[Fluffy]

I remember long ago TylerE showed us a lovelly table with distributions most likelly to make game.

Now I am toying with a relay system, I feel like I need a bit of extra space for my slam bidding here and there, I play a system that lets someone know partner's exact distribution below 3NT for most non freak hands. But some of the hands let you know the distribution at 3♣ while others let you at 3NT, and it is a big difference.

I am thinking about moving something here and there and concentrate on hands that most likelly produce slam so that they show distribution a bit earlier. But at the same time are hands likelly to hold. First thing that comes to mind is 5431s, or perhaps hands with a void are better (althou if a hand witha void fits, it is very lielly you will get preempted).

I don't think this might be computed, but on general, what do people think are the most likelly hands to produce slam?. On my specific case there are 2 different main scenarios: 5 card major hands, and balanced hands that allow 5422's but with no 5 card major nor 6 card minor. 5422's might be shown as 2 suiters if they have weak doubltons so are a bit less likelly.

[Fluffy]

double post

[gwnn]

Sorry Fluffy for an unhelpful line of this post but it's spelled "likely" and "lovely".

I would think that the more extreme the distribution, the more likely slam is. So if you call 8311 a freak then the most likely one-suiter hand that makes slam and is still covered by your relays is 7321. If you call 8410 a freak then 7420 is the two-suited hand which makes slam most likely. I think this simple rule is actually true. However, I would advise against using bidding space like this. 5431 is the most likely two-suited distribution and it should be bid lowest. I think you often find a good fit and can make slam on 28-29 combined points even if there's no amazing distribution. That's where relay systems excel and it's nice to be able to ask for controls at a low level (3♦ or 3♥ are very low) and if relayer finds out about a high control count he will be happy to go on, if he sees that 3 controls are missing we can just play 4M or what not.

[Fluffy]

either I didn't explain well or you missunderstood, the likelyhood of you making slam when you hold 8401 is good, but the likelyhood of you having 8401 is very low compared to other distributions, I am looking to maximice the % of having that distribution multiplied with % of making slam with it. So if you make slam 10x times with 8401 than 4333, but you hold 4333 35 times more often then I'd rather show 4333 than 8401.

thx about the correction, it was a lovely thing from you to do althou I am likely to still make it badly form time to time

[gwnn]

Oh right. I misunderstood. Interesting question, I hope we get an answer.

[kenrexford]

This is a topic that has interested me for some time. I cover a lot of this on my blog (www.cuebiddingatbridge.blogspot.com) and in my book on cuebidding.

For instance, the power of the 5431 hand suggests the use of a switch-focus call in game-try auctions because slam suddenly becomes possible, even though we are just looking for game initially (one assumes). Consider these two hands:

♠AQxxx ♥x ♦AQJx ♣Axx
♠Kxx ♥xxxx ♦Kxxx ♣Kx

After a 1♠ opening, Responder makes a simple raise to 2♠. Opener, with 17 HCP (not that wild) bids 3♦ as a natural game try (ostensibly). Using my focus-switch method, Responder bids 3NT to show superb support for the second suit (five-card would be even better) and a maximum. Opener can agree the focus-shift, and shift tacks to slam try, by bidding 4♣, which asks for more clarification. 12 tricks is easily made (assuming no wild surprises) on 26 HCP only, assuming the remaining questions can be answered, and assuming that the contract is 6♦ (shifted focus).

The same type of thinking goes into my "empathetic splinters" analysis, where on my blog I go into the twilight zone. But, I analyze a type of matrix where a two-fit scenario (typically featuring a 5431 and a 34xx or 35xx pairing) makes slam on as little as the low 20's in HCP count. The idea was not just to recognize the holdings but to formulate "rules" or principles in seeking these slams, by defining suits within this pairing matrix (like focal suit, trick suit, shortness control suit, and honor control suit).

I also thought about developing this more. But, you might enjoy looking for these spots in my book and/or on my blog. On the blog, look for "abandon all hope ye who enter here" for some really out there thoughts. Esoterica to the extreme.

[Cascade]

1000 hand double dummy simulation for hands that could make slam in any suit.

The frequencies for distributions were:

5431 172
4432 154
5332 105
5422 88
6421 66
4333 63
5521 50
6322 48
6331 46
4441 37
6430 32
5440 26
6520 23
5530 17
6511 13
6610 3
Total 943

Therefore 57 hands were with seven card or longer suits.

[Zelandakh]

Ken, your example just happens to contain xxxx opposite x, is it surprising that slam makes on 26hcp? If you play long-suit game tries then you can describe this hand perfectly with a 4H call over 2S. Now with a max and an ideal holding in the splinter suit Responder can make a move. As for the original question, I would imagine that the reasonably balanced hands with a singleton would come out on top - 5431, 6331, 6421, etc - without any analysis to back it up. Whether it is a good idea to design the relay system around this rather than more traditional grounds is another question entirely...

[gwnn]

Zel I think you're missing Ken's point. He was talking about 6♦, which is almost cold, as opposed to 6♠, which is almost down.

[Zelandakh]

You are right gwnn, I missed the "shifting focus" point, something which seems to happen more at the game level (shifting to the other major). That does not alter the point that the hands are carefully constructed to fit perfectly. Note also that using the more common help-suit game tries rather than long-suit game tries the opening hand would not bid 3D. In my own methods Responder would relay and find out about the 5143 shape. I would like to think I could find 6D after this, at least at IMPs and agree that it is (much) more difficult after a 2S response without some specific agreements.

[gwnn]

I know I wouldn't find it and think at least 95% of top pairs wouldn't either. It is a hand from a certain fairy-land, but it is one where Ken's methods work wonders.

[kenrexford]

Zelandakh, on 2011-June-05, 06:10, said:

Ken, your example just happens to contain xxxx opposite x, is it surprising that slam makes on 26hcp? If you play long-suit game tries then you can describe this hand perfectly with a 4H call over 2S. Now with a max and an ideal holding in the splinter suit Responder can make a move.

4♥ by Opener does not distinguish 5143 from 5134, which is extremely important. Let alone, 6133. Moreover, even if Opener is 5044, there is a wild difference between AQJx-Axxx and Axxx-AQJx, as you want internal slow honors and external fast honors.

Contrast this:

1♠-2♠
3♦-3NT(focal shift super-accept)
4♣(agreed, asking)-?

6♦ = four key cards (assuming 6KCB and counting both Queens immediately) -- KQKQ in the internal suits
4♥ = three key cards plus the heart King (two internal Queens and an internal King)
4♠ = only three key cards, no stiff or void anywhere
4♦ = three key cards, plus the club King (two internal Queens and an internal King)
5♣ = three key cards, plus stiff club
4NT = three key cards, plus stiff heart

You get more definition, and you get the confort of knowing that the alternative fit is 4-4 (or 5-4).

Zelandakh, on 2011-June-05, 06:49, said:

You are right gwnn, I missed the "shifting focus" point, something which seems to happen more at the game level (shifting to the other major). That does not alter the point that the hands are carefully constructed to fit perfectly. Note also that using the more common help-suit game tries rather than long-suit game tries the opening hand would not bid 3D. In my own methods Responder would relay and find out about the 5143 shape. I would like to think I could find 6D after this, at least at IMPs and agree that it is (much) more difficult after a 2S response without some specific agreements.

The hands were carefully constructed to fit perfectly to illustrate a principle. The fact that a perfect fit is often needed is why you need the 4♣ relay to unwind this.

BTW -- Responder's hand is not right for 3NT in my methods, anyway. He would actually need the spade Queen instead of the club King, which makes the call purer (and then Opener would have the club King instead of the spade Queen). However, the slam would still likely be bid because Responder would now simply raise to 4♦, showing a less pure raise, which would allow Opener to bid 4♥ (shortness), and NOW Responder can bid the DIAMOND slam, eventually. Ideally, Responder actually uses the 4♣ call to do this, which allows both shortness calls:

1♠-2♠
3♦-4♣(impure super-accept of diamonds]
4♥(shortness, agreed shift presumably)-...

Call 4♣ a "courtesy focus shift" if you want.