You are ignoring the author of this comment. Click to temporarily show the comment.
Michael,
I am certainly not an apologist for “NLTC” or any other “losing trick count” method.

However, it does seem odd that you would pick a pair of hands where the NLTC delivers almost precisely the correct answer for expected tricks for your criticism.

I'm sure you could find hand pairs where NLTC would perform significantly less well so that your criticism of its possible shortcomings would be better illustrated.

Anyway, my point in doing an “NLTC” analysis on this OP problem was to support the contention that it is *East* who overbid more significantly than West.

The West hand does come in with 6.5 “NLTC” losers, and that is without making any upgrade “adjustments” for the minor suit Jacks (which must be worth *something*).
Thus, according to NLTC theory, it just qualifies for an invitational raise (e.g. 1-1-3), albeit perhaps minimum for that category (6.0 losers, even 5.75 would probably still be in range for an invitational raise).

But look at the East hand.
It counts to *10* NLTC losers, making it below average for a minimum range response. Even making upgrade adjustments for the :JT (I believe NLTC references suggest 0.25 loser subtraction for such) leaves it showing worse than the “average” of 9.5 NLTC losers for a minimum response.

Anyway, I selected an NLTC analysis not so much to tout the wonderfulness of that method but rather to provide a quantitative way of comparing the West vs. East hands to determine which was more responsible for reaching an almost no-play 4 (my simulation had 4 making only about 1% of the time, and that is with double-dummy play).

The NLTC method predicts expected number of tricks as 8.75 while my 1000 deal simulation produced 8.703.

So with these hands anyway, the NLTC analysis is quite close to reality. And that analysis clearly shows that it is the East hand which is poor (a lot of losers), much more so than West (which is admitedly minimum for the chosen jump raise but not out of range).
Jan. 31
You are ignoring the author of this comment. Click to temporarily show the comment.
Patrick,
But the “NLTC” is supposed to be a big improvement over Klinger's 1986 book.

The NLTC is well documented in the book:
“Mastering Hand Evaluation” 2015
by Lawrence Diamond

This book was not the original source for NLTC but it does document and explain its details rather thoroughly.

I think the main idea of NLTC vs. Klinger LTC is that in NLTC the Ace, King, and Queen are not treated equally in counting losers.

On the contrary NLTC count is approximately
ace = 1.5; king = 1; queen = 0.5
Also, a small stiff = 1.5 losers (ace missing) and
a small doubleton = 2.5 losers (ace and king missing)

There are additional adjustments and details that I won't detail here (you can look at the Diamond book or find numerous on-line descriptions of “New Losing Trick Count”).

One point I will mention for NLTC as it is important is that after determining the NLTC “loser count” for each hand and adding these together, one then subtracts from *25* (rather than 24 as in Klinger) in order to estimate the expected number of tricks available.

Anyway, because NLTC treats A, K, Q differently, it does not have the need to make secondary reference to HCPs which I think Klinger's older LTC methods needed to do in an attempt to “adjust” for that method's treatment of A, K, Q equally w.r.t “loser” count.
Jan. 31
Craig Zastera edited this comment Jan. 31
You are ignoring the author of this comment. Click to temporarily show the comment.
Looks like 7 losers to me:
: 1.5 s: 1.5 s: 1.5 s: 2.5 = 7

Minimum opener is 7.5 losers, so this hand is 0.5 losers better.

I'll agree this hand has a “plus factor” or two:
* J is the main one–deduct 0.25 losers

* T and J must be worth something but not reflected
in NLTC count

So perhaps with these adjustments, might evaluate this as a bit more than 6.5 losers.
That is why I said I thought this close between 4 and 5.
Jan. 31
You are ignoring the author of this comment. Click to temporarily show the comment.
I think this is close between 4 vs. 5.

I chose 4 as this hand is only 0.5 “losers” better than a minimum–doesn't seem like quite enough to volunteer the 5 level. Partner still has another call.
Jan. 31
You are ignoring the author of this comment. Click to temporarily show the comment.
This is clearly far more East's fault than West's.

To demonstrate this quantitatively, we employ “New Losing Trick Count” (NLTC) which is especially appropriate when the partnership has identified an 8+ card major suit fit.

The West hand has an NLTC of 6.5:
s: 0 s: 2 s: 3 s: 1.5 = 6.5 total

The East hand has an NLTC of 10:
s: 3 s: 3 s: 2.5 s: 1.5 = 10.0
with possibly a small + adjustment for the JT, so say perhaps 9.75 NLTC losers.

NLTC specifies that a minimum opener is about 7.5 losers.
With one loser fewer than this, i.e. 6.5, opening bidder is entitled to make a game invitational rebid after an 8+ card fit is revealed. The original NLTC document says:
` "6.5 losers:    game-invitational values (e.g. jump to 3)  5.5 losers:    game-forcing values`

So, by NLTC theory, West's 3 rebid is right on the money.

The same document says for responder:
` "9.5 losers: minimum values (simple raise)  8.5 losers: game-invitational values  7.5 losers: game-forcing values"`

So this suggests that the East hand is perhaps a slightly *below average* minimum range response.

When East hears West's 3 invitational raise, he should envision opener with 6.0-6.5 NLTC losers for his bid.

For a 10 trick game (4 here), the partnership's combined NLTC loser count should be 15 (the formula for expected tricks is (25 - (NLTC1 + NLTC2))).

Thus, from East's point of view, even if his hand were as good as 9.5 NLTC losers (it isn't), that would mean West would need no more than 5.5 losers (9.5 + 5.5 = 15) for game to be good.

But West's hand can't be that good (5.5) losers, else he would have taken a stronger action than his 3 rebid.

Hence, by NLTC theory, East should pass 3.

Of course, one can arrive at this conclusion using other evaluation techniques or just plain old expert judgment.

But the NLTC explanation provides a proven quantitative technique for estimating the combined value of two hands playing in an 8+ card fit.

A 1000 deal simulation of these E/W hands yielded an average of 8.703 tricks per deal.
Whatever you may think of the accuracy of NLTC in general, it cannot be denied that for these two hands the NLTC estimation for expected tricks of 8.75 (25 - (6.5 + 9.75) = 8.75) proved (astonishingly?) accurate.
Jan. 31
Craig Zastera edited this comment Jan. 31
You are ignoring the author of this comment. Click to temporarily show the comment.
Passing 2 is absolutely wrong playing forcing (or SF) 1NT responses.
When responder has doubleton support for opener's suit, he is expected to preference back to 2M over opener's 2m rebid (unless he's going to make some strong rebid like 2NT).

Passing 2 should show some dead minimum hand with a stiff . Opener could have up to 18 HCPs and/or only 3 s for his 2 rebid.

“Semi-forcing” 1NT isn't much different from fully forcing 1NT in this respect. About the only difference of “semi-forcing” is that responder can't have a 3 card limit raise of opener's suit (as he could with forcing 1NT).
Jan. 31
You are ignoring the author of this comment. Click to temporarily show the comment.
First of all, even if 4 really required a 3-2 break, that would make it mandatory to bid. That would be a 68% game, and VUL at IMPs you should be bidding 40% games.

But second, 4 does not require a 3=2 break.
It can be made against a 4-1 break if the stiff is T or J (and you cash two top s correctly so as to be able to finesse on the 3rd round).
Double dummy can even be made against :JTxx with West, but probably won't find that in real life.
Jan. 30
You are ignoring the author of this comment. Click to temporarily show the comment.
I think South's pass of 2 is by far the most egregious error–so bad as to almost be grounds for getting a new partner.

South simply *MUST* preference back to 2. For all he know, North might have only 3 s.

But the more important reason is that South must give North another chance to bid (North might have 18 points).
On the auction, South knows that all his cards are working overtime since he has fits in both of North's suits and all his honors there too.

After South preferences to 2, North should raise to 3 as he can see that A and Q may easily be enough for game.

Then South would have to decide whether to take the push or not. I would say “yes” because of the ideal location of his honors.
Jan. 30
You are ignoring the author of this comment. Click to temporarily show the comment.
You can use DealMaster to show the winning lead only on those deals where the contract is beatable.

In fact, that is actually the *easier* thing to do.
The procedure is:
Enter the exact hand for East
Enter constraints for North and South
(i.e. North = 15-17 balanced, no 4 card major
South = what you think qualifies for
Stayman then 3NT)

Generate # of deals you want
click on “DF Analysis” to start Deep Finesse

Select contract of 3NT by North

Then, you will notice there is a check box labled:
“Determine best Opening Lead if contract will make”

If you click that box, then run Deep Finesse, the resulting opening lead stats will include ALL DEALS
(i.e. Matchpoint analysis).

However, if you do NOT click that box, then after you run
Deep Finesse, the resulting opening lead stats will include ONLY the deals where 3NT went down, i.e. the opening lead stats will be relevent to IMPs.

You can verify that what I've said is true by noting that the numbers shown for the various opening leads are much larger when you have clicked the “Determine best Opening Lead if contract will make” box (because all deals are included) then they are when you do not check that box.
Jan. 30
You are ignoring the author of this comment. Click to temporarily show the comment.
My simulations gave the same results (not surprising) as long as you are talking about MATCHPOINTS.

When I looked at results for IMPS where only the deals where 3NT was beatable (13% for you–about what I got too) are relevent, THEN the small lead was the clear favorite.

So the upshot is, when the contract can be beaten (not too often), a lead is the best chance to beat it.

But over all deals, a lead will (on average) give the defense the highest expected number of tricks, hence best at matchpoints where holding down the overtricks on the (majority) of the deals where 3NT is always makeable becomes the major concern.
Jan. 30
You are ignoring the author of this comment. Click to temporarily show the comment.
I think all the terror of playing 2N-3N as a relay to 4 (or perhaps something else) is unfounded.

My partners and I have been playing this way for *years* and have had a “forget” exactly once–the first time it occurred.

So we use 2N-3 as a relay to 3NT:
(a) for play in 3NT
(b) to show GF (slam interest) minor 2-suiters
i. 2N-3-3N-4/4/4N show:
3=1=4=5/1=3=4=5/2=2=4=5
ii. 3N-3-3N-4-4-4/4/4N show:
3=1=5=4/1=3=5=4/2=2=5=4
iii. 3N-3-3N-4 shows x=y=5=5

© to play 4 or 5 with a weak hand, long s
(after 2N-3-3N-, 4 relays to 4)

We use 2N-3N as a relay to 4:
(a) to play 4 (or raise to 5 for play): wk, long
(b) 3-suited (usually 4441) slam tries:
after 2N-3N-4, bid short suit (4N=short s)

So my advice is that if you want to design a scientific structure for responding to 2NT and find using a 3NT response as some artificial relay, do not let fear of forgetting stop you–it won't take much practice before this becomes 2nd nature.
Jan. 30
You are ignoring the author of this comment. Click to temporarily show the comment.
natural and forcing.
Jan. 30
You are ignoring the author of this comment. Click to temporarily show the comment.
It is possible to continue to play transfer advances when responder makes any call up to and including a single raise of opener's suit.

If his bid is lower than the raise, the TAs still just start with the cue-bid as usual and end with a simple raise of overcaller's suit.

When responder's bid is a single raise (most common), then DOUBLE is “stolen bid”, i.e. means whatever the cue-bid in the TA scheme would have meant. Then other calls by advancer above 2 of their suit up through the single raise of overcaller's suit retain their normal TA meaning.

We modify this slightly in one specific auction (not necessary, but we like it):
(1)-2-(2)-??
Here, we use double to show both minors (i.e. “responsive”), 2NT to show s (normally in TA schemes, NT advances are natural), 3 to show s, and 3 to show a good raise (as usual), and 3 as less than a LR.
Jan. 28
You are ignoring the author of this comment. Click to temporarily show the comment.
The ideal holding for a help suit try is Axxx.
The ideal holding for accepting a HSGT is Kx.

The value of shortness (stiff or especially doubleton) is dependent on trump length/strength.

A small doubleton is the HS isn't that good when responder has only 3 trump (particularly if not strong).

But with 4 trump, particularly if including A or K, a small doubleton in HS isn't too bad at all.

Any honors in the HS (particularly secondary honors) are good.

Obviously, opener can't always have Axxx.
So Axx, Kxx(x), Qxxx, even JTxx in a pinch are OK holdings.
Some 2 honor holdings like AJxx or KJxx are OK, but of course they leave fewer honors for partner to hold.

But “xxx” is definitely NOT acceptable for a help suit try.
There is another convention called “weak suit game tries” (played by few I think) where “xxx(x)” would be ideal.

Opposite a “weak suit” try, shortness (especially a stiff) is good, but any honors (particularly non-ace) are *bad*, so very different from a “help suit” try.
Jan. 28
Craig Zastera edited this comment Jan. 28
You are ignoring the author of this comment. Click to temporarily show the comment.
I do not think it is “standard” for a cue-bid by advancer to promise “support for partner.”

To make that work (assuming advancer is not a PH), one would need either:
(a) new suit advances are forcing
or
(b) some transfer advance scheme into unbid suits so
that the bid of the suit immediately below over-
caller's suit can be an unambiguous strong raise.

Even (b), as transfer advances are usually played (Rubens), involves new suit advances in suits that occur below the cue-bid to be played as forcing because the TAs usually start with the cue-bid. If these new suit advances were non-forcing, you'd have no way to show strong hands with that suit (I suppose a jump shift advance could be forcing, but even that agreement would not solve the problem since often the suit won't be long/strong enough to support a jump-shift advance).
Jan. 27
You are ignoring the author of this comment. Click to temporarily show the comment.
I'd like to answer, but the phrasing and options are too vague for that to be possible.

What does “outside of your 1NT overcall range” mean?
Does that include hands that are too strong for 1NT or just those that are too weak?

And among those that are too weak (say 14 HCPs or a flat 15),
what is my honor dispersion?

If I have none (or very few) of my points in s with just short of a 1NT overcall, that would make a “double” with some flat hand (say 4=3=3=3) more appealing that it would be with the same shape but some secondary values in s. Also, 11-12 HCP hands are a lot different from 14-15 HCP hands.

It would be better to give some (say 8) specific hands and ask on which of these hands would you double a Precision (1) opening.
Jan. 27
You are ignoring the author of this comment. Click to temporarily show the comment.
I either
(a) don't bid 1NT the first time
or
(b) pass 3 the second time

IMO, 6-11 is too wide a range. If it makes you think you should bid 3NT after partner says he wants to play 3, that suggests that you must think it is too wide a range too.

Why not just an invitational 2NT the first time? (or some conventional device to show a balanced NT invite).

That would be my choice.

But if somehow a 1NT response was dicated by my system with this hand, I would stay consistent and pass 3.
Jan. 27
You are ignoring the author of this comment. Click to temporarily show the comment.
I believe that in balancing seat, favorable vul, matchpoints, that balancer with 5=5 majors ought to make the partnership call showing “both majors” with almost any hand regardless of HCPs.

Who cares how few HCPs he has? The fewer he has, the more partner will have.

The opponents have stopped in 1NT. So they have AT MOST 24 HCPs, and will usually have fewer.

So as long as balancer has good shape, he should act.
With both majors, we have reasonable hope of outbidding the opponents, and even if we DONT, we may push them into a lower scoring minor suit contract.

I play DONT over strong 1NT, so our side does not have a strong (“penalty”) double.
Thus, partner (directly over their 1NT opener) can have a very strong balanced hand and still pass.

Perhaps my thinking would be somewhat more conservative if playing methods where partner would be expected to double anytime he has 15+ (or whatever) HCPs in a balanced hand.
But nowadays, I think few players play DOUBLE as “strong and penalty oriented” over their strong 1NT
(over their weak 1NT openers, it is a different story–in that case DOUBLE needs to be strong).
Jan. 27
Craig Zastera edited this comment Jan. 27
You are ignoring the author of this comment. Click to temporarily show the comment.
Bridge World Standard:
`"Forcing vs. nonforcing:   When a call could logically be interpreted as either   forcing or nonforcing, and there is no explicit   agreement:      In a competitive situation, treat as nonforcing;      in a noncompetitive situation, treat as forcing or      nonforcing by which seems more sensible to the      observer." `

Note: this is the default for undefined situations.
In BWS, new suit advances of simple overcalls of 1 level openings
are explicitly defined as non-forcing.
Jan. 27
Craig Zastera edited this comment Jan. 27
You are ignoring the author of this comment. Click to temporarily show the comment.
Bridge World Standard:
`  "A direct-position one-notrump overcall shows a   strong 15 to 18 HCP, regardless of the suit opened." `
Jan. 27
.

Bottom Home Top