Join Bridge Winners
All comments by Craig Zastera
You are ignoring the author of this comment. Click to temporarily show the comment.
I don't think 3 is enough, particularly at IMPs (might try that at matchpoints).

Give partner some awful hand like Kxx-Kxxxxx-Kx-Kx and 4 still has play.
There are plenty of other hands where he will pass 3 where 4 is even better, e.g:
Kxx-KQxxxx-KJx-x
Jan. 12
You are ignoring the author of this comment. Click to temporarily show the comment.
You might be surprised to learn that a simulation (1000 deals) suggests that 3NT is a much superior contract to 4 AT MATCHPOINTS (and, of course, based on double-dummy play/defense).

On these 1000 random deals:
4H beat 3NT on only 258 deals
3NT beat 4H on 742 deals (there were no ties)

Interestingly, though, at IMPs, 4 was slightly superior:
NV: 4H beat 3NT by 532 IMPs
VUL: 4H beat 3NT by 598 IMPs

This is because 4 made on 97.3% of the deals while
3NT made on only 94%.

The big win for 3NT at matchpoints was because there were only 198 deals where 4 scored a “small win” (20-50 points), whereas there were 715 deals where 3NT scored a “small win (and on all but one of those, the ”small win" for 3NT was only 10 points, hence *0* IMPs).

Of course, this says little about how many pairs would actually reach 3NT vs. 4 in the real world, although since 3NT is so decisively superior, one might hope that good bidding pairs would find a way to reach that contract.

What does that mean about how 4 should be played?
Perhaps that (as Kit said) 4 declarers should play for a (relatively unlikely) lay-out where 4 can outscore 3NT.
Jan. 12
Craig Zastera edited this comment Jan. 12
You are ignoring the author of this comment. Click to temporarily show the comment.
Perhaps I should have “abstained” for lack of sufficient explanation by OPer.

I don't like the “fake” inverted 2 here, but that is not the issue.

I presumed that partner's 3 is 100% non-forcing. I play that it shows a minimum opener with 5 s and explicitly states no game interest opposite a minimum “inverted” 2.

Given that, I believe that pulling partner's double of (4) to 5 with this defenseless hand is clear-cut.
Jan. 12
You are ignoring the author of this comment. Click to temporarily show the comment.
I would never bid 3 with this hand.
Must show the suit and the invitational strength.

Ideal is 2NT followed by “correction” to 3 to show invitational strength with exactly four s.

I would do this even without explicit discussion as I think the meaning of this sequence is likely to be understood.

If I somehow didn't think partner would understand 2N then 3 as intended, I would jump to 3 “invitational” instead.
Jan. 12
You are ignoring the author of this comment. Click to temporarily show the comment.
Bad hand evaluation on your partner's part.

His is just about a maximum, especially after your try.

His T is good because it's in combination with AJ.
His T is good because that's your help suit.
His T is good because that's trumps
His 4th is good because it can produce a or ruff
after drawing trump.

I would value his hand on this auction as about 10 support points, hence easy accept.
Jan. 11
You are ignoring the author of this comment. Click to temporarily show the comment.
Speaking of slams, on the actual lay-out, 6 is a make.

On the auction start given, I was considering whether a 4 splinter jump (in support of s) might have been a reasonable alternative to the actual 4?

I'll agree that with limited opening bids (not specified in OP, but assumed from familiarity with OPer's current methods), searching for a “perfecto” for a slam might be too aggressive (although East's overcall increases the odds).

But since this is IMPs, perhaps 5 might be a reasonable alternative to 4 since the fit is longer (might even be 10 cards–opener could have 5=5 reds).
Jan. 10
Craig Zastera edited this comment Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, but R. Pavlicek's data suggests that statistically 6NT contracts make in real life only about 2.4% less often than double dummy results:
http://www.rpbridge.net/8j45.htm

So double dummy make percentage of 80% is not going to change enough with “real life” play/defense to invalidate my conclusion that this hand merits committing to slam.
Jan. 10
Craig Zastera edited this comment Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
Well, I am again surprised.

I, like the large majority, did pick “double” to show a maximum hand without 3 card support. I consider this an “optional penalty double.” Partner is invited to leave it in, but he can pull with a second suit or long s.

But I thought the obvious 2nd choice (and a “logical alternative”) was *3NT*. After all, we do have 12 HCPs with s stopped and are VUL vs. NOT (so the penalty we reap vs. (3) might not be enough if we can make a game).

Yet *no one* (at this point) chose 3NT. ??
Really, can 3NT be *that* unreasonable?
Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
Many people play that you have to go through Stayman to invite game in NT (i.e. Stayman does not promise a 4 card major).

Alternatives include:
(a) use 2N (1N-2N) as balanced NT invite
(b) use 2 (1N-2) as a range ask
© don't have balanced invites over 1NT with no
interest in a major suit contract–responder either
passes 1NT or raises to 3NT.

(a) suffers from using a potentially valuable bid for a very limited purpose.
(e.g. 2NT might better be used as a transfer or to show minor suit 2-suiters).

(b) suffers from ambiguity when 2 response is also used as a transfer. There is then a “conflict” for opener as to whether his “super-accept” is supposed to show good s or a maximum HCP hand.

© suffers (maybe) in cases where responder has a “pefect”
invite (i.e. would benefit from having opener make
the game vs. partscore decision), but is forced to
choose himself.
Jan. 10
Craig Zastera edited this comment Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
I did a large simulation of this hand.

Difficult to claim simulation can reliably determine best lead without more info, e.g. exactly what would partner need to double 3NT “for a lead”?

Also, would want more info about what hand types responder may or may not hold for his direct raise to 3NT, e.g. if he's 4333 with a 4 card major, would he bid this way?.
What types of unbalanced hands (say with a long minor) would raise to 3NT directly?

However, one question I believe can be reliably answered by a simulation is “among leads, which card is best?”

The simulation clearly shows that the answer to that limited question is the J (or T), because that lead defeats 3NT on more deals than either K or a small does (actually, the simulation showed that the K was by far the worst lead of any of East's 13 cards).

Note that I am not claiming (any) lead is necessarily best choice here. Just that among leads, the J or T will succeed in defeating 3NT more often than other card would.

Note: my simulaton actually had lead being considerably more successful than lead. Also 9 was better, but that was without any constraints to eliminate hands with which partner would have doubled 3NT for a lead.
Jan. 10
Craig Zastera edited this comment Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
Frances,
But I am not trying to design an entire alternative 1NT engine from the ground up.

Obviously, there are many different such engines possible.
Some are very different from standard US methods, but may work very well once understood. No doubt one can design a non-standard alternative engine that handles responder's slam try strength hands with a 4 card suit in some very different ways.

But all I am trying to do is to “tweak” the standard US methods where after Stayman and a 2M reply, responder's 3OM continuation shows a slam try in M with shortness and his 4 (or 4) continuation shows a slam try in M without shortness to handle the case where opener's reply to Stayman is 2 when responder has a slam try in *s*.

So here is responder with four s and a slam try, perhaps with shortness somewhere. So he bids Stayman, all set to rebid 3 over the hoped for 2 reply to Stayman to show his splinter slam try.

But, alas, opener replies 2. Now opener might very well have the four card suit that responder was looking for, but he hasn't shown it yet.

My method is just a way (I believe the only way at this point in the auction and assuming these methods) to allow responder to find out if opener has four s and then to continue the auction just as he would have had opener replied 2 originally to the Stayman inquiry.

None of this has anything to do with whether one could design some completely different response structure to 1NT (perhaps using transfers into four card suits or some other exotica) that would handle these responding hand types in some entirely different way.
Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
Michael,
My point is very simple. You seem not to have read enough to appreciate it.

When the auction goes:
1N-2-2-??
using standard US methods, responder has ways to indicate a 4=4 fit with slam interest, to wit:
(a) 3: to show unspecified shortness
opener relays with 3 to ask for location

(b) 4 (reverse Baze) or 4 (Baze):
to show interest in slam without shortness.

But what does responder do with the same hand types when
after 1N-2, opener inconveniently replies 2 instead of 2?

Opener may well (also) hold four s. How does responder check on that possibility while still retaining his descriptive slam try options as described in (a) and (b) above?

The answer is to play that responder's 2 continuation is a one round force. This still allows for his making the bid with only game invitational strength (and four s), but he may now also bid 2 with *slam interest* (and four s).

To make this work, opener's replies when he holds four s must be designed so that responder will still have the 3 slam try with shortness” follow-up available.

This is accomplished by limiting opener's replies when he holds four s to 3 (non-max) or 3 (max).

That way, responder with a slam try including shortness can continue with 3 to show that hand type in the “usual” way as he would have had opener replied 2 to his original Stayman inquiry.

This improvement does not conflict with responder's bidding 2 (1N-2-2-2) with a game-invitational strength hand.

If opener lacks four s, he replies 2NT (reject) or 3NT (accept), and we arrive at the same contract we would have had 2 been only a game-invite.

If opener has four s, he rebids 3 with a max and responder raises to 4.
If opener has four s and a non-max, he bids 3.
Now responder can either sign off in 3 or issue a “re-invite” with 3 if he wants to be in 4 when opener has a “middling” hand.
Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
Richard's example is correct.

Another declarer hand where underruff would be fatal is:
AKQ52-xx-xxx-KQT

But notice that this example also gives West an 8 card suit.

I do not think there is an example where West has only 7 s where East's underruffing at trick 2 will allow the contract to make when a round suit discard would defeat it (assuming West has the T).
Jan. 10
Craig Zastera edited this comment Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
No, what I am saying is that this hand is too good to put the burden of bidding a slam onto partner.

He will often have a minimum hand with no fit for either black suit, yet slam will still be good which is what my simulations suggest.
Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
If declarer's hand is:
AKQTxx-x-xxxx-Kx
then ONLY a “discard” of a small (not the 9) by East at trick 2 will defeat the contract.

The same defense (discard a small at trick 2) is also required if declarer's hand is:
AKQTx-xx-xxxx-KQ
Jan. 9
Craig Zastera edited this comment Jan. 10
You are ignoring the author of this comment. Click to temporarily show the comment.
4NT is way too conservative on this hand.

A simulation giving partner 15 HCPs, two s, three s, and a balanced hand had 6NT making 80% of the time.

So this hand should insist on slam.

If you think your partner might have four s, you might consider 5NT (pick a slam), but I would fear that if he chose 6 that would be too likely to be outscored by 6NT even when 6 could produce an overtrick.

A 1000 deal simulation giving partner 15 HCPs with four s and two s gave the following:
tricks in NT: 13: 181 12: 556 11: 226 10: 33 8: 4
tricks in s: 13: 398 12: 423 11: 156 10: 22 9: 1

But even though 6 makes more often than 6NT (821 vs. 737),
6NT handily beat 6 at matchpoints:
6NT beat 6: 749 deals
6NT ties 6: 69 deals
6 beat 6NT: 182 deals

Thus, a simple leap to 6NT over partner's 3NT seems best.
Jan. 9
You are ignoring the author of this comment. Click to temporarily show the comment.
If you can propose another solution that preserves all the slam try sequences that responder has after 1N-2-2M when responder has four card support for the major shown by opener (i.e. Reverse Baze and 3OM to agree M and show shortness) in the more difficult case where the auction starts:
1N-2-2
and responder has a slam try with four *s* (and opener might still have four s also), I would be interested in hearing about it.

In order to preserve responder's (eventual) 3 to show a slam try in s with shortness, the method needs both to allow responder to show his interest in a 4=4 fit *and* to allow opener to confirm that he hold four s all without going beyond 3 (so that responder will have room to bid 3 to show his slam try with shortness).

I do not see how to accomplish all that other than by the structure I have suggested.

Further my structure gives up almost nothing to achieve these goals. About the only possible criticism I can think of is that it is not possible to play in precisely *2* after the auction starts 1N-2-2.

I consider that pretty insignificant–if opener has < 4 s, it likely will be better to play NT. And if responder has 4 s, then given that responder has at least game invitational values and a 4=4 fit, I (as responder) would not want opener to be passing 2 with four s anyway.
Jan. 9
You are ignoring the author of this comment. Click to temporarily show the comment.
After 1N-2-2M, we have “reverse Baze” available to show slam try in the major bid by opener:
3OM: agrees M, shows unspecified shortness, opener relays to ask for location of shortness. Kickback is “on.”

4: agrees M, slam interest, no shortness
(and opener always has at least 4 available
as a “Last Train” type bid.

But what about when the auction starts:
1N-2-2-??
and responder has *s* with a slam interest hand?

Ideally, he would like to be able to find out if opener also has four s, *AND* still have the above “reverse Baze” tools available to show or deny shortness with slam interest.

Here is the solution:
1. After 1N-2-2:
2 by responder is *forcing*, shows four s and
either *game invitational* or *slam invitational+“
strength.

If responder has four s with GF strength (but less
than slam invite strength), he jumps to 3NT.

2. Over 1N-2-2-2, opener replies as follows:
a. 2N: minimum with fewer than four spades.
Responder passes with only a game invite but
bids (usually) 3NT with the slam invite.
If stronger, he can bid higher (e.g. 4N).

b. 3N: non-minimum (accepts game invite) with
fewer than four s.
If responder has a slam invite, he can continue.

c. 3: 5 s, more than a minimum, forcing.

d. 3: four spades, non-maximum
e. 3: four spades, maximum

Over both ”2d.“ and ”2e.“, responder now has the usual
”reverse Baze“ tools available (s agreed):
3: slam interest in s with unspecified shortness
4: slam interest in s, no shortness.
4NT RKCB s (at any point by either partner)

When responder has only game-invitational strength:
over 3 (”2e.“): he jumps to 4.
over 3 (”2d."):
i. he can sign off in 3
OR
ii. bid 3 as a further inquiry
(opener replies:
3 with a complete minimum
3 with an intermediate (e.g. 16) hand)
Jan. 9
You are ignoring the author of this comment. Click to temporarily show the comment.
If I had SS game tries available (as I do in my methods), I would choose that rather than a help suit try in s because KQ and KQ are roughly equally valuable (think of a SS game try as a help suit try in the other two side suits).

But without SS tries, a 3 long/help suit try is preferable to one in s.
Jan. 8
You are ignoring the author of this comment. Click to temporarily show the comment.
3 strongly suggests you have only three s.

Funny that you think this hand is too strong for 4.
I regard 4 as an immensely strong bid since partner could have a Yarborough.
I actually debated whether or not this hand was strong enough for 4, but finally decided it was.
One might actually consider just inviting game (e.g. via 3).
Jan. 8
.

Bottom Home Top