Join Bridge Winners
All comments by Craig Zastera
You are ignoring the author of this comment. Click to temporarily show the comment.
Jeff,
I'd like to try some simulations for this problem but am not quite sure what constraints to put in for opener's hand.

You say 11-15 HCPs with 2+ s, but I'm suspecting that not all hands that satisfy those two conditions are 1 openings in your methods (?). For example, what is your 1NT opening range?

If you could give me a little more detail about the set of hands that qualify for 1 opener, I will try to use simulation to investigate relative merits of 1 (–> ) vs. 1 (–> 1N) responses.

Also, you might confirm whether opener will rebid 1 over 1 will *all* hands (meeting 1 opener requirements) that have exactly (and only) three s and rebid 2 with all hands that contain exactly four s, or whether he might sometimes deviate from this (say by re-bidding 1NT with three s or only 1 with four).
Sept. 2, 2018
Craig Zastera edited this comment Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Peg,
I'm not trying to “blame” anyone here. As I observed, range of possible strengths for advancer is way too wide for accurate game bidding.

So I suggested my “practical” guideline that absent contra-indicating bidding from the opponents, doubler should “assume” 5-8 range for advancer (that's support points, not just HCPs) and base his Pass vs. invite vs. bid game himself decision on that assumption.

I'm not a fan of the idea of a 2nd round 2 cue-bid by doubler here. His HCP strength is perfect for that call, but I think this should be reserved for strong hands with 3 card support for advancer's suit, except possibly in some very rare case where doubler is *so* strong that he has serious slam aspirations.

On OP problem, both hands are borderline–doubler between 3 and 4 and advancer between passing 3 and boosting to 4. As I mentioned, the somewhat unexpected great value of advancer's doubleton is important here.

I will observe, though, that opposite doubler's hand, 4 would be a reasonable contract opposite xx-Kxxx-xxxx-T9x (a quick simulation with that hand had 4 making 62% of the time).
This observation, to me at least, suggests that doubler's mere invite of 3 is perhaps a bit too conservative.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
A disgusting example of how “low” one might have to go with a 1NT advance after (1)-X.
Sure, 1NT advance is supposed to show 7-10 or perhaps even 8-11 in theory. But over 1M, especially 1, practical players know that it may be necessary to go lower (perhaps as here much lower) than that. Sad but true.
I must say I can't see a lot of virtue in 2 with this hand.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Danny,
It is obviously impossible to bid accurately if we assume a 0-8 point range for advancer.

Even a *4* point range is really stretching the accuracy of game vs. partscore invites.

So as a practical matter, doubler just has to “assume” that advancer is in the 5-8 point range for the simple suit advance and base his invites or game bids on that assumption.

So advancer “accepts” invite with 7-8 and declines with 5-6. When doubler has 20+ and 4 card support, he just bids game.

When advancer has less than 5 and this can't be diagosed by helpful opponents' bidding, we likely get too high.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
The arguments about invitational strength depend on the exact nature of the auction.
In 1NT-2NT types, little is known about “mesh” of honors between the hands in different suits, so more of a pure point count (with high spots and good/long suits being relevent too) decision.

In other invitational auctions, say those utilizing “help suit” or “short suit” raises where there is information about in which specific suit(s) secondary honors will have more value, more aggressive invitations can be used.

Also, there is a *huge* difference even in 1N-2N invite situations between matchpoints vs. IMPs NV, and IMPs VUL.
In the latter case, I have found little if any gain from responder's inviting rather than just deciding between 1N and 3N himself. This is because of the combination of extremely low “make percentage” required for 3NT to break even or gain vs. stop in a partscore *AND* the penalty paid by inviting with 2NT when opener declines (having to play a level higher).
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Richard,
I see a potential problem with your idea that opener's 3 over a 2N () or 3 () transfer should be NF.

What is opener to do with GF strength hand that needs to see if responder has enough of a tolerance to make 4 viable? Thus, I would think that 3 ought to be forcing.

I'm also curious about what range of strengths (and, to a lesser extent, suit length and quality) required for the minor suit transfers.
I would suppose that they could be quite strong (to cover the hands that would have bid 3m GF without transfers), but also extend down some what in strength to less than GF values. But how low can we go without the range becoming so wide that accurate constructive bidding thereafter becomes impossible?
I'm guessing that these minor suit transfers probably need to deliver at least game invitational strength to be playable, but I have no practical experience with them.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Interesting idea using the cue-bid as *not* a raise.
Schemes I've seen use 2NT as 4 card LR and the cue-bid as a (possibly 3 card) GF raise.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, playing “Good/Bad” 2NT in many competitive auctions, I've found that it works extremely well when the opponents do not bid further over the suit-ambiguous 2NT (whether it be played as the “good” bid or the “bad” one).

But when they do, the ambiguity of which suit the 2NT is showing causes difficulty when the other hand might like to compete further in one of the possible suits but not in other(s). This problem is present whether the 2NT is “good” or “bad”.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Thanks, Frances.
I had thought of transfers as an option and should have listed that explicitly.

They have more appeal to me than G/B here, although lots of details would have to be worked out, e.g. what is required for opener to do other than simple acceptance of a transfer into a minor.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
If you mean an Ace or a King, I would agree.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I don't think it is “clear” at all, particularly at matchpoints where hyper-aggressive game bidding is often a losing strategy. VUL at IMPs (perhaps even NV), I would think differently.

My view is that with 20 points and 4 card support, doubler should just raise to 4 and not whine “but you might have had nothing for your 2 advance.”

Merely to invite with that strong a hand is simply putting too much pressure on poor partner to bid again with a 5 count.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
My view is that in 2nd chair with s that are at best headed by QJ, most players would have 7 for a 3 opener. Perhaps :QJTxxx and a side ace would be possible too, so perhaps I should have included that in the simulation.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
As usual, it would be nice to know if 3 promises extra values (i.e. does “good/bad 2NT” apply here) or can just be a competitive noise with good shape but minimum values.

Anyway, 4 can't be too ridiculous with this hand.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I'm not a big fan of off-shape, out-of-range, or 5 card major 1NT openings.

But even to me, this hand seems like one where we need to open 1NT as “least of evils.”

The primary considerations for me are that I consider this hand definitely too weak to reverse (1 then 2), and also unsuitable for 1 then 2 (mainly, suit too short/weak).
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
True enough. And take away the J, and now game is not so good (but does have play). All of which just shows there is a lot of guess-work here, and luck about unknowable details will play a big role in determing the result.
Sept. 2, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Does it make sense to define 4NT as a slam try but *not* have opener responder key-cards?.

I presume this 4NT is forcing. How is partner supposed to respond if not # of keycards?
Sept. 1, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I chose 4NT under the assumption that is keycard for s.

Presumably, every partnership has a clear agreement about what 4NT is on this auction type.

If key-card 4NT not available, this is very tough.
6 (or 7) could be lay-down, or partner might have e.g.
Jx-Kx-QJxxx-KQJx.
I'd be torn between 4 and 6.
Very hard to say what the probability is that partner has what is needed for 6 (or more) on this auction.
Sept. 1, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
I think this is just unlucky.
While I won't go so far as to say partner has too much for his raise to 3, he certainly has an absolute max.

Suppose his hand were instead: xx-AQJx-AKxx-KJx.
That is a fine, sound raise to 3 (not at all scraping the bottom), yet 4 has zero play opposite your hand.

The problem is that there is just not enough bidding space for accurate bidding. If partner is going to worry too much about your having near zero for your forced 2 advance, he is going to miss a lot of games when he passes or raises only to 3 and you have something.

The raise to 3 should ideally show around 17-18 “support points”, so that if you are at the top of your bid (7-8 points), you bid game.

Here, he has 20 support points, which is quite a bit above the expected 17-18.

You have 6 “points” (counting 1 extra for the 5th ). So you are a bit thin to raise 3 to 4, but close. In a sense, you actually have 7 “support points” because the doubleton turns out to be valuable. But normally, advancer doesn't count anything for a doubleton in the suit partner doubled, because it is expected that doubler will have two or fewer cards in that suit.
Here, that proved to be incorrect which also contributed to missing a cold game.

So, in my view, you have a max pass of 3 and partner is a bit too heavy for 3 (although that choice could be the winner if he catches you with “nothing”).
Sept. 1, 2018
Craig Zastera edited this comment Sept. 1, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Best way to start with a problem like this is to run simulations consistent with the bidding and South hand.

The constraints are pretty well-defined. Could do even better if East's (favorable vul) pre-empting style had been specified.

Anyway, I assumed East has 7 s to the QJ, QJT, or QT with no more than 9 HCPs.

Partner can have a balanced hand of up to 11 HCPs or most any hand up to 9 points (but I eliminated weak 2s or higher pre-emptive openers). I disallowed 10-11 HCP hands that count to 13 or more “Goren points” (3/2/1 for void/stiff/doubleton) under the assumption that partner might have opened those.

I did a 2000 deal simulation with above assumptions.

3NT made slightly more often than 4–58% vs. 54%

However, variance in 3NT was huge. Sometimes as few as 2 tricks are available in NT, and sometimes up to 13.
3NT made fewer than 7 tricks on 18% of the deals, while 3 made fewer than 7 tricks on only 3% of the deals.

This variance resulted in 3NT and 4 being virtual ties at IMPs: NV, 3N beat out 4 by 178 IMPs, while VUL 4 beat 3NT by 22 IMPs (remember, this is over 2000 deals).

If this were matchpoints, 3NT would prevail over 4 with about 57% (BAM scoring).

Just the above would suggest (VUL at IMPs), it is a virtual toss-up between 4 vs. 3NT.

However, the added flexibility of 3, I think, makes that call the winner. By bidding 3, we will sometimes play in 3 (when partner is weak with modest s) and sometimes in 4 (when partner has a decent combination of support and values).

This flexibility should push the 3 overcall into the lead.
We will often reach 4 when it is a good contract and often stop in 3 when 4 would be too high.

Contrast this with 3NT which will almost always end the auction (for better or worse).
Sept. 1, 2018
You are ignoring the author of this comment. Click to temporarily show the comment.
Anyone who opens 1NT with this hand should not be allowed to write “15-17” on their card.
Sept. 1, 2018
.

Bottom Home Top