance to rebid S if holding 4 of them. So this method seems okay from a competitive standpoint, though we would need to look at a whole bunch of hands to confirm this.

Let's look at our two prime options now:
1C: 4Hs, unbalanced, can have longer minor, less than 2C strength
1D: Either minor, unbalanced, not 4Hs, less than 2C strength.

1C: Either minor, unbalanced, not 4Hs, less than 2C strength.
1D: 4Hs, unbalanced, can have longer minor, less than 2C strength

What's the better scheme? Since the either minor hand type covers more hands, perhaps it should be in 1C, to give it more room to unwind. What we would do now is consider response structures for the two approaches, and we would focus on the either minor opening first, as it promises to get messy. For discussion purposes we will just concentrate on the responses for finding a major suit fit.

After 1D: Either minor, unbalanced, not 4Hs, less than 2C strength.
-- 1H: Relay, now opener's 1S=4Ss
-- 1S: 5+Ss.
-- 1NT: 5+Hs
That seems workable.

After 1C: Either minor, unbalanced, not 4Hs, less than 2C strength.
-- 1D: Relay, now opener's 1S=4Ss, 1H by opener is extra values.
-- 1H: 5+Hs
-- 1S: 5+Ss.

This last one is like a Montreal Relay (the 1D bid denying a 5 card major) combined with a Kokish Relay (the 1H rebid by opener) - since both are invented by Kokish, this would be Double-Barrelled Kokish. With a chance to play Double-Barrelled Kokish, we might just stop the design work right here, under the general principles that we must have arrived at the utopia of bidding design. However a system designer looks for areas where a system can follow a somewhat symmetric approach, so the memory work of the users will be reduced. Here the designer might go with a one-under scheme:

--1H: 4Ss (so one-under the major like the 1D opening bid)
--1S: Cs (following the general one-under idea)
--1NT: Both minors & extras (extras so opener has values if ends up in NT contract)
--2C: Ds
--2D: Both minors, no extras.

So what was the better scheme, opening 1C or 1D with the 4+H hand type? Now the system designer might return to world championship hands, trying to see what are the important characteristics of each scheme. Another consideration for the designer is what is allowed to be played in their country or region. Say the designer was in North America. Now opening 1H as 15-17 balanced would be ACBL Mid-Chart, but a defense would need to get approved as well. The designer might submit one for approval, and while we know defenses don't seem to be getting approved at the moment, the designer might still hope that it does. The sequence 1D(either minor)-1H(relay) would not be Mid-Chart, but the sequence 1C(either minor)-1D(relay, forcing) would be General Convention Chart (GCC), so the designer might go with the scheme that is covered under the GCC.

So the system design might look like:
1C: Either minor, unbalanced, 10/11 to near GF values, fewer than 4Hs, not-forcing
--1D: relay, opener rebids one-under Or bids 1NT or 2D with both minors, 1NT showing extras
--1H, 1S: five or longer in major

1D: 4 or longer Hs, unbalanced, can have longer minor, 10/11 to near GF values

1H: 15-17 balanced

1S: 5 or longer Ss, 10/11 to near GF values

1NT: 12-14 balanced

2C: Strong, 22+ balanced or any GF

2D: 18-19 balanced

2NT: 20-21 balanced

Now as the system is put into play, some might try 1NT as 15-17 and 1H as 12-14. Others might incorporate the 18-19 balanced into 2C. Some might try increasing the range of 1H, making it 15-18, or making it with two ranges, such as 15-17 with 10-11, or 15-17 with 20-21, or 12-14 with 18-20. The system designer can expect that if other partnerships use the system, they will make some modifications to it: putting their own stamp on it and, thus, creating their own unique system.

Tuesday, April 24, 2007

A great site to see links to the latest on bridge is Claire Martel's What's new?:

Here there are links to such things as:

Strong Diamond System by Noble Shore and Mike Gill
Danny Kleinman Bridge site (then check out his bidding and conventions section)
A complete system based on the French System Majeure by Geoffrey Ostrin (this really is complete - over 200 pages!)
Polish Club Manchester Style by David Collier
Sistema di Fulvio Fantoni in italian on Infobridge (use Google translate to view in English)
Bacon Torpedo Bridge bidding system developed at Durham University (including schemes for a singleton beer card)

If it is out there on the Internet about bridge, Claire Martel finds it!

Monday, April 23, 2007

One of things that system designers look at is how opener's 17/18+ hand types handle common jam situations. With responder just passing and opener to make first rebid, three measure points often considered are:

Opponents reach three of the highest unbid major (e.g. 1C-1S-Pass-3S--?)
Opponents reach four of the highest unbid major (e.g. 1S-2H-Pass-4H--?)
LHO (of opener) jumps to three of highest unbid minor (e.g. 1C-3D-Pass-Pass--?)

When these measure points are looked at, the standard and multi-way club opening systems are in some trouble. For example, it goes 1C-2S-Pass-3S--? in standard: now what does 18-19 balanced do - double for takeout, or try 3NT and find partner with nothing, or pass and find partner had a little something? Likewise, it goes 1C-1S-Pass-4S--? with a Polish Club 1C - are most 18-19 hands going to double here, and if they don't then too many times the partnership will end up in a bad spot in the five level.

If one plays a Mexican 2D opening (or uses something like BRASS - see ), then the 18-19 balanced hands are taken out of the minor suit openings. So after 1C-1S-Pass-4S--Double-Pass-?, responder knows that opener has a club suit and an unbalanced hand, and is better positioned to take appropriate action than in standard. One can see examples of this by looking at hands in the past few years of the Italian World champions.

One idea is to use the 2C opening to handle 18-21 quasi-balanced (named QBal: balanced or close to balanced, can be a 5-4-3-1 for example). This allows opener to express the approximate values in one shot. Let's look at an example framework:

1C: Natural, non-forcing, either:
-- a) Cs, unbalanced, 10/11-17
-- b) Balanced, 11/12-14
-- c) Cs, 17/18+ to near GF, not-QBal

1D: Natural, forcing, with:
-- a) Ds, unbalanced, 10/11-17
-- b) Ds, 17/18+ to near GF, not-QBal

1H/S: Natural, 5 or longer, 10-17, or 17/18+ to near GF not-QBal

1NT: 14/15-17 balanced

2C: 18-21 QBal, not forcing (can be passed with Cs and weak)

2D: Either:
-- a) 21+ QBal, but if just 21-22 then with a singleton
-- b) Any game force

2NT: 21-22 balanced

Rest: Preemptive package

So the 2D opening is somewhat overloaded - it is hard to unwind that much in so little bidding space - however the frequency is relatively low, so any damage will not occur that often. What one gains in this framework is putting the 1-level openings in a nice position, both for constructive and competitive bidding.

Like most system designs, the proof is in the pudding - the pudding of playing thousands of hands in reasonable strong fields and seeing what happens. For system lovers, this means trying to get some partner to play it - then after the thousands of hands if it turns out to be ineffective, one has a not-too-pleased partner, or ex-partner.

--- ---

A lot of system design is considering sets of hands, and then placing these sets in one spot or another. Here's a posting I made on in 2000 about the placement of these sets:

- A lot of system design revolves around the meanings to attach to the opening bids of 1C, 1D, and 2C, and what range of a balanced hand that a 1NT opening should show. Designers must determine how to unwind bids later in the auction, especially when faced with interference by the opponents.

- For discussion purposes about the set of meanings, assume that ranges of meanings can be Ordinary (about 11-14), Intermediate (14/15-17), Strong (17/18-21), or Big (22+). Also meanings can be Balanced or Unbalanced.

- For an opening bid, the meaning containing the lowest range of strength is the one that responder has to be first concerned with, in order that the bidding does not reach too high of a contract. For example, after a standard 1D opening, responder must first be concerned about opener having 11-14, even though opener could have much higher values. This is extenuated by the probability that opener is more likely to have less points than more points (assuming a minimum of 10/11).

- In competition responder must be concerned about opener's balanced hands, since these will be less likely able to bid freely. The distribution of unbalanced hands often provides opener with another call later in the auction.

- For a standard 1C opening, we can say it has the following meanings:
--1) Ordinary, Balanced, 3+ clubs.
--2) Ordinary, Unbalanced, 4+ clubs (5+ usual).
--3) Intermediate, Unbalanced, 4+ clubs (5+ usual).
--4) Strong, Balanced, 3+ clubs.
--5) Strong, Unbalanced, 4+ clubs (5+ usual).

Responder's first concern is 1), and then next is 2). If opener has 3) and responder is unable to describe hand or show values, opener will often be able to take another call. Likewise for 5). 4) is an interesting case - opener will sometimes be able to make another call, such as in the sequence 1C-1H-Pass-Pass--1NT, but sometimes opener is faced with a real guess such as 1C-3H-Pass-Pass--?. However what helps the standard 1C opener in making the guess is that responder, in the example case, has been unable to support clubs, or make a negative double.

- Playing a weak notrump standard system (e.g KS), the opening 1C bid does not have meaning 1) above, but now has meaning: 1b) Intermediate, balanced, 3+ clubs. For this system responder tends to be concerned about whether opener has meaning 1b) but also must be careful not to press too high in case opener has meaning 2). Hand type 4) becomes a little easier in competition since responder will endeavour to bid in case opener has hand 1b). So after 1C-3H-Pass-Pass--?, opener knows responder will try to find a bid in case opener has 1b), so with 4) close cases can be passed.

- In the proposed idea (that this post came from), 1D shows 11-18 balanced so there are essentially two meanings:
--A) Ordinary, balanced
--B) Intermediate, balanced

Responder must first be concerned with A) in the face of competition. However B) does not provide easy entry into the later auction. For example after 1D-3H-Pass-Pass--?, opener knows responder must worry about A) first, but holding B) opener knows that partnership might have the values for game, yet opener is not strong enough to bid over 3H, except perhaps with a double if short in hearts.

- It would be better if 1D had a stronger minimum strength or used a split range. If 1D was 15-21 balanced, responder could worry about the 15-17 balanced hand, and opener, if holding 18-21 could take a later bid. Likewise if 1D was 11-14 or 18-21 balanced, responder could worry about the 11-14 hand type and opener could bid again with the 18-21 hand type.

- One might ask if it is better to use a split or stronger minimum approach. Personally I believe in the Either/Or Principle (Anders Wirgren) so I like the split.

- The Nightmare system has a well thought out scheme for openings. In particular 1C is either:
--N1) Intermediate, Balanced
--N2) Intermediate, Unbalanced, 4+ clubs (5+ clubs usual)
--N3) Strong, Unbalanced, 4+ clubs, (5+ clubs usual)
--N4) Big, Any.

Responder's first concern is N1, knowing opener can often call again if N2 or N3, and that N2 will often know when to stay quiet if responder could not bid. 1D is either:
--D1) Ordinary, Unbalanced, 4+ diamonds (5+ diamonds usual)
--D2) Intermediate, Unbalanced, 4+ diamonds (5+ diamonds usual)
--D3) Strong, Unbalanced, 4+ diamonds (5+ diamonds usual)
--D4) Strong, Balanced

Responder's first concern is D1, knowing the opener often has another bid holding the other meanings.

- In one of my systems I try this for 1C:
--G1) Ordinary, Balanced, 3+ clubs
--G2) Ordinary, Unbalanced, 4+ clubs but not too distributional (not decent 6-4 or 7+ clubs)
--G3) Strong, Balanced.
--G4) Strong, Unbalanced, 4+ clubs (5+ usual)
--G5) Big, any without 5+ in major unbalanced or very distributional without clubs.

Responder's first concern is G1 and then G2. Basically 1C is a standard 1C opening with the Intermediate Unbalanced taken out [hand type 3) way above], G5 added in, and G3 allowed to have any balanced, not just with 3+ clubs and no five card major.

- One can map out the meanings for systems such as Precision or Polish.

Anyways if you are still reading at this point congratulations and I would appreciate any comments you have on the issues given above, especially if you disagree.

--- ---

That ending ("if you are still reading") is as apropos today as it was back then! These system design issues are interesting, but only to a limited few of us.

--- ---

So a key concept that could be used is that an opening consists of a base meaning, and also of stronger meanings that are related, but that have sufficient strength and/or shape to be able to handle themselves if the bidding gets competitive. Also, in non-competitive auctions, these stronger meanings can be shown by jump rebids, since they have the strength and/or shape to play there.

Let's look at a multi-tiered example, using some influences of Miles's Unbalanced Diamond, and the far earlier Katz-Cohen Breakthrough system. The "we have to get to the two level" break point here will be between 18 and 19.

1C: Intermediate or Big QBal, artificial, forcing, either:
-- a) All 15-18 QBals with 18+
-- b) All 23+ QBals with 18+

1D: Unbalanced "Diamond" (either minor) approach, either
-- a) 10-14, not balanced, no five card or longer major
-- b) Long minor, 15 to 18, not-QBal

1H/S: Natural, five or longer, either:
--a) 9-14
--b) 15 to 18 not-QBal

1NT: 11/12-14, balanced

2C: 19-22 QBal

2D: Multi. Either:
-- a) Weak-two in either major.
-- b) 5 card or longer major suit, 19+, not-QBal

2NT: 19+, long minor, not-QBal

Rest: Preemptive package

So for the 1 level suit openings, responder focuses on the high frequency a) option, knowing if opener has b), opener will be able to show it later. As you can see there are many twists that one could apply here to create your own versions.
I'm considering another system blend involving Savage type ideas. In Savage a core concept was the treatment of Quasi-Balanced hands (abbreviated QBal, these are hands that are balanced or near-balanced); with a big hand and not-QBal one could open a one bid, knowing if later a big hand was shown, it must be distributional. For example:

1H (if 18+ must be not-QBal)-4S(Overcall)-Pass-Pass-Double: this double promises a not-QBal, so responder knows to pull the double if shapely.

One problem with Savage was the Catch-All 1D opener, handling non-Qbals in either minor, from 13/14+. If we give up the weak-two in Cs (sadly losing a key EHAA influence), we can go with something like:

-- a) Cs, unbalanced, 10/11-17 (poor 17)
-- b) Balanced, 11/12-14
-- c) All QBals with 17/18+, including QBals with a five card major

1D: Natural, forcing, with:
-- a) Ds, unbalanced, 10/11-17
-- b) Ds, 17/18+, not-QBal

1H/S: Natural, 5 or longer, 10-17 (poor 17), or 17/18+ to near GF not-QBal

1NT: 14/15-17, balanced

2C: Artificial, forcing. Either:
-- a) Cs, 17/18+, not-QBal
-- b) 5 card or longer major suit, Game Force, not-QBal (this could be moved to a 2D Multi opening)
-- c) Something else, since opening not doing a lot of work here - could be weak with both majors etc.

Rest: Preemptive package

--- ----

This brings up thinking about limited one-level major suit openings, like in Precision. One concept has been that balanced hands define strength, unbalanced hands show shape. However the limit major suit openings, which tend not to be balanced, are one of the most effective parts of a Precision system, so why does it seem to be an exceptional success to the concept?

One key factor is how quickly responder can get to the right spot after an opening. Take the 1NT opening - responder will often pass, or bid 3NT, or bid Stayman, and then rebid to the spot, or invite. Very direct, 1 or 2 bids.

With a limit major suit opening, one can bid a game, or make a raise on a whole bunch of hands. If responder bids 1NT, or 2/1 first, responder can often place the contract on the next bid. Once more very direct.

When opener has a minor, things take longer to unwind. One usually needs to investigate for a major fit, and perhaps untangle what opener has. So we have frequent sequences like:

1D-1H--1NT-2C(Checkback)--etc., with responder taking at least three bids

Since responder will often take several bids over a minor opening, one might as well package a bunch of hand types into the opening, since opener will have time to unwind. From a design point of view, one wants to have the packaging such that if the opponents compete, some of opener's hands do not need to take any further bids (unless responder shows values), and big hands can either double or show shape.

--- ----

So returning to the system blend above, let's try an alteration with limited major suit openings:

1H/S: Natural, 5 or longer, 10-17 (poor 17)

2C: Artificial, strong, forcing. Either:
-- a) Cs, 17/18+, not-QBal
-- b) 5 card or longer major suit, 17/18+, not-QBal

That certainly fits in the ACBL General Convention Chart

--- ----

The in-between hands of the 1C opening could be a concern if the opponents get into the bidding. 15-17 Cs unbalanced is more than a 11-14 minimum, yet not quite a 17/18+ Big QBal, and might not be QBal to boot. So one looks at ways of moving this hand type out of 1C, such as:

-- a) Cs, unbalanced, 10/11-14
-- b) Balanced, 11/12-14
-- c) All QBals with 17/18+, including QBals with a five card major

1D: as before, but not forcing

2C: Natural, non-forcing. Either:
-- a) Cs, 15 to near GF, not-QBal
-- b) Cs, 15-17, QBal but not appropriate for 1NT opening

2D: Artificial, game force, a not-Qbal hand

--- ----

Let's conclude with a design walk on the wild (non-GCC) side:

1H/S: Natural, 5 or longer, 10-17 (poor 17)

2C: Artificial, forcing. Either:
-- a) Cs, 17/18+, not-QBal
-- b) Weak two in Ds
-- c) Weak with both majors

2D: Multi. Either:
-- a) Weak-two in either major.
-- b) 5 card or longer major suit, 17/18+, not-QBal

2H/S: Major + minor, weak

2NT: Minors, weak

So no weak two in clubs (sadness) but we have weak twos in the other three suits, and can show all two-suited weak hands. If the opponents compete over 2C and 2D, the strong hand type has many options to show strength. For the Multi opening, responder can make pass or correct bids in the majors, without concern for the strong hand type, since opener always has a major for the opening. 2NT for the minors is likely not the best treatment for the opening but it completes the two-suited weak openings.

Sunday, April 22, 2007

Justin Lall gives a good run down of bridge blogs here:

In particular I want to note two. First:

When Jonathan was in Ottawa, he played with most of the local experts, including myself, and it was a great time for bridge in the local area. Jonathan's blog promises to be a lot of fun, as he will speak very freely (as he says: "outspoken, blunt, direct"). I'm glad I didn't play online with him the other night (I was tired) for who knows what would have shown up on his blog shortly after that!


If you like my blog, you will certainly like that one, but the vis-versa is not necessarily true. This is because David explains a lot about bridge theory concepts, which makes his blog a great read. By contrast, here I dive right into the deep end of bridge ideas, so this blog is only targeted to a small part of the bridge community who can swim in this stuff (this is deliberate - I'm not trying to get 100's to use the suggested methods, but just trying to get select 10's considering the ideas - this is R&D for bridge scientists, not marketing). However the main site continues to have stuff of interest to most (that is marketing), so please check it out if you haven't already done so.

Seeing the approaches of others, I wonder from time to time if this blog, and perhaps the bridgematters site in general, should target a wider audience by reducing the complexity and density of material covered. However I worry that doing so would weaken the R&D aspects of the work here, and gain in the adoption of the suggested methods, which might not be a good thing, especially if it was widespread. For example, with 2dlay (the easiest of recent suggestions - see: ), this could be promoted to a wider audience (more examples, postings on various forums, article in a bridge magazine etc., things I have resisted doing so far), but should R&D stuff be rolled out for general use? I would rather see somebody pick up the 2dlay R&D and further advance it in some direction.

As an example of moving R&D forward, take the excellent development work that Henk Uijterwaal did with Transfer Walsh, from my small article in the Bridge World, to research discussions (in Dutch) and then to full fledged sequences covered in the Bruwil Bidding System book. So now we have some strong partnerships using T-Walsh type approaches, but the average player does not - this seems a good balance to me.

So after considering that, if you feel this blog should move in a certain direction that would help the collaborative R&D efforts, drop me a line or even better post a comment.

--- ----

A problem with J leads against suits from JTx, JTxx etc. is that partner doesn't know what to do if Axx(xx). For example, partner leads J in a side suit, dummy has Qxx, and you hold Axxx - should you play the ace in case declarer has a singleton K, or play low, waiting to later cover the Q with the A? Would a parity approach help? Say one leads J only from JTx and an odd number of cards in the suit - with even one leads a spot card. From KJT one would continue to lead J, regards of suit count. Could this approach be viable?

Saturday, April 14, 2007

Some more new articles on the site:
- 6MIA, TIM, and Mazzilli
- ETM A, a 4 page example system to show how the latest ETM methods can be worked into a system

If you study the ETM plug-n-play systems you will see different major suit raise packages for the major suit openings. This is not because one (the last) is better than the others, but instead the variety is provided to show some of the options that one can select in putting together your own system.

ETM Gold Premium is now available on the site in the complete form - all 259 pages in 10 point font, so there is an overwhelming amount of detail in these notes, albeit with a ton of useful stuff that can be incorporated into your own system.

If you do create your own system, send me a private copy of the notes to look at, if you can, as I love bridge bidding systems, if you haven't guessed that already by looking at this site!

Thursday, April 05, 2007

Another approach to implementing 5-4-4-4 systems is to use a tighter ladder for balanced hands without a five card major:

11/12-13: Open lowest four card suit with 1C, 1D, or 1H. Pass any 4-3-3-3 if 11-12.

13/14-16: 1NT. All 13s with a five card minor open 1NT, and 4-3-3-3 exactly & 13 opens 1NT.

16/17-17: Open 1 of a minor, and rebid 2NT, except if 4-4 in the majors and balanced, open 1H.

18+: Open at two level using 2C, 2NT, and BRASS 2C or Mexican 2D

So we have this 1 level framework:

a) 4 or longer minor, unbalanced
b) 4 or 5 in the minor, balanced, 11/12-13, not 13 if five in the minor, can be 3-3-4-3 or 3-3-3-4 only if 13.
c) balanced, 16/17-17, if 1C 3 to 5 Cs, if 1D 4 or 5 Ds

a) 5 or longer Hs
b) 4Hs balanced, 11/12-13, no 4 card minor, can be 4-4 majors, can be 3-4-3-3 only if 13.
c) balanced, 16/17-17, 4-4 in the majors.

1S: 5 or longer Ss

1NT: 13/14-16, includes 13s with a five card minor, and 4-3-3-3 exactly and 13.

1m-1NT: is to play opposite 11/12-13
1m-2NT: is natural, game force balanced
So by using the tighter range there is no game invite - assume opener has a 12 count and either go to game or stop in 1NT.
1m-3m is to play opposite 11/12-13, forcing to 3NT opposite 16/17-17. So 7/8-11 points and a 4 card or longer suit.
In responding to 1m, one will usually have 6+ points, so if opener rebids 2NT showing 16/17-17, then 2NT is reached with 6+17=23, so should be an ok spot - however the field is opening a 15-17 1NT and playing there so if 2NT fails it is a terrible result - this is a function of the notrump range structure, which will produce anti-field results that can be quite good and really bad.

1H-1NT: is to play opposite 11/12-13 & 4Hs, has fewer than 4Ss (bid 1S over 1H with 4Ss)
1H-2NT: is game forcing, balanced/semi-balanced, can have 3Hs.

So the framework is a 5-4-4-4(3) system, with the 3 being a balanced 17 hand with 3Cs and a four card major (4-3-3-3 or 3-4-3-3 exactly). However one idea is to downgrade these into 1NT, producing this pure 5-4-4-4 framework:

a) 4 or longer minor, unbalanced
b) 4 or 5 in the minor, balanced, 11/12-13, not 13 if five in the minor, can be 3-3-4-3 or 3-3-3-4 only if 13.
c) 4 or 5 in the minor, balanced, 16/17-17.

a) 5 or longer Hs
b) 4Hs balanced, 11/12-13, no 4 card minor, can be 4-4 majors, can be 3-4-3-3 only if 13.
c) balanced, 16/17-17, 4-4 in the majors.

1S: 5 or longer Ss

1NT: 13/14-16(17), includes 13s with a five card minor. Can be 17 if 4-3-3-3 or 3-4-3-3 exactly.

In this scheme, pass in 1st and 2nd seat if 4-3-3-3 exactly and just 13, for otherwise with 4-3-3-3 exactly one would be opening 1NT with the too wide of a range of 13-17: however in 3rd or 4th seat do open 1NT if 4-3-3-3 exactly and 13 since one needs to get this hand in somewhere - sometimes this will get one too high but most of the time it will land in the right spot, and 1NT on this 13 count can function as a blocking bid on the opponents - if one opens 1C with this shape in standard, the opponents will have the rest of the one level to come in.

So what we have here is a pure 5-4-4-4 system with a 2NT response to a 1 level suit opening being natural and game forcing: tres natural.

Wednesday, April 04, 2007

5-4-4-4 systems (where 1S is 5 or longer, all other 1 level suit openings promise 4 or longer) are highly natural. However there is the problem where to park the exactly 4-3-3-3 hand type. One option is to open at the two level with all 18-19 balanced (e.g using BRASS or a Mexican 2D), and in first/second seat if 4-3-3-3 exactly to open 1NT if 14 and pass if 13. In third and fourth seat, the system becomes 5-4-4-3, allowing 3 card 1C openings on 10-14 balanced.

Here's an uber-natural snapshot design:

1C: 4+Cs, 10/11-21, balanced only if 11/12-14
1D: 4+Ds, 10/11-21, balanced only if 11/12-14
1H: 4+Hs, 10/11-21, 4Hs only if 11/12-14 and balanced without a four card minor (can be 4-4 majors)
1S: 5+Ss, 10/11-21
1NT: 14-17 balanced. 14s with a 5 card minor open 1NT, and 14 with 4-3-3-3 exactly opens 1NT in first or second seat.
2C: BRASS - usual 2C hand types or 18-19 balanced without a five card major
2NT: 20-21 balanced

In third or fourth seat, 1C: 3+Cs, 10-21, balanced only if 10-14 - 10-11s with 3Cs will open but not rebid.

Granted the 4-3-3-3 exactly and 14s will get too high sometimes by opening 1NT - however the system gains lots of 1C-3C bounces, where 3C shows four or longer Cs and less than a game invite. 4-3-3-3 exactly and 12-13 passes in 1st or 2nd seat, knowing there is a strong chance to be able to bid again, using a takeout double if the opponents open. Note that if you do open 1C with 4-3-3-3 and 12-13 (playing a standard type system), and LHO becomes declarer, your partner will often lead to your club "suit", so not opening 3 card minors has its advantages.

When balanced hands with just 4Hs are put into the 1H opening, it means that 1H-1NT can no longer include game invite hands, since the flat hand with 4Hs needs to pass 1NT even if 13-14.

Here's the core framework of a 1H response structure to handle this:

1S: 4 or longer spades

1NT: non-forcing, can be distributional

2C, 2D: Game invite with 6 or longer in the minor, or a game force 3 or longer in the minor. Opener rebids 2NT if 4Hs balanced. 3 of the minor over any of opener's rebids shows the game invite hand.

2S: artificial, 3+Hs, game invite, if just 3Hs then fewer than 4Ss

2NT: game invite, natural, 1 or 2 Hs, fewer than 4Ss, no six card or longer minor

With 5-5 minors, or if 5-4/4-5 in the minors and a bad doubleton in spades, either bid 1NT or make a game force in a minor - there is no invite for these hand types.
Here's another snapshot design, blending elements of SOB and A-BOMBS:

1C: 17+ any OR 11-13 4Ss balanced with fewer than 4Hs
1D: 10/11-16, 4+Ds, can have 4Ds & 5Cs
1H: 10/11-16, 4+Hs, can be balanced 11-13 with 4Hs
1S: 10/11-16, 5+Ss
1NT: 13/14-16 balanced. 13 with 5Cs opens 1NT.
2C: 10/11-16 5+Cs, only a 4 card major when 11-13
2D: Multi, weak two in either major
2H: Both majors, weak
2S: 4S, 5+Cs, 14-16

In this case one has 1C-1D--1H available as Kokish (17+ Hs or big balanced). Btw Kokish likes Kokish to be called Birthright but this would not be namedright imo - something like HOBBI (hearts or big balanced integrated) could have worked at the start but now it will always be relayed as Kokish, as it should be.

Tuesday, April 03, 2007

A-BOMBS Club system is now up

It is like Precision, but has four key advantages:
- Instead of one diamond being nebulous, it is 4+Ds natural
- One club has a weak possibility, so the opponents can't jam the opening as much
- Two clubs only has a four card major if 11-13 - a style I liked from San Diego Precision
- It has a "free" 2D opening - it is used as a weak two in the notes but one can use it other ways.

It does have 14-16 with a 4 card major and 4+Cs opening 1H or 1S - this means responder can't bounce to 4M if just 3 in M - however while that works on a bunch of hands bouncing takes away 3NT as a contract, and often 5-3 major fits have a better landing there.

SOB Club would be niftier to play, but this system is far easier to learn and handle.