Join Bridge Winners
All comments by Kai-Ching Lin
You are ignoring the author of this comment. Click to temporarily show the comment.
Deleted.
Feb. 12, 2014
Kai-Ching Lin edited this comment Feb. 12, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Or

1 - 1 (4+ spades)
?
1NT = 4+ clubs unless 2-5-3-3 or 3-5-2-3
2 = 4+ diamonds unless 3-5-3-2
2 = 6+ hearts
2 = 3-card spade support, minimum or super extras (18+)
2 = 4-card support, minimum

With only extras (15-17) and 3-card support, you transfer first, then bid 2, as Barry said.

This version loses out to Zirconia sometimes.
Feb. 12, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
If you play Kaplan Interchange, transfer rebids have many advantages:

1 - 1 (at most 4 spades)
?
1NT = 4+ clubs unless 2-5-3-3 or 3-5-2-3
2 = 4+ diamonds unless 3-5-3-2
2 = 6+ hearts
2 = 5-4, minimum values
2 = 5-4, usual reverse
2NT and up = your preference

The benefits are:
1. No need for Flannery opening - because of the 2 rebid
2. No need for Gazilli - opener will have another chance to speak if he has a strong hand. In fact, you may need to think hard on how to define opener's higher-level rebids (2NT and up).
3. No need for Zirconia - responder can now force with 2 over 2 (diamonds). Recall that responder can't have 5+ spades.
4. In fact, responder's 2 (over 2 and 2) can be thought as third-suit game-forcing. You can certainly start your relays from there, if desired.

If your Kaplan Interchange is in a big-club context (like mine), then the following structure is worth consideration:

1 (11-15, 5+ hearts) - 1 (at most 4 spades)
?
1NT = 5(332) or 1-suited 6 hearts (min), non-forcing
2 = 4+ clubs
2 = 4+ diamonds
2 = 4 spades
2 = at least 6 - 4
2NT = 6+ hearts, maximum

The major advantage is opener's minor-suit rebids are real 4+-card suits - no more forcing no-trump death hands. It may seem odd that 1NT may have 6 hearts. But 1NT can easily be the best contract when responder is short in hearts. And if he has some heart tolerance, he can choose to bid on.

Erik - I have been incorporating game-forcing relays into 1-1 as well. I do the same for 1-1NT too. So 1M-2 can be used for something else.

Feb. 11, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Okay, below is the full structure after opener's 2 rebid. It may not be as effective as the major-suit cases, but it does meet all the purposes you stated above (and more).

Over opener's 2, if responder has 8-10, some club tolerance, may or may not have a 4-card major, he bids 3 to play. Opener usually does not move on unless he has an extraordinary hand. Here, we may stop at 3 even when opener has maximum.

With an invitational hand (say 11-13), he bids 2NT to invite. Opener is free to correct to 3 to play or bid 3/3M with a maximum hand. (You probably want 3 = 6-4; 3M = stoppers or shortness, depending on your preference)

With a game-forcing hand, he asks further with 2:

2 - 2
2 - 2: game-forcing inquiry
?
2NT = 6-4, min or max
3 and up = 1-suited clubs, min or max

2 - 2
2 - 2
?
3 = min, 1-suited clubs, some shortness. See the max cases below for continuations.
3 = balanced, some 6322, min (3) or max (3NT)
3 = max, short in diamonds, typically 3-3-1-6
3 = max, short in hearts, typically 3-1-3-6
3NT = max, short in spades, typically 1-3-3-6

2 - 2
2 - 2
2NT (6-4) - 3: game forcing inquiry
?
3 = min (His later 3 = short in hearts; 3NT = short in spades)
3-3NT = max., same shortness details as in the spade case earlier.

Yes, with this structure we don't get to play at 3 when it is right, as in the major-suit case. But that's a small price to pay.


Feb. 9, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Below is the structure I play:

2-2
?
2 = no 4-card majors, hence 1-suited clubs or 6-4
2 = 4 hearts, min or max
2NT = 4 spades, min
3 = 4 spades, max

So it is similar to Stayman over 1NT. Not too bad for the memorization. Over opener's 2 (4 hearts), responder can ask with 2NT. Over which opener's 3 = min; and everything else = max. Responder can ask similarly when opener bids 2 (= no majors).

With this structure, if opener has no major or wrong major, responder can still retreat safely to 3 to play, even when opener has maximum strength. Therefore, the requirement for the 2 inquiry can be lowered - 8-10 HCP, a 4-card major and a club tolerance may suffice.

As to shortness, below is an example:

2 - 2
2NT( min, 4 spades) - 3: asking
?
3 = short in diamonds (his later 3NT = min, 4-2-1-6; 4 = min, 4-3-0-6)
3 = min, 4-1-2-6
3NT = min, 4-0-3-6

So we always get to know opener's shortness before or at the level of 3NT and his void before or at the level of 4.
Feb. 9, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
It is probably not hard to quantify the playing record. A Bermuda win = x points, a Spingold = y points, etc. If someone accumulates enough points, he is automatically in HoF. We can then debate if playing record is all he or she needs.
Feb. 9, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
If playing record is all being considered for HoF (I was told), why can't we simply quantify it? A world title = x points, a Spingold = y points, etc.
Feb. 9, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks guys for the hard work. The result is close to what Dean predicted - he “guessed” the expected value to be +70 when vulnerable.

Could it be possible that in the future when we play online bridge we will have a deal generator next to us which would generate hands and conduct an expected value analysis at every stage of bidding and play? We then determine what is our best course or strategy at that point. Dean had tried to do that in some of his posts. This may be the future of computer bridge.

Now Thomas brought up another question - the Price of Dealership. He observed that there is an advantage of declaring 1NT first. This is less obvious than the vulnerability question, even theoretically - after all, the defense, who leads first, is one tempo ahead. On the other hand, in real life, 1NT is probably the contract that a declarer has most guesses to do. Anyway, the price, as far as 1NT is concerned, is perhaps small.

A bigger question, and harder to answer, is what the dealership is really worth. We all heard about the plausible advantage of being the “first to strike”. So the HCP requirement for opening a hand had gone down over the years, from Goren's 13 to Precision's 11, and then to some players' 8 count today. We probably won't be able to answer the question completely any time soon. But it does seem reasonable to open a sub-minimum hand when favorable, acknowledging that we will go down more often than not - the expected value of a favorable dealer's sub-minimum hand can be -100 points. Bocchi's idea of using the same system but with varied strengths at different seats/vulnerability is an excellent one.
Feb. 9, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Dean - I have this theoretical question for you, “The Price of Vulnerability”.

When you are white against red, what is the expected value of the deal before you even look at your hand? I believe it is not zero - game or slam bonuses are larger when vulnerable, which would offset easily vulnerable under-tricks. So the expectation, I think, is a negative number.

Put another way. Suppose you are playing a 4-deal Chicago rubber bridge (1st hand = love all; 2nd = red against white; third = white against red; and 4th = both vul). Let say you passed out the first two hands. Before the third deal is played, are you already behind? If the game has to stop now for some reason, can your opponents request some compensation? If so, what is that number?

Bocchi's article got me thinking about this. How about simulating a large number of hands, using double dummy to determine optimal contracts, scoring them two ways (red against white and white against red, and then calculating the differences? For example, if there is a 3NT making exactly 9 tricks, then it would be scored +600 (red against white) and +400 (white against red). The difference = 200. You do this for every deal in the simulation, add up the results and divide the sum by the number of deals you simulate. Note that the optimal contracts (or par) can depend on vulnerability, so it can get tricky.

Worth a try? Anybody wants to guess how big or small the number is?
Feb. 5, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Fantunes's 2 promises only 5. The Precision that Pollack-Feldman plays shows 6+ diamonds, which is about 2.5 times more likely than semi 3-suited hands with short diamonds. This is the version I personally prefer.
Feb. 1, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Barry, my definition of “relays” here is somewhat different from yours. Yes, you are right - 1NT opener can never find out responder's exact shape if opener is the relay captain (responder's hands are far too wide-ranging). My relays can only show an approximate distribution. For example, the sequence below:

1NT-2 (4+ hearts, canapés to minors possible)
2-2 (4 hearts, a longer minor or both minors)
2NT-3 (4 hearts, 5+ clubs)
3-3 (4 hearts, 5+ clubs, short in diamonds)

shows 3-4-1-5 typically, but a 2-4-1-6 or a 3-4-0-6 is also possible. Note that opener is required to make cheapest bids at every stage to allow responder maximum space to show all his features, including short suits. (He can violate this rule only when he has a special message to send, e.g. which shortness is irrelevant or want to set trumps early). So the bidding above does look like a relay bidding with 1NT as captain. That is what I meant. The structure isn't complicated (published as “Denial Stayman” in the 1994 Nov/Dec issue of Bridge Today), there are rules to follow, as in other relays.

The ability of showing shortness is quite important, as you are surely aware. It allows opener to assess first if they should get out of the no-trump or not when there is no major-suit fit (even if there is, 3NT can be better than 4M). On the other hand, relays with 1NT opener as teller is of more limited use, in my opinion. They are useful when you discover there is eg a 2-2 fit in some suit (then you probably don't want to play at 3NT); or when you discover that a 4-4 major fit is not profitable; or in slam bidding, an exact distribution is required to produce another trick. However, when responder is unbalanced, especially with a shortness, conducting relays with opener telling is not effective - responder does not know how much wastage the short suit faces.

Not knowing the exact shape is not a big issue for me. By the level of 3NT, opener would know all responder's 4+-card suits, in addition to his shortness. He can tell if responder has a 4-card or 5-card major, but he will not be able to distinguish a 4-card from 5-card minor, or a 5-card from 6-card minor. Once opener decides that 3NT is not likely to play well, he can use the whole 4-level to scramble to find a playable spot, which includes 4-3 major-suit fit or 5 of a minor.

I also have something to say (or complain?) about the 1M-2 relay. Both 1M-1NT and 1M-2 are asking (more or less the same)information from opener. This is a duplication of effort, similar to non-forcing 2 and game-forcing 2 Stayman over 1NT opening. If you are willing to play 100% forcing 1NT over 1M, they should be combined. This would free 2 for other purposes.
Feb. 1, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I suppose pass is a must if your partner would preempt with 4 at the 4-level with a 4-2-6-1 or open 4 with 6-2-3-2 (see Part 2, though you would probably have chosen 3 if available), where he has already taken majority of your hands into consideration.

Bidding or opening at 4-level with a 6-card suit reminds me of Marty Bergen's weak-two openings with 4-card suits in 80s. His “protection” came from “two-under”, i.e. opening 2 = spades; and yours is based on vulnerability (white against red).

I am not saying your style is wrong. In fact, your simulation/calculation of expectation values at any given point of bidding is impressive. Based on this calculation, your theory is that you want to reach the statistical optimal contract as quickly as possible, hence the 4 preempt or the 3/4 opening. And your partner should stay out of the auction unless he has an extreme hand (it will be helpful if you can also quantify this). At other tables, the bidding might go slower, with partner bidding 3 only and opponents has more room to maneuver.

However, the theory should be applied differently at different levels. When opponents can bid at 4-level, they are more risk-neutral (i.e. more aggressive, based on expectation values); while at 5-level, they are more risk-averse (i.e. more conservative, tend to take easy profits). So, preempting at 4 has a lot going for it and opening 4 does not.

By the way, I do not agree your pass over 1 opening, holding AKxxxx, for many obvious reasons.

Again, it is nice to see those expectation value calculations to quantify/justify the aggressive preemptions.

Feb. 1, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I, too, think that distributional invitation is the best use for 2. I also agree that relays is perhaps the best structure over 1NT - but with opener as captain and responder as teller. The system won't be able to distinguish 1-4-5-3 from 1-4-6-2 or 0-4-6-3, but responder can tell opener that he has 4 hearts, 5+ diamonds and short spades. That is, the structure allows responder to show all his 4+-card suits, as well as his short suits (if any) below or at the level of 3NT. Opener is then in a good position to judge the degree of fit of the two hands, especially if they belong in no-trump.
Jan. 31, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
This is probably a relatively easy hand for Precision players whose 2 shows 6+ diamonds, 10-15 HCP. The bidding can go, more or less,

2 - 4 (kickback)
4 (0 or 3) - 6
Jan. 31, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
“Trident identifies secondary suits and locates shortness at or below the level of 3NT.”

If this is the main purpose of Trident, then the following structure (“Denial Stayman”, published in Bridge Today, 1994 Nov/Dec) is perhaps easier and less vulnerable to opponents' preemption. It is similar to 4-suit transfers and each bid carries a single meaning.

2: Denial Stayman, a variant of Stayman
2: 4+ hearts, canapés to minors possible
2: 5+ spades
2: 4 spades plus (longer) minor(s)
2NT: 6+ clubs
3: 6+ diamonds
3: balanced slam try
3M: both minors, short in the bid major

Over the 2 transfer,
1NT-2 (4+ hearts)
2-?
Responder's
2 = 4 hearts plus (longer) minor(s)
2NT = 4+ hearts and 4 spades
3 and up = 5+ hearts and may have a secondary minor

I will give only the details for the case of 2 = 4 hearts plus (longer) minor(s)here:
1NT-2 (4+ hearts)
2-2
2NT-?
Responder's
3 = 4-5(opener can relay with 3, and responder bids similarly as the 4-5 case below)
3 = 4-4-4, short in spades
3 = 4-5, short in clubs
3 = 4-5, short in spades
3NT = 2=4=5=2

We use a lot of symmetries. You can probably guess how the bidding would go when responder has 5+ hearts and makes a bid at 3 or higher (his 3 = 1-suited hearts). Similarly, when responder has 4+ spades, he will use two bids 2 (5+ spades) and 2 (4 spades) to describe his hand.

Jan. 30, 2014
Kai-Ching Lin edited this comment Jan. 30, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
How about South bidding 4 over 4? If his earlier 3 may not be taken as natural, this 4 should be. South does not need to cue-bid 4 at this point to assure that the partnership has a diamond control, given that North has 0-7 HCP and has made a club cue-bid.

The auction can then continue as before, with North correcting 6 to 6, and South correcting 6 to 6NT at the end:

1 - 1
2 - 3
3 - 4
4 - 5
6 - 6
6NT

I thought the slam was well-bid, with the structure given. Standard bidders could well have the same issues too, if they play 2-2 waiting.



Jan. 28, 2014
Kai-Ching Lin edited this comment Jan. 28, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
I play a relay system over 1NT, with opener as relayer and responder teller. The system allows responder to show all his 4+-card suits, as well as his short suits(if any) below the level of 3NT. So, for this pair, East would show 4-5-2-2 to opener with slam interest and then bid 5NT as choice of slam:

1NT - 2 (4+ hearts, unbalanced, canapé possible)
2 - 2NT (4+ spades, GF)
3 - 4 (4-5-2-2, slam interest. 3NT would have been a lesser hand with same shape)
4(to play) - 5NT (choice of slam)
6NT
Jan. 26, 2014
Kai-Ching Lin edited this comment Jan. 26, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
What would South's 3NT over 3H have shown? If the partnership agreed to play as some sort of artificial slam try, then it may be easier to reach slam.

Let's say with Henry's balanced hand,?Qx ?AKxxx ?Kxx ?Axx, you would bid 3NT; and with real clubs which needs help, you bid 4C (as Michael wants).

Many partnerships use 3NT as serious/non-serious when a major-suit fit is found, sort of a quantitative investigation. Here, North had limited his hand with his 3H call, so a shape-oriented differentiation between 3NT and a cue-bid is perhaps more helpful.
Jan. 25, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
Is this construction of North's hand, KQxxx, QJxx, AK, xx,
reasonable? I would have bid 4H over South's 2H - opposite A10xxxx and out, 4H has a good play, assuming no spade ruffs. Furthermore, what is East's hand, if both clubs are off? AJ10x - QJ10xxxx xx is possible, I guess. It is more likely that East has some club honor(s).
Jan. 25, 2014
You are ignoring the author of this comment. Click to temporarily show the comment.
True…but that's another discussion on interference. Actually, this hand is not so much about big-club relays. If West opens 1H and East relays, they will get to the same place. In fact, if West starts with Flannery as Martin suggests, East could RKCB (in H) with 4C over 3H and then do the same spiral scans on high cards to reach 7H with 100% confidence.
Jan. 12, 2014
Kai-Ching Lin edited this comment Jan. 12, 2014
.

Bottom Home Top