Keri

[blackshoe]

I'm currently reading Ron Klinger's book Bid Better, Much Better, which lays out this system of responses to 1NT. So far, it's pretty interesting, although I can see great difficulty in trying to get partner to give up Stayman. Aside from that, I think I read somewhere that Klinger himself no longer plays it. So I have two questions: does anyone have experience with Keri and would like to comment? And does anyone know whether Klinger in fact no longer plays it and if so, why not?

[Phil]

Lots of experience with old Keri. Last time we discussed it here, Klinger was playing a newer version.

 

You can find a lot of old posts where people discuss the pros and cons.

[awm]

I played Keri and modifications thereof for a pretty long time, and still do with some partners. My observations (and reasons for quitting Keri in at least one partnership):

 

(1) The three-level splinters are a really good treatment, and help both in finding light slams and avoiding bad 3NT contracts.

 

(2) The two-level diamond signoff can be helpful, but losing "garbage stayman" is also significant. I feel like these are a wash at IMP scoring, but because majors outscore minors the "garbage stayman" approach might be a bit more useful at MP.

 

(3) Keri leads you to play a lot of 4-3 major suit fits in 2M instead of 2NT. This is made out to be a win in Ron Klinger's writeup, but I haven't found it to be so at the table. One issue is that the "three" hand is always balanced; 4-3 fits where the 3 hand is 4333 are almost always worse than 2NT for example. Of course, it's possible that I'm not playing these 4-3 major fits as well as I'm supposed to.

 

(4) You do occasionally win by playing a 5-3 major fit in 2M instead of 3M. However, I would rarely pass the invitational 2M when holding 4-card support because of the possibility of a nine-card fit; most 4-4 major fits end up in 3M anyway.

 

(5) Keri is quite bad at finding minor-suit slams when both hands are balanced. This is obviously more of an issue playing strong notrump than weak notrump.

 

In all my view was that the splinters are great but easy to append to a non-Keri notrump structure. The rest of the structure seems not so good to me (worse opposite stronger 1NT ranges, better opposite weaker ones due to the slam issue and some right-siding issues on invites).

[jdonn]

I had a small amount of experience playing it and a medium amount playing against it. My experience was almost exclusively that the bid showing invitational with 4 or 5 in the major worked so badly that essentially it couldn't be offset and nothing else mattered.

[MickyB]

http://dcrcbridge.blogspot.com/2008/04/stayman-v-keri.html

[blackshoe]

thanks for the info. I'll do a search here later.

 

Sounds like a book I can put on the shelf as "interesting, but..." :)

[the hog]

I played it for ages and discussed it a lot with Ron. My experience is that it is far better than Stayman. I disagree totally that the bid which shows an invit hand with 4/5 cards worked badly - quite the opposite in fact.

Rdk plays with Nagy or Hoffman in the seniors and Nagy likes traditional methods. With Hoffman he plays 5 card stayman.

He plays a newer version with others and "no", I don't have the notes.

[NickRW]

I play a version of it - more similar to the one here than to the version in the book. I thought Keri was weak on some big ♦ hands and this version seems better on that point.

 

On the matter of whether Keri is good or bad with the invitational major sequences, I've seen opinions both ways. I looked at a sample of some 200 hands where I tried to work out whether we would have been better off or not - frankly it was close - and you could easily have a long run of hands where it looks bad - and a long run of hands where it looks good. One thing I observed - that it was often bad passing partners 2M with a min and 3 card support if 4333 shape - Klinger makes this point in relation to some game sequences, but I don't recall the book mentioning it about the invites.

 

Keri isn't devoid of stayman. 1N-2♣-2♦-2N introduces a staymanic sequence, albeit GF.

 

I've played 2♣ puppets 2♦ over a weak NT a good deal longer than I've played Keri as such - the ability to stop in 2♦ is a plus - but be careful at MP to only use when truly weak - if you use it when only somewhat weak and ♦ looks safer, too often 2♦ rolls in with an over and 1N also does the same so you can lose with it if you don't apply it sensibly.

 

One downside is that it is quite a lot of hard work to learn - and not many people play it, so it is really for the more serious partnership. You do score the odd good plus when you can find 3N is better than your major fit that conventional methods miss - but on the negative side, again especially for the more serious partnership, some sequences leak information and good defenders are listening.

 

The splinters can be mega - when they come up - and if you play the version above you'll be able to use some extra splinter type sequences.

 

I personally don't miss garbage stayman - but then I was already using 2♦ over a weak NT as stayman - so had already ditched it.

 

Contrary to another poster, I think Keri is better over a strong NT. Frankly I have no time for either Keri or conventional methods over a weak NT.

 

Nick

[straube]

I really enjoyed reading Ron Klinger's book and it gave me a lot of really good ideas. Its use of splintering before fit has been found and its use of a form of Puppet Stayman (Game Probe) were especially helpful.

 

Having said that, I think Keri 2C is fundamentally flawed. My best evidence for this is the loss of sequences it causes.

 

Stayman allows for frequent use of 1N-2C, 2D and 1N-2C, 2H and 1N-2C, 2S sequences. After each of the three allowed rebids, responder can react to opener's bid by making an invitation or splintering, making a Smolen jump or introducing a minor. Keri loses all of this because it is a puppet and not an ask. It's not efficient.

 

The best way is to combine Stayman with the best ideas of Klinger. My impression is that RM Precision does this sort of thing (though I think their NT structure predates Klinger's book) and I've written a structure myself that does this.

[the hog]

snipped

Stayman allows for frequent use of 1N-2C, 2D and 1N-2C, 2H and 1N-2C, 2S sequences. After each of the three allowed rebids, responder can react to opener's bid by making an invitation or splintering, making a Smolen jump or introducing a minor. Keri loses all of this because it is a puppet and not an ask. It's not efficient.

snipped

Keri loses nothing of this at all. You have your splinters; you have your ways of showing 5/4 major suit hands; you have your invitational bids; you can introduce minors. In fact EVERY ONE of the things you said are "lost" are not lost at all.

[awm]

There is one observation about Keri losing something via the puppet that seems to make sense to me. It's occasionally useful to have a shape relay available; the particular situation I have in mind is two balanced hands looking for slam in a 4-4 or 4-5 minor suit fit.

 

Playing stayman, you can bid 1NT-2♣-rebid and then use 3♣ as a game forcing relay to find all 4-card or longer suit(s) in opener's hand below the level of 3NT. The simple scheme I use for opener's rebids is: (1) 3♦ shows exactly one four-card minor, after which 3♥/3♠ inquires about clubs/diams respectively (2) 3M after a major suit response to stayman is natural, showing 4-4 majors or a 5-card major respectively (3) 3M after a 2♦ response to stayman shows five cards in the respective minor (4) 3NT denies any other bid, so (32)44 after 2♦ to stayman, or (34)33 after a major suit rebid.

 

There doesn't seem to be any way to insert such a relay into Keri. In principle you can investigate via 1N-2♣-2♦-2N (this is what it says to do in the book) but at best you are finding minor suit fits at the four-level. Trying to make 2♠ a direct relay doesn't work either (you're just too high to get all the information out).

 

Of course, the weaker the notrump the less likely such a slam on two balanced hands will be. But this does seem like an issue. The same relay can potentially be used to avoid 4M on a pair of 4333 hands (one of the advantages of Keri).

 

Obviously there is a big advantage to having in-depth notrump systems. Certain other aspects of Keri (like freeing up 2N after a transfer for GF hands, or the splinter bids) are definitely good. But it's the 2♣...2♦ sequences that really seem to define Keri as a method. If you believe the 4-3 major suit fit in 2M is a substantial win then Keri is for you, but like many other posters I haven't had great experiences with this (especially at MP scoring).

[MickyB]

There are lots of hands below GF strength with interest in both majors that Stayman handles much better than Keri, or any other method I have thought of.

 

4441

(43)51

4450

 

can bid Stayman, intending to pass 2♦, or 2M in a 4-3 fit. If they find a 4-4 fit they can pass if weak or raise on hands short of being "invitational" if a fit was not found.

 

Hands with 5 hearts and 4 spades can bid Stayman, looking for a 4-4 or 5-4 fit, and settle for playing in a safe 2♥ contract if one is not found. Again, this can be done on weak hands just looking for the best part-score, or constructive hands intending to bid game if a good fit exists.

 

Invitational hands with five spades, with or without four hearts, can bid 2♣ then 2♠ in my preferred methods. The knowledge of the fifth spade is crucial - opener will pass most hands with a doubleton spade and raise on a reasonable proportion with three spades. In Keri, you are having to bid on any hand with a doubleton spade, leading to less safe part-score and poorer evaluation for bidding game.

[doclands]

We've played Keri for several years and find it much preferable to Stayman etc. Those passing after 1N - 2♣ - 2♦ - 2M should be bidding 2N more often and only passing suitable hands - the idea isn't to pass whenever you are minimum and have 3cd support! We even play the Keri 2♣ sequences after a 1N rebid.

 

Lots of pluses and few minuses. It does help that neither of us plays with any other partner in real life!

 

Cheers

 

doclands

[straube]

snipped

Stayman allows for frequent use of 1N-2C, 2D and 1N-2C, 2H and 1N-2C, 2S sequences.  After each of the three allowed rebids, responder can react to opener's bid by making an invitation or splintering, making a Smolen jump or introducing a minor.  Keri loses all of this because it is a puppet and not an ask.  It's not efficient.

snipped

Keri loses nothing of this at all. You have your splinters; you have your ways of showing 5/4 major suit hands; you have your invitational bids; you can introduce minors. In fact EVERY ONE of the things you said are "lost" are not lost at all.

Keri loses sequences.

 

You can see some evidence in things it loses here and there. For instance, it can't splinter all hands. I complimented the immediate splinters, but one can't splinter

4-1-5-3 or 1-4-6-2 or 5-4-3-1. It can't splinter 4-3-5-1 or 3-4-5-1. Etc.

 

It's imprecise with its 5M/4M GIs. One transfers to hearts and rebids 2S to show either one. With a Stayman system, I can get to the best strain.

 

One loses the ability to show 4M/4M GI hands. One makes a GI in hearts and only sometimes can a superior spade fit (4/4) be found over a 4/3 heart fit. Granted one stays at the 2-level.

 

Again, my point is really a mathematical one. Keri loses sequences and that has to cost somewhere.

 

Separate issue...if you read Klinger's book it's filled with very biased example hands where a 4/3 major suit fit is superior to a NT contract. Typically the 4cd suit is beefy and the 3 cd suit has ruffing power. That's not necessarily the case in real life.

[hanp]

No Straube, take it from people who have played the method for ages: KERI DOES NOT LOSE SEQUENCES AND IT DOES NOT HAVE DISADVANTAGES!! IT IS FAR BETTER THAN REGULAR STAYMAN!

 

Now, will you stop arguing with arguments?

[straube]

No Straube, take it from people who have played the method for ages: KERI DOES NOT LOSE SEQUENCES AND IT DOES NOT HAVE DISADVANTAGES!! IT IS FAR BETTER THAN REGULAR STAYMAN!

 

Now, will you stop arguing with arguments?

How can you say that Keri doesn't lose sequences? Keri has no 1N-2C, 2H or 1N-2C,2S sequences. That's a loss of two very important sequences.

 

Stayman allows for these sequences and it allows responder to react to what opener has shown.

 

awm gave an example of this...1N-2C, 2D-3C as a relay (ask) searching for 4-cd suits. Whether that's best use or not of the sequence, I don't know, but Meckwell uses a similar sequence as a puppet to 3D so that responder can show shortness.

[gwnn]

rule number 1 in efficient forum reading:

 

hanp does not always mean literally what he writes.

[mikegill]

As someone who still plays the system that was linked to (with a few small changes) and has played it for 5+ years now, I have some observations:

 

1) I am relatively certain that overall this has not been a huge win or a huge loss. I did not keep detailed notes about its success, so I will not try to guess based on my scattered memories whether or not it has been a small win or a small loss in that time.

 

2) Honestly if I were going to stop playing it, the primary reason would probably be that it sometimes sides contracts differently than they would be in standard and I would like to create fewer random swings.

 

3) Here are what I see are the primary advantages:

- I think the many splinters and exact shape-showing sequences are a win vs. standard NT systems. As has been mentioned, much of this can be incorporated into a stayman structure if you're willing to futz with it, but I think it fits cleaner into this system.

- Being able to sign off in 2♦ is nice, and I can remember wins from this.

- The convoluted stayman sequence does let you avoid 4M with 4333 opposite 4333

- Being able to handle light invitational hands with (4x)(6x) and (5x)(5x) is usually a win when it comes up.

- There are other sequences that seem like they should be good but they never seem to come up.

 

4) Here are the disadvantages:

- You give them a chance to double clubs and diamonds on normal GF stayman hands.

- You don't have garbage stayman (this probably about cancels the benefits of being able to play 2♦ imo).

- 2♠ is artificial (I would rather play 2NT nat inv all other things being equal)

- There is no way to show (4x)(6x) with a min GF (to invite slam you have to be willing to play 4N).

- Lack of smolen both sometimes wrongsides and makes it awkward to bid 4-5+ and 5+-4+ major suit hands. I can't remember this ever mattering but it's definitely not a pretty part of the system.

- Keri is more vulnerable to opponent bids than is stayman. It's generally obvious what to do when they bid over stayman, I don't find it so over Keri.

 

I see the 2♣ - 2♦ - 2M sequences as approximately break-even. I don't really ever remember +110 when the entire field is -100 in 3, but I also don't remember +110 losing to a long string of +120s. I wouldn't advertise this sequence as the reason to play this system, though.

[whereagles]

I'm currently reading Ron Klinger's book Bid Better, Much Better, which lays out this system of responses to 1NT. So far, it's pretty interesting, although I can see great difficulty in trying to get partner to give up Stayman. Aside from that, I think I read somewhere that Klinger himself no longer plays it. So I have two questions: does anyone have experience with Keri and would like to comment? And does anyone know whether Klinger in fact no longer plays it and if so, why not?

I played the book version for some time and while it was interesting and effective, it's more of a pro convention. It requires a lot of memory work and practice. The 2nd version (around this forum somewhere) is similar. Some improvements, but still hard to memorize.

 

Then Ron Klinger came up with his book "5-card major stayman", which is basically a revised and simplified version of keri that's much more playable and simple. Get this book if you want to move away from regular stayman into something equally simple and efective.

[gwnn]

I played 2♣ forces 2♦ and then 2M showed 4-5 cards, I didn't like it; when I was opener sometimes I wanted to show a maximum if partner had 5 cards but play on the 2 level if partner had 4. That's impossible, no matter how nice your scheme is afterwards; it is of course open to debate whether the above wish is reasonable, maybe and maybe not. Heeman looks like a lot of fun, I have never played it but it looks nifty. I think you should try it instead if you are interested in abandoning Stayman and transfers.

[NickRW]

Keri loses sequences.

 

You can see some evidence in things it loses here and there.  For instance, it can't splinter all hands.  I complimented the immediate splinters, but one can't splinter

4-1-5-3 or 1-4-6-2 or 5-4-3-1.  It can't splinter 4-3-5-1 or 3-4-5-1. Etc.

Geez. Yes it is different and inevitably with *any* change you will tend to lose some sequences that you used to be able to express easily - and you gain others. The acid test is whether your new method is better overall. With Keri versus more "normal" methods it is perhaps too close to call for the casual player - and the frequency of being able to play "normal" methods is almost certainly the deciding factor.

 

On the splinters thing, in the version I'm playing, I can splinter (if I choose):

33(07)

33(16)

(34)(15)

44(14)

(04)(45)

(14)44

(13)(45)

(03)55

(03)(46)

plus anything slammy with a 6 carder in both majors or minors assuming opener is at all interested, plus on some other sequences opener can find shortage principally a 4cM with longer clubs. That isn't everything it is true - but how many can most people do with "normal" methods?!

[hanp]

What do you mean with "normal" methods? If pair one plays an entire book of Keri stuff, and the other pair just got together and agreed to play Stayman and Jacoby transfers, well duh, of course the full Keri complex would be better.

 

If on the other hand we would compare the Keri pair with Meckwell, who as far as I know play Stayman but probably have detailed descriptions of more auctions than most of us., then I wouldn't bet against the Meckwell system.

 

To get a fair comparison you would need to compare Keri with a Stayman based approach of similar complexity.

[blackshoe]

Heh. Now the pro side comes out. :)

 

I found a description of Heeman. Looks much like Keri at first glance. I'll look closer later.

 

Thanks all, and especially to Mike Gill for the pros and cons lists.

 

I'll look into 5 card major stayman as well.

[NickRW]

What do you mean with "normal" methods? If pair one plays an entire book of Keri stuff, and the other pair just got together and agreed to play Stayman and Jacoby transfers, well duh, of course the full Keri complex would be better.

 

If on the other hand we would compare the Keri pair with Meckwell, who as far as I know play Stayman but probably have detailed descriptions of more auctions than most of us., then I wouldn't bet against the Meckwell system.

 

To get a fair comparison you would need to compare Keri with a Stayman based approach of similar complexity.

Perhaps I am unaware, but I don't know of a book published about Meckwell over 1NT, so I don't suppose I can comment on their specific methods.

 

I am not beating some big drum for Keri - I was trying to give a more balanced picture than that - but straube seemed a bit prejudiced against it.

 

Nick

[hanp]

You can pick any top level pair whose methods you like and read their names instead of meckwell. The point was that if you are comparing your elaborate methods with the methods of "most people", then surely you are going to win by a landslide. Most people can write their notrump methods on the back of a... how do you call one of those round cardboard things that you put your glass on? Usually with a beer brand logo on it.

 

By the way, did you think the Klinger fans posting here gave a balanced picture?