[nige1]

lamford, on 2017-February-07, 05:03, said:

The SIM showed that 4NT was overwhelmingly likely to make opposite 26-27, not opposite 24. So your conclusions are erroneous. I did not do a SIM opposite 24, but opposite 25, 4NT went off 7% of the time. How can someone get a PP for what appears to be the correct bid opposite 25-27 balanced, which partner is assumed to have for an in-tempo 3NT?

IMO, the results of a simulation should be backed up with the dealer-script code. For bridge-law purposes, a peer-poll would be more relevant than a simulation.

South's hesitation implies doubt.

He is less likely to hold 25-27 balanced.

He is likely to be stretching with a sub-minimum or underbidding with a super-maximum (e.g. 28 HCP). He'll pass 4N with 24 HCP, bid 6N with 28 HCP.

• If South holds 26 or a good 25 HCP, then many would prefer to stay out of 6N with a combined 30-31 HCP. The hesitation makes these holdings less likely.
• If South holds 24 HCP. then 4N should be OK with a combined 29 HCP. Of course, if South goes one down, then you'll regret that decision.
• If South holds a good 27 or 28 HCP, then 6N on a combined 32-33 HCP is likely to be an acceptable contract,

Hence, the UI makes a quantitative 4N safer.
.

[sfi]

wank, on 2017-January-31, 08:56, said:

---

3NT was described as "25-27, but he once had 28".

double of 2D was natural.

south thought for quite a while before 3NT (screens but agreed). at the table north bid 4NT and they made slam. i think we can all agree that pass is a logical alternative, but what does the hesitation suggest?

the director ruled result stands and the committee upheld it. quite a ruckus has ensued. opinions?

It sounds like their effective agreement really is 25-28, and the explanation supports this. I would find this range unmanageable, but it seems like 4NT is reasonable under these circumstances.

How many 28 counts has South held in this partnership when they chose a different action?

[barmar]

sfi, on 2017-February-08, 18:55, said:

It sounds like their effective agreement really is 25-28, and the explanation supports this. I would find this range unmanageable, but it seems like 4NT is reasonable under these circumstances.

How many 28 counts has South held in this partnership when they chose a different action?

Another way to look at it is that they don't have an agreement about what to do with 28. The one time it came up, he had to improvise, and he chose 3NT. Does that really make it an agreement from then on? What if the other partner would choose a different action?

Or maybe his action that one time was influenced by other factors (perhaps it was a flat 28, so he downgraded it). Is this any different from people who play 15-17 1NT, but choose to open Q32 QJ2 AQ32 AQJ with 1NT? There's a good chance they have more partnership experience with the latter type, too.

I don't think you can really conclude anything about their agreements from something that happened just once. About the only thing that supports your idea is that it seems like his action this time catered to a repeat of it.

[lamford]

sfi, on 2017-February-08, 18:55, said:

It sounds like their effective agreement really is 25-28, and the explanation supports this. I would find this range unmanageable, but it seems like 4NT is reasonable under these circumstances.

How many 28 counts has South held in this partnership when they chose a different action?

Given that the distribution of hands with 25 to 28 is as follows:
28 0.00
27 0.00
26 0.02
25 0.05
With the figures being to two decimal places, I expect the answer to your question is a round number. I would also think that the 28 count recalled was the only time this partnership had a 28-count. And I don't think it is unmanageable at all, 4C could be fit-finding with partner bidding 4NT with 27-28 and otherwise he bids a 4-card suit.

[lamford]

nige1, on 2017-February-08, 16:33, said:

IMO, the results of a simulation should be backed up with the dealer-script code. For bridge-law purposes, a peer-poll would be more relevant than a simulation.

I did not write the code myself, but used bespoke software, and I am nowhere near competent enough to generate the code! A peer poll is useful, but I, for now, often change my mind when "Computer says no", as David Walliams would say.

[nige1]

lamford, on 2017-February-09, 10:07, said:

I did not write the code myself, but used bespoke software, and I am nowhere near competent enough to generate the code! A peer poll is useful, but I for now, often change my mind when "Computer says no", as David Walliams would say.

OK. I simulated results for various strength openers using the following crude dealer-script code:

#
# Dealer script. 
# [url="http://bridgebase.com/tools/dealer/dealer.php"][/url]
# You hold SQ874 H86 DJT7 CQ865
# Partner has a balanced 26 HCP.
#
produce 10 # Bump this parameter to produce more deals.
predeal south SQ874, H86, DJT7, CQ865
W2D = hcp (west, diamonds) > 4 and diamonds (west) > 4
N3N =  hcp (north, diamonds) > 2 and 
hcp (north, diamonds) < 7 and
shape (north, any 4333 + any 4423 + any 5332 - 5xxx - x5xx - xx5x) and
hcp (north) == 26 # Change parameter for different opener strengths,
condition W2D and N3N
action printall # Change to printpbn for a pbn file for separate analysis.

Analysis of results confirms Lamford's conclusion that 6N is often a good contract.
Nevertheless, IMO, the hesitation makes 4N safer by reducing the likelihood of a midrange opener.

[gordontd]

nige1, on 2017-February-10, 06:17, said:

Analysis of results confirms Lambert's conclusion that 6N is often a good contract.

Who is this Lambert to whom you keep referring?

[nige1]

gordontd, on 2017-February-10, 07:07, said:

Who is this Lambert to whom you keep referring?

Thank you for pointing our my mistake, Gordon. Sorry Paul: I've tried to correct my misspellings

[nige1]

barmar, on 2017-February-03, 09:45, said:

Looks like there's now a BW thread discussing this same hand.

http://bridgewinners...e-2-7ckut8691n/

-- with a poll -- currrently 60% vote that the TD and AC should both roll the contract back to 3N+3.

[sfi]

lamford, on 2017-February-09, 10:04, said:

Given that the distribution of hands with 25 to 28 is as follows:
28 0.00
27 0.00
26 0.02
25 0.05
With the figures being to two decimal places, I expect the answer to your question is a round number.

So do I. Which is why the description of "one time he had 28" is, to me, a strong indication that it's worth catering for that possibility rather than suggesting that partner bids most 28 counts another way.

[mycroft]

My issue with all of that logic work is that "tank-3NT" means that almost all of those 7% are - gone. Partner doesn't have anything but the worst 25 (and my guess is that people "almost never" downgrade, and "very rarely" even *think* about downgrading); and no 26s or 27s. Partner either has something offshape (in which case bidding to give partner another chance to show is probably okay) or a hand where he really really doesn't want me to pass, but I could have nothing (but I don't have nothing).

I guess the biggest other question is "my diamond stopper is inflexible" - Kx or the like. In which case, you're happy to be in NT; but I don't think that hand suggests that "catering to the overstrength 3NT call" is a bad idea - partner will just happily pass, and we'll be 4NT+1 instead of 3NT+2.

My argument boils down to a feeling that a quick happy 3NT would have been explained as "25-27" and that one time partner had 28 wouldn't come to mind. The tank brought it to mind for the explanation, and the 4NT was catering to it happening again.

But I'm biased; when people hesitate-call against me, they welcome action from partner. And they frequently get it.