Join Bridge Winners
All comments by Craig Zastera
You are ignoring the author of this comment. Click to temporarily show the comment.
2NT here is not natural for those of us who play “good/bad” (or “bad/good”) 2NT.

This is actually a classic auction for that convention.

The idea is to be able to differentiate between opener's mere competitive hands (which might be just a 1-suiter in “m”, or if “m' = , it could be a minor 2-suiter) vs. hands with the same shape but with game-invitational strength.

2NT shows one range (traditionally the competitive hands) while 3m (or 3 when ”m" = and opener has a good hand with 5=5 or 6=5 minors) shows the stronger ones.
June 27, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
He might have said it, but I do not believe that view is in accordance with majority modern views which, in general, would advocate overcall with this shape (as the voting supports).

Of course, shape alone should not be the only criteria.
Relative honor strength in the two suits as well as overall hand strength are also relevent.

Thus, if you made the hand AKQJ-xxxxx-x-KTx, you might get a different vote, even though this hand has same shape and HCP strength.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
As I read your comment, it sounds like you are advocating that this *is* a FP situation, but just using “pass/double inversion” semantics.

You say “North doesn't need a penalty double in this situation.” Why? The only reason I can imagine is that you must believe that South must double if North passes (or at least, not pass). Because if that weren't the case, then North *would* need a penalty double for all those cases where he thinks N/S can defeat (5).
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Leonard,
No, I think it is an algorithm–for any specific hand, I can tell you for sure what I would do. Now, there are surely hands where my current algorithm is *wrong*, i.e. some other action is demonstrably (via careful simulation) superior.

As those cases are discovered, I may tweak my algorithm slightly to bring it closer to matching theoretical results.

What I was saying is that the full details of my algorithm are too difficult (perhaps impossible) to encode into “DealMaster Pro”. That software only allows a maximum of *20* hand type descriptions for each hand (N/S/E/W).

That number (20) is often not enough to descibe exactly every detail of the factors that I would in fact consider relevent to my bidding choice at the table.
So when setting up simulations using DealMaster Pro, I generally have to simplify and approximate my at the table algorithms so as to be able to describe them within the limitations of DMP.

Sometimes, due to sheer laziness, I may not even reach the limits imposed by DMP when describing constraints.

For example, imagine some simple uncontested auction like 1N-3N vs. 1N-2N (invitational).

One should specify for the E/W (defensive hands), a complete description of every hand type they *cannot* have because they passed throughout rather than doubling, overcalling, pre-empting, etc. etc. But to describe all these “negative” cases would be very time consuming and would often run up against the “20” limit.

So often I explicitly eliminate only a few of the most obvious hand types the (passing) opponents cannot have
and hope that any ommissions will not materially affect the conclusions I draw from the simulations.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Need an agreement about what 3 shows.

I do think a popular agreement is that 3 is a cue-bid in suppoort of a contract. In this style, opener would bid 3 to confirm that suit as trump, else 3NT if not wanting to encourage a high contract.

But an alternative agreement would be for opener to assume responder is usually just looking for the best game rather than likely having slam interest. Thus, his transfer to s followed by s should be thought of as suggesting “worry” about one (or both) of the unbid suits for 3NT.

In this style, opener would still bid 3 if wanting to agree that suit, but could use 3 and 3 to show good values/stoppers in that suit while suggesting that 3NT might not be good unless responder has some help in the remaining unbid suit. Opener's 3NT instead would suggest adequate holdings in both pointed suits, only a doubleton , no great enthusiasm for 5 or 6.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I assume such 1 advance is forcing? That's the way we play it anyway.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I advocate playing “forcing pass” in more situations than seems to be fashionable nowadays (“forcing pass” popularity has been in decline for some time I think).

Nevertheless, even I would not advocate for FP on this auction.

We have not voluntarily bid game here–we kind of forced partner into 4 unless he wanted to risk passing (4X).
Typically FP only applies on auctions where we have voluntarily bid game, presumably with expectations of making it. Not the case here.

We are limited (no 2 opening) and partner has not promised any values at all.

But an interesting question is: “what if South really wants to encourage North to bid 5 if he has anything at all?”

If pass is not forcing, how does South send that message?
Tough one.

In theory, since PASS wouldn't be forcing, DOUBLE would be the only way to show “more” (of an offensive nature) than already indicated.

But wouldn't South also want to double here with some strong hand that is confident they aren't making 11 tricks without seriously wanting to invite partner to bid 5? I would think so.

The only solutions I see are:
(a) they win. South's double of (5) is just
accepting the consolation prize of a probably
smallish plus score. Not inviting North to bid
5.
or
(b) If South Doubles (5), he is showing offensive
extras and *is* inviting North to bid 5.
If he doesn't want to do that, he must just pass
(5). This isn't forcing, but hopefully North
will often DOUBLE anyway as they are very high
and opener has shown a very good hand.
or
© play forcing pass here, so that
* opener's pass of (5) explicitly invites 5
* opener must double to suggest “we're done.”
* we cannot defend (5) undoubled.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
While I picked 4, hence agree with the almost unanimous vote, I do not think that passing (4X) is completely crazy.

This hand is weak enough that it will not be a surprise if we cannot make 4 (particularly with the liklihood of bad breaks increased by the pre-empt).

Meanwhile, our side will usually have enough to beat (4X) as partner should be pretty strong for his 4 level double.

This is matchpoints.

So playing for neither side's being able to make 10+ tricks in their major, so that passing (4X) will lead to the best result available is not an outrageously anti-percentage position IMO.

Now if it were IMPs, passing (4X) would be a considerably worse choice, IMO, because the downside (when one or both the major suit games is a make) is very large. Better to bid 4 where the worst that can happen is we get a small minus when we might have had a small plus.

But at matchpoints, all that matters is frequency.
I'd be interested in what you all would estimate are the relative probabilities of the following two possibilities:
(a) at least one of (4) or 4 is a make
vs.
(b) *both* (4) and 4 will go down

My view is that the relative odds between the above two possibilities is not far from 50-50.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I'm actually surprised that no one picked 2 Michaels, although I have noticed before that this strategy appears to be less popular than I would have thought (Mike Lawrence has advocated it in print more than once, though).

The given hand is just about the right strength to try this tactic, although stronger s and weak s, say
AQJ8-K9654-x-KTx, would make it even better.

This assumes that one plays (as I think most do nowadays) “continuous range Michaels” where Michaels cue-bids (and unusual 2NT too) can be any strength from whatever the partnership decides is minimum on up, as opposed to the older style of “split range Michaels” (either weak or strong) where intermediate strength hands with the same shapes start by overcalling the higher suit.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Nicholas,
Addressing your comment about the difference between 5332 hands in the strong NT HCP range when the 5 card major is s vs. when it is s, I agree with you that there is a difference.

The difference revolves around the possible 1 response to a 1 opening as you say. In order to treat balanced opener's with 5 s and with 5 s identically, one would have to be willing to rebid in a 3 card minor after 1-1, just as one would after 1M-1N(F).

This is an idea that has some appeal to me because I really dislike opening 1NT with a decent 5 card major because of the undeniable fact that this will lead to missing playing in that major as a superior strain fairly often.

But I have not experimented with this in practice, mainly because my partners like the idea that 1-1-2m promises at least four cards in the bid minor.
And I don't have data to show that view is wrong–it is certainly undeniable that ambiguity about the length of opener's minor will also sometimes lead to problems.

My feeling is that probably these problems are not as frequent or as severe as missing good fits by opening 1NT with 16-17 balanced and a good 5 card suit, but that is just my speculation.

So currently, my algorithm for choosing between 1M and 1NT opening is slightly different when the major is s than when it is s.

With , I generally open 1 with 16-17 but 1NT with only 15 (because with 15, I fear 1-1N-2m-2-2N will get us too high too often).

But with s, I usually open 1N with 15 or 16, but usually 1 with 17. That is to avoid having to rebid a 3 card minor after 1-1.
But with 17, I'm (usually) willing to risk this problem or to raise 1NT to 2NT if I think the hand is close enough to justifying an “upgrade”.

Even the above is an over-simplification as judgment about individual specific hands also comes into play.
For example, there are some (poorish) 15 counts with 3=5 majors where I will open 1 because I will be content to raise a 1 response to 2 if necessary.

Also, the quality of the major suit, the strength of the doubleton, and spot card quality all modify the above general algorithm.

Because it would be very difficult to try to specify all these secondary factors when doing *simulations*, I usually simplify my 1M vs. 1NT criteria when doing simulations to:
1NT with 5 s only with 15 HCPs
1NT with 5 s with 15-16 HCPs
1M with 5 in either major with 17 HCPs
even though that is only an approximation of what I do at the table.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I think 1 is better than DBL because missing a 5-3 fit is too likely if you double.

I think a 2 Michael's cue-bid would be a reasonable alternative–partner should pick s unless he has longer s, so this is likely to work OK.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Why would it be a surprise to make 4?
Give partner: AQxx-x-KQxx-AKxx
which seems like a typical sort of hand for his double and 4 should have good play.
June 26, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
No mistake in my conclusion. What I quoted above was (and I stated it as such) an *example*. The results I reported were precisely what I measured when simulating this particular example.

I did say that as a *general rule*, 5332 hands opposite 4333 hands with combined HCP strength around 25 will generally make 3NT slightly more often than 4M in a 5-3 (even 5-4) major suit fit. The example hand I simulated produces results consistent with that observation, although even if it didn't, as it is only one example it would not demonstrate anything about the general case.

I also pointed out that this example hand does better in 4 AT MATCHPOINTS than in 3NT opposite random 4333 7 counts (with 3 s). I did not state whether or not I believe this is true in general or what specific properties the 5332 hand would need to have to make this statistically valid.

I will say, though, that in general and on average a 5-3 (not to mention 5-4) fit will produce *more tricks* playing in the suit fit than in NT.

Whether or not this will be enough to make it better to play in NT or in the suit depends on many factors–partscore vs. game; matchpoints vs IMPs; weak doubleton vs. strong doubleton; strength of the 5 card suit, etc.

I will also state as a bonus piece of information that when the responding hand, with 3 in opener's 5 card major is *NOT* 4333, even if relatively balanced (4432 or 5332), with around 25 combined HCPs, playing 4M will on average be significantly better than 3NT at matchpoints or IMPs.

BTW, many think that this example hand should be opened 1 because it is “too strong for a 15-17 1NT.” However, if that is their reason for preferring 1, they are simply wrong.

I have done extensive simulations with this particular hand which show that while it performs better than an “average 17 HCP balanced hand”, its performance is still closer to that of an average 17 HCP balanced hand that it is to the performance of an average 18 HCP balanced hand.
Thus, an “upgrade” is not justified on the basis of the hand's playing strength.

That is entirely different from saying that it should be opened 1 rather than 1NT. I agree with that. But it is not because it is “too strong” for 1NT, but because opening 1 will lead to a superior contract opposite many hands where that suit fit would not be found if 1NT is opened instead.
After opening 1 and hearing a (forcing) 1NT response, you can decide whether to “upgrade” to 18 and rebid 2NT
or to treat it as 17 and rebid 2, planning on continuing with 2NT should responder preference back to 2.
My simulations suggest that the latter strategy is sufficient with this hand, although it is fairly close.
June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Leonard,
Be very careful in what you conclude from the simulation results to which you allude.

I would agree that 4333 opposite 5332 in the borderline game strength range (about 25 HCPs) will (on average, statiscally speaking) make 3NT slightly more often than 4M in a 5-3 or even a 5-4 fit (double dummy analysis, of course).

So at IMPs, that is sufficient to prefer 3NT in these cases (assuming you can identify them in the auction).

But what about matchpoints?

As an example, I recently did a 1000 deal simulation of
AKT43-KQT-86-AJ2 (a hand discussed in another recent BWs post) opposite random 8 HCP hands with exactly 3 s and 3-4 cards in each of the other suits (i.e “four triple three” shape).

On this simulation, I found:
3NT made on 619 deals
4 made on 591 deals

So this agrees with the assertion that 5332 opposite 4333 generally makes 3NT a bit more often than 4M in a 5-3.

But comparing 4 vs. 3NT on these 1000 deals at Matchpoints reveals that:
3NT beats 4 on 412 deals
3NT ties 4 on 129 deals (both down the same #)
4 beats 3N on 459 deals

So 4 turns out to be better at matchpoints.
This is easily explained because s often makes 1 more
trick than NT.
June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
But that is just the hand most likely to have been opened 1NT with a 5 card major (see above discussion).
June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
(a) my partner's don't generally respond to 1M with 5.

(b) with *15* I usually open 1NT with 5332 and a 5 card
major for the reason you suggest.

© with 16-17, 5332 and a decent 5 card major, I am
100% convinced from experience that opening 1M gains
more than it loses, despite the possibility of maybe
getting to 2NT with 16 opposite 6 occasionally.

If you instead open 1NT, you will miss your 5=4
major fit whenever responder has fewer than 8 (and
often even with 8 he will pass 1NT).
You will miss your 5-3 major suit fits almost all
the time (unless you play Puppet Stayman which brings
its own problems).
June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Quite a few people play “bad/good” 2NT instead of the traditional “good/bad”, i.e. they use 2NT to show the “good” and immediate 3 level suit bid to show “bad.”

I think there are reasonable arguments both ways on this issue.
One point that is clear–on whichever hand types you bid 2NT (whether “good” or “bad”), that call will leave partner in the dark as to which suit you are coming in.

This probably doesn't make a lot of difference when next hand (i.e. opponent) just passes, but it can be a problem when that opponent instead takes up space by bidding again.

Then, if partner knew your suit(s), he *might* have a good fitting hand where he would compete. But because the 2NT is ambiguous, he will often fear to bid again, so sometimes we get boxed out when we have a good fit.

Consider the auction:
1 (1) DBL (2)
2NT?

Opener might have:
(a) a 1-suiter
(b) a / 2-suiter
© 4 card support
either minimum (if playing “good/bad”) or game invitational (if playing “bad/good”).

When overcaller now continues with (3), responder might well want to act again if opener has one of these hand types, but not if he has one of the others.

For the above auction, in one partnership we have agreed:
* with (a) or (b) we play “good/bad” so that 2NT
shows minimum with s or minimum /,
while immediate 3 or 3 is stronger (game invite)

* but with ©, we play “bad/good” so that immediate
3 is just competitive while game invitational
hands with support start with 2NT.

I think you can see some logic in the above, although it certainly isn't a panacea when they bid (3) over 2NT.

My point is that even in traditional Lebensohl (i.e. “Good/Bad 2NT” situations), it is debatable whether 2NT should be used for the “good” or the “bad” hands.
June 25, 2018
Craig Zastera edited this comment June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I have refuted this point previously.
Easy to show strong NT with 5 card major after 1M-1N(F)–you bid 2m, then 2NT next time (typically over 2M).

If partner passes 2m (quite rare and never with 2+ in M), we should be in an OK spot.

What is irrefutable is that if you open 1NT with a decent 5 card major, you will miss playing in that major quite frequently when responder has 3 or even 4 cards there.
This doesn't have to be bad, but it will be fairly often.
June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Does anyone play 2 is forcing on this auction?
June 25, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Glad to see that 3 different choices have received almost equal numbers of votes and a fourth choice has also attracted a fair amount of support.

But I was hoping for more discussion of the merits and drawbacks of the different approaches.

Here is one issue:
How many of you who choose 4M (major of your choice) are planning to give up if LHO continues with 5 and that comes back around to you?

I'm guessing that not many would do so (but maybe I'm wrong).

If that is true, then what is the distinction between bidding 5 (majors) immediately, vs. say bidding 4 and then 5 if/when 5 comes back around to you?

My view is that immediate “2 suited” bids tend to show either equal length/strength in the two suits or possibly slightly longer/stronger in the *lower suit*.
This style leads to the idea that partner should pick the lower suit with no clear preference.

It also suggests that when “2-suited guy” is longer/stronger in the higher suit, he should generally bid that suit first and then the lower suit later (if possible).

If you buy those ideas, then bidding 4 here, planning to bid 5 next over (5) would seem to be a bad idea as that would definitely suggest longer/stronger s (perhaps 6=5), just what you don't have.

An immediate 5 would therefore appear to be a better choice with this hand (if willing to reach the 5 level regardless), as it gives a better description of the relative holdings in the majors (partner won't pick s unless he has more of those than s).

The downsides to immdediate 5 are:
(a) gets us to the 5 level perhaps unnecessarily
(b) might (depending on partnership agreements/style)
make partner think we have a stronger hand and
therefore bid 6M (or 6) thinking we have
a slam when we dont (e.g. QJx-Kxx-Qxxx-Axx).

Comments?
June 25, 2018
.

Bottom Home Top