[hrothgar]

Cascade, on Jul 18 2009, 11:49 AM, said:

hrothgar, on Jul 18 2009, 08:09 AM, said:

Cascade, on Jul 17 2009, 11:02 PM, said:

I thought I did too.

I wasn't sure whether your simulation included both the change in the conditional probability and the effect on the frequency of a first/2nd/whatever seat opening...

1st column - frequencies for 100 000 000 hands

2nd column - frequencies for 100 000 000 hands constraining 1st seat to have a PASS by the rules I outlined

etc

Hi Wayne

That accounts for the conditional probability that you have X HCPs

Art is claiming the following:

Lets assume that We are playing a 10-12 HCP 1NT opening
We're sitting in third seat seat.

In order for us to have the option to open anything, 1st seat and second seat needs to pass. So, we need to multiply column 3 by

(1 - The probability that first seat opened) * (1 - probability that second seat opened)

Art is also claiming that this probability changes significantly based on whether we hold a 10-12 HCP NT opening in third seat

[TimG]

I generated 10,000,000 deals and counted how many times south would get to open a mini (10-12), weak (12-14), or strong (15-17) NT in 1st, 2nd, 3rd, or 4th seat.

 [space] [space] [space] [space] [space] [space]1st [space] [space] [space] 2nd [space] [space][space] [space]3rd [space] [space] [space] 4th
mini [space]	1,028,996 [space]	622,642 [space] 318,994 [space] 103,323
weak [space] [space]	809,465 [space] 535,865 [space] 317,150 [space] 146,547
strong [space]	407,163 [space] 303,392 [space] 213,693 [space] 137,097

Notrump openings were 4333, 4432 or 5m332. Opening bids were 12+ HCP with any shape or 6+ HCP with 6+ spades, hearts or diamonds or 7+ clubs. Even in 3rd seat. In other words, everything was very loosely defined, but I think this gives a general idea of what you were looking for.

Spoiler

Initial random seed 1247927503

[Cascade]

What happened to 4th seat?

[blackshoe]

He went home.

[ArtK78]

The probability that the player is rational greatly reduces the probability of a 10-12 1NT opening in 3rd seat. As I have said before, opening a 10-12 1NT opposite a passed partner is not rational (suicidal is the description that I would use).

As for 4th seat, an argument can be made that it makes some sense, since both opponents have passed. But I would not wait around for 3 passes to a balanced 10-12 HCP hand. You will be waiting a long time.

[dake50]

Similar to Cascade, old chart. Notice the obstructive value in third seat: 4-seat after three passes AVERAGES 15 hcp! Obstruction is waaay to often ignored. What is the value of mini-NT taking 1-level away? Does opponents game-partial decision accuracy go down? How much? Enough to balance penalty disasters? I think YES.

HCP\% by seat: previous seats opened by rule of 20. Last column if rule of 19 3-seat. 1-seat calculated, simulation rest.

1-seat 2-seat 3-seat 4-seat 4if 3wRuleof19
10.00hcp 10.85hcp 12.18hcp 14.72hcp 15.20hcp avg
0 0.36390 0.1836 0.0461 -- --
1 0.78844 0.4376 0.1296 -- --
2 1.35612 0.8160 0.2921 -- --
3 2.46236 1.5955 0.6548 -- --
4 3.84544 2.6702 1.2603 0.0036 0.0002
5 5.18619 3.8514 2.0780 0.0244 0.0042
Sum0-5 14.00245 9.5543 4.4609 0.028 0.0044

6 6.55410 5.1854 3.1909 0.1428 0.0425
7 8.02809 6.7909 4.6620 0.5232 0.2227
8 8.89219 7.9726 6.0817 1.3814 0.7762
9 9.35623 8.9188 7.5014 2.8356 1.8407
10 9.40511 9.4352 8.7345 4.8932 3.6540
11 8.94468 9.4158 9.5244 7.2309 6.0546
sum6-11 51.18040 47.7188 39.6949 17.0071 12.5907

12 8.02687 8.8816 9.7374 9.4149 8.5070
13 6.91433 8.0119 9.4431 11.0059 10.5347
14 5.69332 6.8862 8.6968 11.7888 11.8666
15 4.42368 5.5606 7.4542 11.4838 11.9597

sum12-15 25.0582 29.3403 35.3315 43.6934 42.8680

16 3.31092 4.3177 6.1438 10.3759 11.2121
17 2.36169 3.1826 4.7367 8.7604 9.7109

sum16-17 5.67261 7.5003 10.8805 19.1363 20.9230

18 1.60508 2.2363 3.5064 6.8304 7.7863
19 1.03617 1.4846 2.3929 4.9613 5.7881
20 0.64354 0.9411 1.5814 3.4009 4.0401
21 0.37787 0.5680 0.9730 2.1848 2.6064
22 0.21004 0.3203 0.5682 1.3081 1.6044
23 0.11190 0.1728 0.3107 0.7321 0.9061
24 0.05590 0.0896 0.1635 0.3857 0.4721
25 0.02643 0.0428 0.0770 0.1832 0.2325
26 0.01167 0.0193 0.0369 0.0904 0.1060
27 0.00491 0.0085 0.0142 0.0383 0.0465
28 0.00186 0.0029 0.0053 0.0124 0.0149
29 0.00067 0.0009 0.0017 0.0043 0.0081
sum30+ 0.00030 0.0005 0.0009 0.0019 0.0026
sum18+ 4.08644 5.8876 9.6321 20.1338 23.6141

[Mbodell]

ArtK78, on Jul 18 2009, 07:10 PM, said:

The probability that the player is rational greatly reduces the probability of a 10-12 1NT opening in 3rd seat. As I have said before, opening a 10-12 1NT opposite a passed partner is not rational (suicidal is the description that I would use).

As for 4th seat, an argument can be made that it makes some sense, since both opponents have passed. But I would not wait around for 3 passes to a balanced 10-12 HCP hand. You will be waiting a long time.

I've been opening 1nt 10-13 in 3rd seat or 1nt 10-12 in 3rd seat (any colors) for over a year now and can assure you in practice that while it is a little swingy it leads to many more opponent disasters than our side disasters.

When we are red, our 3rd seat 10-12 nt are opposite partners who have less than 10 if unbalanced and less than 12 if balanced. When white, our 3rd seat 10-12 nt are opposite partners who have less than 10 always.

I agree it is possible that against a sufficiently good class of players this is unsound and losing. I disagree that this is clearly the case because the side effect of exposing yourself to the odd phone number is countered by exposing your opponents to a preempt over which it is quite hard to find the right constructive spot. But on the next level if they do end up in the right spot they do have an advantage in the declaring based on what information you've exposed. It is a pretty complex system to analyze.

Also, in general, we do open most 5M332 hands nt as well as 5m332 and 5m422 and most 6m322. So it does come up quite a lot.

[gerry]

Cascade, on Jul 15 2009, 06:27 PM, said:

awm, on Jul 16 2009, 10:55 AM, said:

Similarly to what Cascade said...

It used to be that a 1♣ opening showed at least four clubs back in the days of Culbertson and Goren and old-fashioned Acol.

Since the advent of five-card major systems, more and more of us have been opening 1♣ with only three cards, and this is now quite common in the context of "natural" bidding.

In the last decade or so, it has become increasingly popular for a 1♣ opening to promise only two clubs such that the 1♦ opening can promise an unbalanced hand. Some people even take this so far as to open 1♣ on singleton so that 1♦ can promise five, and this approach is no longer extremely unusual even amongst otherwise "natural" bidders.

So obviously 1♣ opening which could be 4441 (i.e. any hand without five cards in any other suit) is just a natural bid right? Shouldn't be regulated, shouldn't be alerted?

Yet Cascade himself is one of the chief opponents of this position. Why is his position so different about 1NT openings?

Any standard which we define for "natural" bids will draw some arbitrary-seeming lines. And at some point some players will want to cross those lines. But as long as we believe that some regulation of methods is permissible, it will be necessary to have some definitions for what is allowed and to require that players who frequently violate those definitions stop doing so. In the case of 1NT openings, the "line" is "no singletons or voids."

I think you don't understand my position on the subject at all.

the whole point is that the existence of a 'bar' inevitably leads to inconsistencies; particularly because there is not one bar but many - every situation has its own.

The 1♣ short-BS fiasco saw anti-system types engaged in logical contortions to define one artificial bid as natural so as to outlaw other artificial bids. Wayne's initiation of discussion of 1nt-with-singleton was, it seems to me, designed at least in part to bring this into sharper focus.

Wayne, as I understand it, is not a 'chief opponent of this position' (the naturalness of a short 1♣). He, like myself and presumably others, doesn't give a rats which calls are defined as natural or artificial - we just ask for consistency. Of course the Alexanderian(is this a word?) way of achieving this would be to make the whole debate moot by removing the bar altogether; but I wont hold my breath waiting for another Alexander.

[mycroft]

Count me as another that plays 10-12 3rd NV to few ill effects.

Might as well say "open 2S with 6-10 in 3rd, given that partner has <= 10" is suicidal. It is likely to go minus, too (granted, less likely than 1NT, because at least you have that fifth (and possibly sixth) spade). And if it's wrong, there's even less chance that partner will be able to rescue you into something sane, as it's on the *3* level. People still do it, though, because ... it works.

I won't play any rescue system that has "pass forces XX". Especially with a third-seat 10-12 :-). At favourable, go ahead and double. I'll try to get out for the same 4 tricks we're making if you play NT... At both NV, it's a little scarier, but still. Combine that with three suited bailouts and a "takeout double of clubs" if fourth hand doesn't double (what, you don't believe in transfers, either, Mycroft? Heresy!), and, a rescue system which, say, gives a chance to get out in any single suit, any pair of suits, or opener having a 5-carder if she does, there is a very good chance that you will land in your best place to play, or at least in a better place to play than they would get if they didn't defend, or if you hadn't taken their system away.

If that "very good" chance is > 50%, you win at matchpoints, remember, no matter how many phone numbers you get the rest of the time. At IMPs, of course, it's a little riskier.

[ArtK78]

Comparing a 10-12 1NT opening in 3rd seat to a weak 2 bid in 3rd seat is nonsensical.

[mycroft]

Why? Because one usually works and the other ... seems to usually work?

[TimG]

TimG, on Jul 18 2009, 09:59 AM, said:

I generated 10,000,000 deals and counted how many times south would get to open a mini (10-12), weak (12-14), or strong (15-17) NT in 1st, 2nd, 3rd, or 4th seat.

 [space] [space] [space] [space] [space] [space]1st [space] [space] [space] 2nd [space] [space][space] [space]3rd [space] [space] [space] 4th
mini [space]	1,028,996 [space]	622,642 [space] 318,994 [space] 103,323
weak [space] [space]	809,465 [space] 535,865 [space] 317,150 [space] 146,547
strong [space]	407,163 [space] 303,392 [space] 213,693 [space] 137,097

Notrump openings were 4333, 4432 or 5m332.  Opening bids were 12+ HCP with any shape or 6+ HCP with 6+ spades, hearts or diamonds or 7+ clubs.  Even in 3rd seat.  In other words, everything was very loosely defined, but I think this gives a general idea of what you were looking for.

Spoiler

Initial random seed 1247927503

It occurs to me that for 3rd (and 4th) seat mini-NTs, I did not include the possibility that partner would open a 10-11 point balanced hand in 1st or 2nd. So, the opportunity to open a mini-NT in 3rd (and 4th seat) is over-stated in my table.