[EricK]

whereagles, on Nov 26 2004, 10:13 AM, said:

I'm not convinced keri super-accepts are superior to other methods. I have the impression the "magic 9-trick-only 3NT game" happens less frequently than the "magic no-wastage 4M game", so there might be a case for changing that bit of the system.

Remember that in Keri, most invitational hands will go via 2♣ not via a transfer

Eric

[whereagles]

I know. The point of those super-accepts is to bid and make a game when responder was intending to pass the transfer, and not to invite. It is under these conditions that I think a 20-22 point game of the "fitting shortness" kind is more likely than the 20-22 point game of the "long running major" kind.

[EricK]

Even for those hands, you will need quick tricks outside, else you almost always have 4 losers.

Eric

[Flame]

I dont like too many stoper checkings, no point in telling the opponents exactly how to diffend. I think this system will help especially when responder got a long suit. you will know you have the suit running and quick tricks outside.
On more balance hands maybe you need just the opposite to want to be in 3nt, more solf points. with bad trumps and soft tricks 3nt might be better then 4M.

[tlgoodwin]

A question about Revised Keri, as described in outlines in this thread. The sequence 1NT - 2C ; 2NT/3C - 3M is defined as forcing to game with 5+ in M. But I do not see that such a hand responds 2C in the first place (as opposed to a transfer to the major). So, isn't the definition given for that 3M rebid "impossible"?
Maybe it is meant that the sequence shows a forcing hand with 5+ diamonds and 4 in the major (i.e., what 3M would mean if opener rebid 2D instead of 2NT/3C).

Can anyone clarify this?

TLGoodwin

[pclayton]

tlgoodwin, on Nov 26 2004, 05:58 PM, said:

A question about Revised Keri, as described in outlines in this thread. The sequence 1NT - 2C ; 2NT/3C - 3M is defined as forcing to game with 5+ in M. But I do not see that such a hand responds 2C in the first place (as opposed to a transfer to the major). So, isn't the definition given for that 3M rebid "impossible"?
Maybe it is meant that the sequence shows a forcing hand with 5+ diamonds and 4 in the major (i.e., what 3M would mean if opener rebid 2D instead of 2NT/3C).

Can anyone clarify this?

TLGoodwin

The 'fast' Keri sequence (1N - transfer - accept - 3N) shows a 5332 and a WEAK suit (Qxxxx) or less.

The slow Keri sequence: 1N - 2♣ - 2♦ - 2N - 3♣ / 3♦ - 3 major shows the better suit.

[tlgoodwin]

That is correct for "book" Keri: 1NT-2C-2D-2NT is forcing to game, and responder might be on his way to revealing a five-card major. But in "revised" Keri (as promulgated in this thread), 1NT-2C-2D-2NT is just an invitational raise of notrump, with no interest in a major; it isn't forcing. I do not believe there is in "revised" Keri any strong hand with 5+ in a major that responds 2C, so 3M by responder cannot mean that after opener bids 2NT/3C to show a six-card minor. Responder does bid 2C with a strong hand containing five or more diamonds and a secondary four-card major, intending to rebid 3M over opener's 2D. Perhaps 3M by responder still means that. (It would be improbable for responder to have five diamonds when opener has six of them, in the sequence 1NT-2C-2NT-3M; but one must not confuse "improbable" with "impossible." The 5D + 4M meaning is of course more probable if opener shows six clubs.) There won't be a 4-4 fit in the major, but it may still be valuable to describe the general hand type.

As I have suggested elsewhere in BBO forum, I am not at all convinced that there is merit in opener showing a six-card minor (and a minimum) directly over 2C. I am dropping it from my partnership agreement (or rather, I am not implementing it at all, since I never did include it). I do like opener's new 2M rebids over 2C, though, and those have now made it into my Notes.

TLGoodwin

[tlgoodwin]

There is (at least) one more glitch in Revised Keri, as set out in this thread. When it goes 1NT - 2C ; 2D - 2H (invitational, 4-5 hearts, fewer spades) and opener has a maximum, the outline says opener bids: 3D, 2H + 5S; 3H, 3H + <5S; 3S, 2H + 4S; 3NT, 2H + 3S. Thus, if opener has only two hearts, he can show three spades (3NT), four spades (3S), or five spades (3D). Left unsaid is how opener shows two hearts and two spades, but that is easily included in 3NT, since responder will not have five spades and won't care whether opener has two or three.

The situation is worse if opener has three hearts. The only rebid in the outline for that, 3H, shows fewer than five spades. Accordingly, if opener has, say, KJxxx Axx KQx Kx, he is out of luck if responder bids 2D, then 2H. Responder might have three spades and four hearts, but there doesn't seem to be any way to determine that (according to the outline).

This is easily fixed. After 1NT - 2C ; 2D - 2H, let opener (with a maximum) bid 3D with only two hearts and 4-5 spades: there is plenty of room for responder to check back for five spades after that. Opener bids 3H over 2H with three hearts and fewer than five spades; 3S over 2H with three hearts and five spades; and 3NT over 2H with two hearts and 2-3 spades.

TLGoodwin

[tlgoodwin]

In the next-to-last paragraph of that last post, it should say "2C, then 2H," not "2D, then 2H." Responder would have a hard time bidding 2D, then 2H, without a friendly double from the opponents . . . .

[whereagles]

I had another look at revised keri.

It seems simpler and more mnemonic than the book version. For instance, the minor suit slam investigations all go via 2S now. Also, apparently one now transfers to the major regardless of suit quality, whereas before a bad 5 major suit would transfer and a good one would go via the 2C gadget. Much of the new stuff seems geared towards dealing with 1NT opening on a 5-card major.

I'm not convinced this revised version is superior to the book one, but it has some good new ideas. I need to dissect it even more lol.

[whereagles]

Ok, I have a question or two.

1. Why is it that

1NT 2C
2D 2NT

is now an invitation, instead of the original forcing relay? Doesn't this defeat the one of the purposes of the original keri, which was to know when to play 3NT instead of 4M? I mean you could bid a strong major suit (HHxxx) via the 2NT relay, to find out if opener had a 4333, so that you could play 3NT in that case. With a weakish major (Hxxxx or worse), you'd transfer and bid 3NT, so that opener would decide to pass or correct, in full knowledge responder had a bad suit. You can't do this trick anymore in the new keri.

2. Dropping the 2NT relay also overloaded the 1NT-2S sequence. Opener can no longer say if he's min or max, and the 1-suiter hands have to go beyond the 3-level more often to ask for aces.

All this gives me the impression that revised keri is more like "beginner's keri", given it's simpler to memorize than the book version, but also less precise. Is this impression correct?

[pclayton]

When I first read about Keri two years ago, I thought it was really cimbersome.

After playing it for awhile, the logic of the structure was very sound. The idea about a 4432 playing opposite a 4333 played in the 4-4 major, but the 5332 opposite the 4333 played in 3N.

We like the 'xyz' concept about playing the invitational 4 or 5 card major at the 2 level.

The exploratory sequences after 1N - 2♣ - 2♦ - 2N are intriguing, and finding the 4-4 minor fit at the 4 level and knowing Opener's exact shape (by inference, not by relay) and kicking off an 'optional' rkc sequence is draining, but effective.

Now, just about all of the above points have been discarded in favor of Opener having the option of immediately showing a 5 card major or six card minor. The 2♠ response (formerly) range ask is now overloaded with single suited slam tries and "natural" forcing stayman sequences. The delayed 2N structure is thrown in the round-file.

About the only recognizable sequences are the jacoby transfers and 3 level transfer splinters, although the transfer sequences seem to be toned down somewhat.

If I had seen this method of extended stayman prior to playing Keri stayman before I knew Keri, I would have preferred the new method, because of its simplicity. But after playing Keri for awhile, I think I still prefer the old structure, not the newer 'homogenized' method.

[pclayton]

The six card minor in a weak NT scenario is an interesting one.

I was playing in a club swiss game with my regular pard and Chris Larsen was our teammate. In the final match we were in 2nd place playing the first place team and about 3 VPs behind them. My pard - 2nd chair, white on red, opens a 10-13 NT with a hand that looked more or less like: KJ, AQx, Jxxxxx, Qx. Not a big deal, but they ran off the first SIX club tricks for -50. At the other table, our partners competed with clubs, but our counterparts settled in a quiet diamond partial making +90 or 110. We lost the match by about 5 IMPs, and Chris didn't think my partner's opening was very well-judged. I defended my pard saying that our requirements about opening 1N with a 6 card minor were lax.

Fast forward to three weeks later. Chris approached us and says, "You are playing IMPs, and pick up: xxx, KJx, AKxx, Jxx. Pard opens 1N which is 10-13. Whats your call"?

We both in unison said, "Pass" - "A flat 12 isn't worth a game try opposite a 10-13 NT". We had completely forgotten about the other hand, which he then broke out.

"This is why its a bad idea to open with a 6 card minor 1N".

1♦ followed by 2♦ is a perfectly good decription of this hand which leads to a painless 3N. Chris played a 10-12 NT with his wife Kay long before it was fashionable to do so on the west coast.

"Vintage Keri" systemically made a 1N opening very difficult with a 6 card minor. I'm not convinced its a good idea to revamp the system to allow for this, although there are other reasons why Klinger changed it around of course.

[EricK]

pclayton, on Dec 7 2004, 06:04 PM, said:

The six card minor in a weak NT scenario is an interesting one.

I was playing in a club swiss game with my regular pard and Chris Larsen was our teammate. In the final match we were in 2nd place playing the first place team and about 3 VPs behind them. My pard - 2nd chair, white on red, opens a 10-13 NT with a hand that looked more or less like: KJ, AQx, Jxxxxx, Qx. Not a big deal, but they ran off the first SIX club tricks for -50. At the other table, our partners competed with clubs, but our counterparts settled in a quiet diamond partial making +90 or 110. We lost the match by about 5 IMPs, and Chris didn't think my partner's opening was very well-judged. I defended my pard saying that our requirements about opening 1N with a 6 card minor were lax.

Fast forward to three weeks later. Chris approached us and says, "You are playing IMPs, and pick up: xxx, KJx, AKxx, Jxx. Pard opens 1N which is 10-13. Whats your call"?

We both in unison said, "Pass" - "A flat 12 isn't worth a game try opposite a 10-13 NT". We had completely forgotten about the other hand, which he then broke out.

"This is why its a bad idea to open with a 6 card minor 1N".

1♦ followed by 2♦ is a perfectly good decription of this hand which leads to a painless 3N. Chris played a 10-12 NT with his wife Kay long before it was fashionable to do so on the west coast.

"Vintage Keri" systemically made a 1N opening very difficult with a 6 card minor. I'm not convinced its a good idea to revamp the system to allow for this, although there are other reasons why Klinger changed it around of course.

I think if you are going to allow offshape hands (eg 6 card minor or 5422) into NT then you shouldn't use them with maximum hands. The extra shape adds too much playing strength. I would be happy opening a 10 or 11 point hand with these shapes if playing 10-13 but less happy if it had 12 points and wouldn't do it if it had 13 points.

Eric

[david_c]

whereagles, on Dec 7 2004, 11:48 AM, said:

Ok, I have a question or two.

1. Why is it that

1NT 2C
2D 2NT

is now an invitation, instead of the original forcing relay? Doesn't this defeat the one of the purposes of the original keri, which was to know when to play 3NT instead of 4M? I mean you could bid a strong major suit (HHxxx) via the 2NT relay, to find out if opener had a 4333, so that you could play 3NT in that case. With a weakish major (Hxxxx or worse), you'd transfer and bid 3NT, so that opener would decide to pass or correct, in full knowledge responder had a bad suit. You can't do this trick anymore in the new keri.

2. Dropping the 2NT relay also overloaded the 1NT-2S sequence. Opener can no longer say if he's min or max, and the 1-suiter hands have to go beyond the 3-level more often to ask for aces.

All this gives me the impression that revised keri is more like "beginner's keri", given it's simpler to memorize than the book version, but also less precise. Is this impression correct?

Because opener bids 2M over 1NT:2C with a 5-card suit and a minimum, it makes sense to start with 2C on balanced invitational hands, so that you can play 2M rather than 2NT when opener has a 5-card suit. So 1NT:2C,2D:2NT is now non-forcing. That does mean you have a little less space overall, so this is not a change you'd want to make if you don't open 1NT with a 5-card major.

There are still two ways of bidding GF 5M332 hands. With a strong major, you use the new game-forcing stayman. With a weak major you transfer then bid 2NT (unlike in the book version, this doesn't promise a 4-card minor); if partner now shows 3-card support there's still room to investigate playing in 3NT.

The nicest thing about the revised version of Keri is using transfer-then-3NT to show a 4-card suit and a 4432 hand (one 4-card major and one 4-card minor). This is worth playing even if you're not convinced by the other changes, as it gives away as little information as possible about opener's hand.

[MickyB]

Bump!

I was wondering what people make of revised Keri a few months on.

David_c suggests just adding the new transfer sequences to book Keri, which seems sensible to me. What about also making 1N:2♣, 2♦:2N a puppet to 3♣, then making bids mean the same as they do after 1N:2♠, 2N in revised Keri?