Friday, 18 May 2012

May it get warmer, please

Hi Chillians

The revised website is now up and running.

It meets what for me was the most pressing issue – my committment in the book to provide an up-do-date list of amendments and corrections.

It is also a reasonable if slightly terse summary of the system described in the book. There are some bits missing - most notably slam bids after the count - but it does include shunts in the Ideas section. It is also not rigorously checked for errors, so I am sure we will spot one or two of those.

There are a number of new ideas that I will be blogging about in the next few weeks. They include increasing the number of splinters, taking position more into account in overcalling, and the mysterious attitude asks.

More soon,
Best wishes
Alan

The Chilli bidding system is described and defined in the book What Can Possibly Go Wrong? and summarised on the Chilli bidding website.

9 comments:

Takis Pournaras said...

Great work, Alan!

The material published will most certainly help us quickly grasp the system's changes. Thank you

Alan Williams said...

Thanks Takis

Until I started I hadn't realised how much had changed. As you can see, I opted for more or less a complete rewrite!

Alan

Andrea said...

Hi Alan!

First of all, I'm glad because your long silence had me a little worried.

Second, great work as always!

Third, I'm amazed ad how articulate disturbed bids seem to be even without my "beloved" shunts (I think disturbed bids are always critical, and particularly so in any strong club system).

Obviously, I'm planning to buy your book! (Actually receiving it will be another matter...)

All the best,

Andrea

Anonymous said...

On the Fit Auction page on the web site ,the section Values-for-five and the Keycard Ask you have "4♣ is an Autofit agreeing clubs and showing Values-for-five . 4♥ is the Keycard Ask, the 5♣ response showing two or five."

It should read "4♦ is an Autofit agreeing clubs ..."

In the section Style of Fit bidding there is a comment "going to game immediately opposite 2NT would have
been an error" but the third example example of the Values-for-five section has the action start 1S-2N, 4S.

I think this isn't the contradiction it might first appear. It seems reasonable to bid 4S here with a minimum and 6 spades but I wonder if 4C and 4D are really best used as values for 5 and keycard by a limited opening hand? It seems like splinters here might work better.

1S-2N
3C/D/H = long suit try, values in suit.
3S = minimum, probably balanced.
3N = 14/15 HCP, to play.
4C/D/H = splinter, probably 6 card spade suit
4S = 6 spades, to play.

If opener is wildly distributional and wants to use key card they still can by first making a long suit try or splinter. As a plus they may get more useful information first.

Michael

Alan Williams said...

Hi Michael

Thanks for the correction, now fixed.

There is no contradiction in the two examples, but I should rephrase one or both of them to make it clear what I intended. In the first example, opener has a useful suit to trial; in the second, not.

To the substance of your idea: I analysed something very similar a while back, with the intention of making all side-suit bids above three of the trump suit be splinters when the bidding was below three of the trump suit.

There seem to be two problems with this. The major one is the need to weaken the meaning of trial bids. If they are only natural waiting bids, then we will end up bidding them with four small or AKQJxx or anything in between, at which point they lose most of their value as fit evaluation tools.

The lesser problem is the usual heart-spade issue: if the auction starts 1H - 2NT, you have to decide if 3S is a splinter or a trial bid.

Your suggestion - which restricts such activity to limited opener - is clearly better.

The problem for me is that, although one of a major is a limited opening, its playing strength varies by up three tricks.

It was for that very reason it took me a long time to accept that we should take away limited opener's use of the minor relays and instead use them to show 5+ - 4+ hands. That was a good decision in that the new forms get far more use.

Would that be true of the splinter idea? My worry is that - although opener's use of the keycard ask is rare indeed (and probably, as you say, replaceable), in practice 1M - 2NT; 4C showing values-for-five is a quite common sequence.

Consider this hand for opener: Kxxx xx AKQxxx x. The auction starts 1S - 2NT and now what?

(a) 3S and 4S are both underbids
(b) 3D creates the devaluation of the trial bid I mentioned above
(c) 4C (were it a splinter) would break (or abolish) the 'controls in the other two side-suits' rule.

As it is, it is a clearcut 4C values-for-five rebid.

x Kxxx AKQxxx xx starting 1H - 2NT would be even trickier.

Your schedule as laid out is logical and consistent, but the trial bid would definitely become devalued, and that's something which would way heavily against the idea for me because bidding (or not bidding) tight games depending on the quality of fit in a side-suit is a more common occurrence than bidding a slam.

Alan

Anonymous said...

I see the point about not wanting to devalue the long suit trials but I don't think it is a real problem. The sort of hand you describe is always going to game, the tight games bid or not bid based the presence on a secondary fit will still be bid the way they are now. Responder assumes the usual meaning and bids accordingly.

After 1H-2N, 3D-3H - no diamond help and minimum, opener can pass with the traditional game try and bid 4H with the values for game and a strong diamond suit. When opener does have the solid diamond suit any bid other than 3H must show the high card strength for game or slam.

A greater concern for me is holding xx KTxx AKQxxx x where the aution starts 1H-2N,3D-4H, now if opener tries 4N we go off opposite xxx AQxx Jx KQJx but are a good shot for 6S opposite Axx AQxx xxx xxx and I don't even know if this last hand would bid 4H.

I agree that 3S after 1H-2N is a bit tricky but what ever it means it isn't a game try, that boat has been burnt.

We have identified two hand types that can realistically suggest slam immediately opposite a minimum 2N response, KQx AKxxxx x Kxx or x Kxxx xx AKQxxx. Either opposite 3 bare aces has a good chance at slam.

If an immediate key card is not needed by a limited opener then there are 3 bids available to show values for 5 (assuming 3S). If these bids show a solid 6 card suit then 1S-2N, 4H has no meaning. I know that is not sufficient reason in chilli for rejecting a set of responses.

I guess the trick is to find a way consistent with general chilli rules and that cover the hands most likely to come up at the table. These are fairly rare hand types so there's no pressing need to have them as part of the basic system, perhaps an optional extra or replacement for the standard chilli methods.

Looking forward to your blog on more splinters in the system.

Michael

Anonymous said...

Hi Alan,

Some questions about auctions that start 1D-1H/S, 2C/D.

Do these always show singleton in the bid major? It would increase the frequency if they can be bid with 5422 shapes.

How about x AQ JTxxxx AQJx? 3D looks too high on such an anaemic suit. 2D describes the hand fairly well.

What does 2NT mean? Responder is unlimited. Perhaps a good raise of one of the minors? 3C = min, 3D+ = max.

1D - 1S
2D
- 2N good raise
- > 3C min, now P or correct to 3D
- > 3D+ max
- 3C/3D law raise.
- 4C/4D forcing raise.

Michael

Anonymous said...

OK, got it now.

2C/2D here are raw and therefore disturbing. 2NT is inv+.

Regards,
Michael

Alan Williams said...

Hi Michael

There's no reason for the 2C/2D rebid to be restricted to hands with a singleton in the bid major (and on page 58 there's an example of rebidding 1NT with such a hand). It was certainly intended as an option for 5422s, and mandatory for hands with a singleton in the other major.

With x AQ JTxxxx AQJx I might settle for 2D (initially) opposite a 1H response, but opposite 1S I would always choose an immediate 3D to pre-empt their heart fit.

You're right that the 2C/2D rebids are raw and therefore disturbing. However they are also natural suit bids, so 2NT is Strong Fit in the bid suit (as per normal disturbed rules). And since a three-level bid of the other minor is forcing (again normal disturbed rules), that is Strong Fit in the other minor.

So how, you may ask, does responder sign off in the other minor? If the rebid is 2C, no problem - just bid 2D. If the rebid is 2D, however, minimum responder may possibly be stuck for a rebid.

That's the price of sticking to the standard rules, but it doesn't seem a very high one - in thousands of hands played, this sequence has never come unstuck - it is fairly rare anyway, the hands on which responder will struggle are a small subset of these, and if pass is the best responder can do then the opponents may well rescue us with a bid of the other major or a double.