[hrothgar]

Few comments here, most of which are related to a central premise:
The analysis technique that you are choosing is badly flawed:

From my perspective, there are two main causes

1. You don't have enough examples to have anything remotely resembling a statistically valid sample. Its VERY easy to back this statement up:

A) Your analysis of the results of my 3C and 3D preempts show a slight net loss in IMPs. In contrast, my 3H opening bid is aveage +6.7 IMPs... It could be that my 3H opening is brilliant in practice, while my 3C/3D openings are slight suboptimal. However, its much more reasonable to assume that evaluating the efficiency of a 3H opening based on 6 observations is suspect...

B) You and Free are using quite similar methods but disagree radically regarding the relative frequency of the two hand types. Free's estimates that the strong hand type occurs 20-25%. Your experience is that its close to 33%. Here once again, its possible that Free has looser requirements for the single suited hand types, however, I suspect that its more reasonable to assume that there aren't enough data points available.

C) Your own posts state that you don't have enough data points to evaluate the results from the weak hand types.

2. I don't believe that looking at Board Results are an appropriate way to evaluate the efficiency of different bids. In addition to all the sample size issues that I raise, there is another very significant problem: Recall the results of the "Swatting ther MOSCITO's" match. From my perspective, the results of that match shed VERY little light regarding the efficiency of the different bidding styles being used or the skill level of the different participants. To me, the match results demonstrated one point very clearly: Two well oiled partnerships have an enormous advantage competing against two pickup pairs...

In turn, this same issue significantly biased your study when you compare the efficiency of your Misery preempts to BBO or OKB board results.

There is a very simple way to sidestep these issues:

Monte Carlo simulation usig a good hand generator like Hans van Staverns's Dealer or Thomas Andrew's Deal is a much more powerful analytic technique than data mining.

[inquiry]

hrothgar, on Aug 9 2005, 12:09 AM, said:

Free's estimates that the strong hand type occurs 20-25%. Your experience is that its close to 33%.

There is a very simple way to sidestep these issues:

Monte Carlo simulation usig a good hand generator like Hans van Staverns's Dealer or Thomas Andrew's Deal is a much more powerful analytic technique than data mining.

Last comment on frequence. I said 1 to 4 to 1 to 3 from looking at 100's of thousands of deals using Bridgebrowser to pull the hands out. Free estimated his from his feelings based upon expereience. I am willing to accepted 1 to 4 (25%), no problem.

As for monte carlo simulations, awm has told you on at least three occassions he has done this and found the frequency around 1 strong hand for every 2 weak hands. I think he underestimates the frequency (and crap) people preempt with, but his study meets you requirement, and I see no need to re-invent the wheel.

Moving on know with the rules for when partner doesn't raise. We will start with no competition from the oppoents, and follow up with their bidding.

RULE 8. IF PARTNER BIDS HIS OWN SUIT, it is forcing. With a weak hand, opener may,
1) Raise partners suit with a fit, or
2) rebid his own suit.
With a strong hand, opener must rebid in the
3)"fourth suit" (not suit opened, not transfer suit)
4) in 3NT (two suiter without the bid suit
5) the "opened suit" or above a simple raise of partners forcing suit. Bids above "Raise" show their normal meaning (as if partner completed the transfer). 3NT shows the other two suit combination maximum losers, cheaper bid shows fewer losers and as soon as the bidding can back to normal meaning it does.

RULE 9. If responder REBIDS his suit, it intended to play (my suit), and is not a denial cue-bid or otherwise

RULE 10. If responder jumps in a new suit in response to the transfer opening, it shows a SOLID SUIT (no losers). That suit will be trumps, even if opener has strong two suiter.

RULE 11. Responder bids of 3NT are to play if opener is weak, opener rebids normally if holding strong two suiter. The 3NT reply can be made when very weak and extra long in partners suit, so if 3NT bidder then rebids the "transfer suit" over partners two suit showing bid, it is to play similar to rule 9.

RULE 12. IF RESPONDER JUMP RAISES, opener passes all weak hands, and bids on with strong hands. Three and four losers responses and some 2 loser responses are compressed into the cheapest responses, and a few adjustments are made. 4NT by responder now ask for number of losers instead of showing a cover. The assumption in the jump raise will include a cover in the bid suit more often than not, so when replies can show same losers at the same level, the cheaper shows no need for cover in the "raised suit" (worse news).

RULE 13. If the opponents double a denial cue-bid, or an ACE inquiry, or any other artificial bid once a two suiter has been shown, PASS and REDOUBLE become the cheapest two "no signoff bids". For instnace, if they double a denial cue-bid, pass ask for spade of off suit, REDBL shows missing queen in the lower anchor suit, next cheapest non=signoff bid shows missing queen in the higher suit. Etc.

RULE 14. IF opponents double the opened suit (regardless of meaning of the double), pass by responder shows a tolerance for the opened suit. A redouble show game invite in the "weak suit", and complete transfer shows total dislike for the double suit (this allows, 3C-x-p-p when opener has a two suiter with clubs, since responder has shown a tolerance, this allows 3D-x-xx-P-4H when opener has a good preempt). Over responder pass, redouble or completion of transfer, opener rebids as if no interference and LHO passed, except that opener completing his own transfer or redoubling shows weaker and stronger preempt hands.

RULE 15. IF opponents overcall, and partner doubles it is for penatly, to remove you need a strong hand, and you respond as if partner passed. IF partner raise suit, bidding continues as if opponent didn't overcall and partner jumped raised. If opponent overcall (either side of table) and opener doubles, it is penalty, they have bid one of openers strong suits. If they "cue-bid" and opener doubles, it shows the strong two suiter (short in the transfer suit). If they overcall and raise, openers double shows strong two suiter. If they overcall and opener bids anything, he shows the "other two suits" using the rules of lower suit covers. In fact, five different bids can exist, so you can show specific cover needs, for instance...

3C-3S-P-P-?

• Pass - weak hand with diamonds
  
• 3NT - diamonds and club, 4 losers
  
• 4C - diamond and clubs, 3 losers
  Now next five bids show D/C with specific losers
  
• 4D - diamonds and clubs, 2 losers, need heart QUEEN
  
• 4H - diamonds and clubs, 2 losers, need heart King
  
• 4S - diamonds and clubs, 2 losers, need spade ace
  
• 4NT - diamond and clubs, 2 losers, need spade King
  
• 5C - diamonds and clubs, need spade ACE

OF opponent overcalls in a suit or doubles and his partner bids a suit, now openers double shows strong hand without the suit they bid. Double serves as the weakest group response (three or four losers), other bids show fewer losers with same refinements shown above once the level reaches the normal response level without competition. For example..

3D - (DBL*) - Pass - 4S where DBL is takeout
?

DBL = Diamonds and club, 3/4 losers
4NT = diamonds, clubs 2 losers, need a heart
5C = diamonds, clubs, 2 losers, no need for heart
5D to 6C one loser, pin point need (or lack of) for a cover.

As for your "preferred method" for competition richard, I think it is not the best. I believe that awn's method (http://forums.bridgebase.com/index.php?sho...indpost&p=61409) but either is better than what most people might play without specific agreement.

Awm method is similar to yours in some ways, his DBL of the transfer shows "cards" but Hxx in the "bid suits" (I assume here the suit opened and suit promised if weak). The problem with this in theory is after 3C-X-P-? advancer has to worry about the bidding going all pass when opener has clubs since the pass showed club tolerance. That will happen some of the time for sure.

But what I particularily like about his method is his cue-bid ask for a stopper in the preempt suit for 3NT, and his direct seat 3NT is really good values (else double) with a sure stopper (else cue-bid). This is a very effective tool, providing an opportunity not found if the preempt had been in the bid suit. On the other hand, you mix two cue-bid (takeout) with pass then double (penatly). You win when you have trump stack behind opener, you lose when partner has good trumps on his side (since you miss chance to double). In fact, you lose more often because with trump stack in your hand, a pass will often fetch a double from partner, but once you cue-bid the chance to double is gone. So this seems a trade off, while the cue-bid asking for a stopper combined with the double showing stoppers and cards, seem a clear win (gain in clarity).

[inquiry]

Sometimes, MisIry is like magic. Here is a hand from BridgeWorld challenge the champion auction. This is old (July 1988) the hand was deal 6 in Canadians Michael Schoenborn and Harmon Edgar versus Aussie's Barbara Travis and Sue Lusk. The deal was submitted by Gregory Arbour of Vanconver BC.

Scoring: MP



2NT 3♣
3NT 7♥

---

Both pairs reached 6♥. Bidding with Misiry is "easy". West 3NT rebid shows red two suiter, with need for the club KING. Since EAST has the club ACE, he knows two of opener's losers are AK of clubs, the other is the diamond queen. Thus partner has AK of diamonds (five or longer), AKQ of hearts (five or longer), and no spade loser. Thus, after trumps are pulled, partner's club loser can be ruffed after throwning clubs on partners long diamonds becomes a reasonable shot. True, if partner has only five diamonds missing the jack, diamonds have to be 3-2, making grand slam roughly 68% with the worse hand partner can hold, (some 4-0 trump splits could be problems too of course). Of course partner can have six diamonds, and or, partner can have diamond Jack (as here).

Magic, two rounds of bidding.

[awm]

The idea behind my suggested defense is to win as much as possible when opener has the preemptive hand. Supposing 3♣ was opened, showing either a weak hand with diamonds, or various strong two-suiters. In direct seat:

X = cards, something like 14+ hcp, guarantees some cards in CLUBS (minimum of Hxx)
3♦ = asking for a diamond stopper, either a running suit or very strong stopperless NT hand
3♥/3♠/4♣ = natural, fairly normal overcall of a 3♦ opening
3NT = to play, but very sound values (else would try X)
4♦ = strong two suiter, or however you would play 3♦-4♦ normally

After 3♣-P/X-3♦-P-P...

X = takeout, if no original X will be a "balancing" type action
3♥/3♠/4♣ = natural but not as strong as in direct seat
3NT = after an initial X, this is a very serious 3NT (slammish); after pass it's 2-suited t/o
4♦ = four card major and 6+♣ (a tough hand to describe normally), good values

If opener has a preempt, you're essentially strictly better off than you would be normally. Here are the wins:

(1) If you have a pure penalty double, you can double 3♣. Often partner will now make a light takeout double of 3♦ which you can convert. This is not as nice as a "pure" penalty double, but is much superior to passing a natural preempt and just hoping partner balances on his 8-count. You can also win when you have a good hand without classic takeout double shape, and encourage partner to make a light major suit call.

(2) If you have a "dubious" 3NT call (i.e. something like 16-18 flat) you can start with double. Most of the time when 3NT is making partner will act now. This means you will often avoid the bad 3NTs while bidding the good ones.

(3) If you have a borderline bid or double, you can start by passing and then act over 3♦. Partner will know this is a borderline action and be less likely to hang you.

(4) You now have a bid with the tough "stopper ask" hands which are otherwise difficult, as well as the tough 4M/longer minor hands.

The only time you might lose when opener in fact has a preempt is on an auction like 3♣-X-P, but one would expect this to be passed out infrequently. Keep in mind that doubler has a pretty good hand including at least a little bit in clubs. How often will 3♣ actually play better than 3♦ here? I'd bet it's pretty unusual.

When opener actually has the strong hand, you might lose by being unable to bid diamonds directly (well, below the five level). But you can normally make a four-level diamond call at second turn and find a sacrifice if any. If an auction like 3♣-X occurs when opener in fact has a two suiter, this can often mean bad things for the opening side. Most of the points are sitting behind opener, including at least one honor in what's probably one of his suits (clubs). Of course, the odds are against a direct-seat double when opener has the good hand.

[mr1303]

OK, I'm trying to persuade a partner of mine to play MisIry. Does anyone have a simple way of memorising the strong hand continuations?

I managed to learn the symmetric relay by writing it all out, since there's a repeating pattern.

Any suggestions?

Mark

[inquiry]

mr1303, on Sep 1 2005, 07:57 AM, said:

OK, I'm trying to persuade a partner of mine to play MisIry. Does anyone have a simple way of memorising the strong hand continuations?

I managed to learn the symmetric relay by writing it all out, since there's a repeating pattern.

Any suggestions?

Mark

The easy way...

Strong hand can never have the "Preempt Suit" (the suit transferred to), and will ALWAYS have the opened suit (2NT = hearts). That part is very easy. So if you open 2NT you have hearts and spades or hearts and diamonds (NEVER CLUBS)

If you open 3C, strong hand is CLUBS and a MAJOR (never Diamonds)

If you open 3D, strong hand is DIamonds and SPADES or DIAMONDS and CLUBS (never hearts).

That part is easy. Yes?

To show two suiter, rebid SPADES or "suit opened" (Heart if you open 2NT) show spades and suit opened. Any other "rebid" shows the two suiter WITHOUT SPADES. (ONLY EXCEPTION, IF PARTNER BIDS NEW SUIT, YOUR RETURN TO "TRANFER TO SUIT" AND SIMPLE RAISE OF PARTNER SHOWS THE WEAK OPENING).

For level rebid, the higher the level, the fewer the losers. So the cheapest rebid shows 4 losers. After the cheapest rebid, the number of losers go down, with emphasis on the need or lack of need for a cover card in the cheapest side suit. This is only tricky at the four level. The rule is the higher you force responder to choose, the fewer the losers. If there are three options that allow choice at the some level, the cheapest shows 2nd round loser in the LOWER side suit (may have first round loser too). The second cheapest shows 1st round loser in LOWER side suit, and the highest shows no loser in the LOWER SIDE suit. If there is only two such bids (and that is the most with two suiter with ♠'s then the cheaper shows a non-specific need for a lower suit cover. This is only "tricky" at the four level. Higher than that, with spades you show yes or no for LOWER side suit, two suiter without spades above 4 level, always have the three step response. Example is easier than words...

Imagine HEART = SPADE two suiter
2NT - 3C - ?
  - 3H = ♥/♠ two suiter, 4 loser (cheapest bid)
   -3S = ♥/♠ two suiter, 3 loser, need a ♣ (non-specific)
   -4H = ♥/♠ two suiter, 3 loser, no need for ♣
   - 4S = ♥/♠ two suiter, 2 loser, need a ♣ (non-specific)
   - 5H = ♥/♠ two suiter, 2l oser, no need for ♣

NOTE, since 3♠ forces responder to bid 4♥ if he prefers ♥ it shows the same number of losers as 4♥. Since this are at the "same" level, it the cheaper 3♠ shows need in the lower side suit, the higher denies such need. Now imagine a SPADE two suiter without ♥'s.

3C - 3D - ?
  -> 3S = ♠/♣ two suiter, 4 loser
   -> 4♣ = ♠/♣ two suiter, 3 loser
   -> 4♠ = ♠/♣ two suiter, 2 loser need LOWER COVER
   --> 5♣ = ♠/♣ two suiter, 2 loser no need for Lower cover

Here 4♣ can not be used to separate between 3 losers with and without a need in the LOWER side suit, because there would be no second way to show 3 losers and still allow responder to choose between the suits at the four level. While 4♠ shows 2 loser and the diamond need, and 5♣ 2 losers without a ♦ need.

It is this, "how much room" do I have that determines if you show a cover is needed or not in the lower suit. For example, with a ♥/♦ two suiter, you can show need for King or Ace of clubs with three losers.... (openres rebids, 3H = 4 loser, 3NT =3 loser, need CK, 4C = 3Loser, need CA, 4D = 3 loser, no C needed). While with ♥/♣ two suiter you can only show need for non-specific ♦ at the two level (open 3♣ then rebid: 3♥ = 4 loser, 3NT - 3 loser, need a ♦, 4♣ = 3 loser, no need for ♣). And as noted above, with ♠/♣ two suiter, you can not show side suit need until you have 2 loser or less.

Note, even with the ♥/♣ two suiter where with three losers, you can not separate between the need for 2nd or 1st round side control, you can if you have 2 losers or less. Because... once you bypass 4♣, you have three bids to show that specific two suiter at or below 5♣, these are, 4♦, 4NT, and 5♣--- (remember 4♥ would show ♥/♠.

You can refigure these out in two seconds at the table. So memorization is not a problem. If they overcall or your partner bids something other than the transfer suit, the rules are just as simple.

[inquiry]

Jlall, on Sep 1 2005, 10:26 AM, said:

A few comments:

1) I agree with Richard that your method of analyzing is very flawed. Most of these hands seem to deal with weak opposition, and against them you can play strong 3's and still win. Presumably you want to beat better players, and I don't think this bridgebrowser method helps in any way.

I haven't analyzed anyhands with respect to the strength or weakness of the competition, so I am not exactly sure what you mean by this. I simply used BRIDGEBROWSER to find the hands. Period. That is, I didn't try to biases the results by "constructing hands". Others can use dealer or other program to deal the hand types they want. Sadly, I lack the ability to simulate hands, but luckily I have access to tens of millions of hands from okbridge and bbo, and an easy way to extract them.

Quote

2) No one knows what the losses really are against a good pair. How much will giving them extra ways to get in over your preempts lose? I suspect a lot, giving the opps more accurate bidding against preempts is counterproductive and the gain needs to outweigh the loss which brings us to...

Well, the "loser" estimate is based on the standard defination. Of course, if you think AKxxx is "one loser" you will be sorely disappointed if partner either lacks support for BOTH of your suits, as you could lose three tricks in this suit for instance. With misfit in both suits, of course, responder has to be extremely careful.

Quote

3) No one knows what the gains are of this method. You show a bunch of slams that you could bid, but who knows if a good pair would get there as well with more normal methods. No, I don't think bridgebrowser helps determine this. Obviously risk/reward can not be gained when neither risk nor reward is known.

There is one gain, no one has metioned yet, and one of the main reason people who play transfer preempts (or two under transfer preempt) choose to do so. That is, when you play the contract from the other side, on average, you gain about 1/3 of a trick per hand. Now, I did not say this, this was said by an author in Bridge World. But it is obvious that leads up to your partner will be more benifitial than leads through him when you play the hand. And his hand being hidden is more often helpful than the preempters (they can place side honor cards more acccurately when they see what he had).

Quote

4) Many of the "natural" auctions with these hands require good judgement calls to achieve par. Similarly, in your methods judgement will sometimes be required. The way you are analyzing it, the judgement is always perfect. I'm not saying you're giving unrealistic auctions, but at the table the auctions are not always so perfect, and a wrong judgement is made.

Well, judgement is a two way street. By removing the strong hands with two suiters from both opening one bids (the kind of strong hands) and 2♣ openers (the really strong hands), you simplify both your 1suit=bid=jump rebid and your 2C auctions (which no longer can include two suiters). I find, at least for my feeble mind, my judgement is much improved not only on non-competetive auctions, but even more so on competitive ones. It is like the rebid problem hand earlier today, if I can open 2♣ on that hand and have method to describe it accurately, it makes my rebids when I open 1H much easier (Since I have removed that type of hand from the mix).

Quote

5) You lose on hands where you cannot preempt to the max because pard may have the strong hand type. Yes, I am aware sometimes you can preempt to the max and let partner worry about it if he has the strong hand becuase you will have support, etc, or sometimes you can just figure it out. But sometimes you can't, or will require good judgement to do.

This is, in effect, a real world problem. However, if you have some cards in one of the suits your partner must hold if strong, you can freely jump raise. You are not going to blast an advanced save to the five or six level, this is certainly true. The jump raises so far in partnership bidding room, and looking at real world examples, mainly hurt in identifying magic fit slams. There is a situation that you didn't mention specific to the heart suit. Here, your partner open's 3♦ showing hearts or a strong suiter with diamonds and a black suit. Responder with some hearts might be on the fence, do I bid 4♥ to make, or bid 3♥. The problem hand is when you might make game based on heart fit, or else you need to stay low if partner had diamonds and a black suit. This hand type (I used bridgebrowser specifically to look for such hands) is the most problematic.

We may never know how well MisIry works if people don't try it out. The good news (for me at least) is several people are trying it. Non world class, but several very good players. It will take time to get the feedback on when they use the bid, but I also await how it works not only with the weak variety (as well as strong), but also how having the strong two suiters removed from their one bids and 2♣ openings, help their auctions. My experience has been that the improvement in normal bidding due to this change, will far, far outweigh the ability to raise the preempt immediately to five or six level.

[han]

I told Ben about the Bridge World quote, I read it recently. Unfortunately, I don't remember who said it, but I'm going through by bridge worlds now trying to find it. I vaguely remember that it was a challenge the champs contester who said it, but not absolutely sure.

[han]

My experiences with MisIry are very limited, they haven't come up very often. Perhaps either variation (weak or strong) has come up twice now while playing. I will refrain from making a further idiot of my myself by giving qualitative statements about the covention.

However, I will say this: the convention is not hard to learn. The follow-ups are very logical.

Also, the negative inferences are good. For instance, Ben plays 1H-1NT-3C as natural (5-5 or better) with about 4-5 losers, so non-forcing. This is somewhat like in strong club systems: jump shifts are descriptive and limited. I always liked that.

[inquiry]

I should add another advantage, if you open at the 2NT level or higher and happen to have the good hand, it is harder for your opponents to get in your way.

You have removed Unusual 2NT and michaels cue-bid for instance. You make it harder for them to overcall immediately and find a good fit (although one could argue that they might come in later, once you show two suits by showing the other two suits). I have obvserved a lot of hands where if you open one of your long suit at the one level the next hand would automatically overcall 1♠, but realistically never overcall 3♠

[han]

Found it! Here is the quote, from the January 2003 BW:

Choice Cuts, by Bart Bramley:

...

On this complex deal, Sidney Lazard wheeled out one of his pet conventions, a high-level transfer preempt:...

[deal is shown and transfer gets doubled]

... I accepted the transfer to gain advantage by declaring with the unknown hand. (Sidney estimates that this advantage is worth about a quarter of a trick per deal. My own observation, based on a more limited sample, is that the edge is even greater.)

[Al_U_Card]

Hannie, on Sep 1 2005, 11:25 AM, said:

I accepted the transfer to gain advantage by declaring with the unknown hand. (Sidney estimates that this advantage is worth about a quarter of a trick per deal. My own observation, based on a more limited sample, is that the edge is even greater.)

Quite likely cause Sidney gets a higher class of opps.....lol

[inquiry]

Zar has posted several large collection of hands that he has "analyzed" with two simple programs. One that mearly counts starting ZAR and GOREN points added to double dummy bridge solver for "what makes". For example, one file has hands with 27 combined high card points that can make GRAND SLAM (lucky lie of cards and double dummy play necessary on many).

This 27 hcp hand file contains 3407 hands. If someone would like to suggest a starting number from 1 to 3340 or so, we can pull up the first 20 hands that would be bid by misiry starting at that point and and see how the auction would go (YES RICHARD, I KNOW, just another "random test" that proves nothing to you.[1]) The thought being, the system is now well document (I Can send word file that is easier to read, or post it on the web BEFORE the challenge is accepted. The concept here is to show, for those who are interested, how the follow up auctions go after the intial opening and response. Or, you can pick a different one of ZAR files as a starting point. Ones that make small slam, ones with more hcp or less. The advantage of the grand slam (many very very lucky I might add) is that there is a high frequency of such hands (and anyone wanting to look into my strong three suiter methods, a lot of those too).

Anyway, I offer this as a way to display the bidding methods.

[1] Footnote, past challenges in this area have included all strong MisIry hands bid in all ACBL events in the next week, all strong MisIry hands played on April 1st 2005 on BBO in the main room, all MisIry hands played by a particular individual. None of these approaches were acceptable to the purist, who suggest alternatively that randomly choosen hands are not useful, becasue the sample size is not large enough, BBO selected hands are not good enough because of weak competittion (the idea is not what other people bid, it is whether the contract reached is reasonable, so this one has never made sense to me), or who cares what you do with the strong hands, there are SOOOO many weaker hands and you are nailed to the wall on the weak ones so any gains on the big ones is useless....

Let me say, in preemptive reply to these problems people will raise, 1-I don't care one whit about the quality of the field, it is rather the bidding with the strong hands is consitent and the quality of the contracts reached that I am interested in, 2-I lack the techical ability to design software to apply my simple bidding rules to randomly dealt hands and do a statisitical study of the results based upon singl dummy evaluation of the contract, so random hands and look is the only way I can go. By letting others (or random chance) choose the hands to look at, I am not biasing the results by selected hands that favor the method. 3-The purpose of this exercise is to show how the bidding might go if you were dealt the hands and playing misIry. I am willing at anytime for some to say, but what if EAST bid, somethign or doubled, how would it continue.

I got, for instance, this hand from a reader yesterday....

Quote

Hi,
I would be interested to know how you'd bid these hand to help me understand the developments of your system:

Ax
AQxxx
AK109xx
void

xxx
KJxx
Q
Axxxx

A beginning like 2NT 3♣ 4♣ 4♠ as denial and then ...

My reply, in blue follows. So the goal in this potential exercise is to show, for those interested, how the follow up works.

Ok.. a simple way would be...

2NT <<---- CLUBS weak, or HEART and SPADES, or HEARTS and DIA strong
3C <<--- PASS/CORRECT
3NT <<--- HEART/DIAMOND 2 suiter, 3 loser, need club
7HE <<--- Heart ACE, DIAMOND QUEEN, CLUB ACE = 3 covers, 3 losers.

Why 3NT? Because club ACE is working for you.

-----

But a purist would not like that auction. Why? Because you will miss the grand slam if your partner's club ACE was changed for the club two. So the "correct" auction is...

2NT - 3C
4C - 4S
4NT - 5S
7D - 7S

2NT - Clubs weak, or HEART/SPADES or HEARTS/DIAMONDS strong
3C - PASS/CORRECT
4C - DIAMOND/HEART, 3 loser, no need for club
4S <-- DENIES A SPADE CONTROL
4NT <-- missing spade not fatal, I am missing ♦ Queen (5♣ shows missing HQ, 5NT both missing queens)
5S <<--- ruffing value in spades useful (here CLUB ace counts)
7D <<--- yes, if I can ruff a spade
7H <<--- ok

Note, 4NT shows no concern about two spade losers, and that the ♦ Q

Note, also with two covers, responder always knew he was going to at least six,

Note, over 4NT, 5S is the asking for "distributional king". This is described on my webpage and or I can send a much easier text file via email. Even with xxx KJxx Q xxxxxx you want to be in grand, because you can throw your spades on partners AK of diamonds, and ruff his spade loser (as long as he has the spade ACE). Your partner knows you don't have the ACE so he is not accepting any grand try with it missing.

So the correct auction finds slam opposite the actual hand were responder has the club ace, or a hand where responder can throw away his little spades on your good diamonds, and then ruff the spade loser. I hope that helps and you can see the logic of 4NT and 5C showing one missing queen, and 5S looking for a slow cover in spades.

[hrothgar]

inquiry, on Sep 1 2005, 10:34 PM, said:

None of these approaches were acceptable to the purist, who suggest alternatively that randomly choosen hands are not useful, becasue the sample size is not large enough, BBO selected hands are not good enough because of weak competittion (the idea is not what other people bid, it is whether the contract reached is reasonable, so this one has never made sense to me), or who cares what you do with the strong hands, there are SOOOO many weaker hands and you are nailed to the wall on the weak ones so any gains on the big ones is useless....

The "purist" mistakenly thought that you were interested in objectively studying the merits of your method rather than just playing with your tool...

[inquiry]

Jlall, on Sep 1 2005, 04:23 PM, said:

I don't understand why the focus is on the strong hands here.

1) It seems to consist of only 25-33% of your hands when you open at the 3 level (or 2N)

Ok, I will try to explain in simple terms why I focus on the strong hands. IT is because when you open with the weak hands, the opponents can enter the bidding a lot of different ways (in fact, as everyone has pointed out, in more ways than they can if you open a normal preempt). Trying to figure out, at least for me, what SHOULD happen on those hands is a logistical nightmare. I much prefer people to try the preempt out and see how it works. For me, and a few people reporting back to me, the results have been at least ast good as if we had opened a normal preempt.

Experience will tell if the transfer preempt is unplayable, but so far, this is far from true. Now, how about agaisnt expert pairs prepared for it? Well, Ultimate club and tangerine club at least both use transfer preempts (two under at least in ultimate). While Rodwell and Meckstroth no longer play transfer preempts, they use to at teh very highest levels. If transfer preempts were a great deteriment they would not have (no doubt they changed back because it might be, as I have always admitted, a milld detriment). Clearly they were playing them at the very highest level. A hand quoted by LArry Cohen in Bridge World shows that Marty Bergen opened wtih transfer preempts (it was 2D transfering to 2S), and of course he is a champion of his two under preempts. (NOTE, amazingly multi 2♦ itself is a transfer preempt, yet we don't see everyone playing that running for the hills because of the weakness of transfer preempts). Nor does anyone complain about Rubens Advances starting with cue-bid up, where opener then bids the "transfer" suit if he would pass a non-forcing bid there. Seems that would suffer the same complains, yet since that article came out in the late 70;s or early 80's tons of people have successfully adopted that method.

Quote

2) I contend you will likely lose when you have a preemptive hand and win when you have one of the strong versions. None of this is known obviously, and how much of a gain/loss it is is unknown.

I agree with this statement, but I would add, I think I gain on any hand where I don't open with the preempt by removing one important hand type from the mix. Same thing iwth my stuffing acol 2 of a major in 2♣ and acol two of a minor into multi 2♦. Bidding is now EASIER when I don't open with these bids.

Quote

3) You contend, certainly correctly, that you gain on openings like 1H when responder is able to make an inference because you did not open with a 3 level bid.

Yes, see under 2.

Quote

You say you do not care about the quality of the field, but this drastically affects how they handle the bids when you're just preempting.

When I am fishing, errrr advertsing, for people to try MisIry, I ahve to convince them that they can bid their BIG HANDS better when they use it or I will not get far. So, it is not the RESULT versus the field, but the result with the hand (strong hand) that I push. And, I can't predict how the oppoents will bid their hands, but I can predict how the strong hand auctions will (or should) go. Misho still surprizes me with the occassional, "i am jsut goign to hope my cover works" jump to slams... but for the most part, the auctions are "automatic".

Quote

Remember, when you preempt it's likely to be their hand. The quality of your opponent is of huge importance when it's their hand and they need to deal with a preempt. This is, of course, the main loss of your method and all transfer preempt methods. To do a "study" you have to figure out what you gain or lose on preemptive hands, strong 2 suiters, and hands where you are able to make a limited jumpshift.

Well, that is the problem isn't it. How do you quantify all these options. The different ways the opponents might bid if you are weak, and if you catch them when you are strong. How well the strong version works. How well the weak jump shift works. And then one you haven't covered yet, if you open 2♣ your "two suiters" will never be 5-5 or 6-5, so openers second suit is never more than 4 cards. This is helpful as well, opener can return to his first suit and responder always knows opener never has more than in his second (under any circumstance). That is a lot to try to "calculate".

So my goal, as it was, is to find a way to convince people to give MisIry a try. And figure out how it also effects their other bids (their jump shifts for instance), and their 2♣ auctions. No doubt those reading these pages will spring transfer preempt on unprepared opponents, so at first it might do well because of that (sad, best to have prepared defense for your opponents). But as people start using it (if they do), maybe we can get some answers.

BTW, I fully prepared for someone with the skill and the methods to answer these questions using software tools and statistical analysis to have at it. I would be more than happy to review their findings....

[the hog]

Some comments. Partner and I have decided to try this for a number of reasons:
*Fun
*They fit really well into our canape system
*Playing strong C, they offer some relief from haveing to open good 2 suited hands with 1C and getting pre empted out of it

One possible modification - as you have 3D as a H suit and 3H as a H suit available, we are considering playing 3D as a good 2 suiter only. This enables responder to bid 3N with stuffing in the Majors and being confident this will be a good contract unless opener is VERY strong.

[hrothgar]

inquiry, on Sep 2 2005, 12:06 AM, said:

Experience will tell if the transfer preempt is unplayable, but so far, this is far from true. Now, how about agaisnt expert pairs prepared for it? Well, Ultimate club and tangerine club at least both use transfer preempts (two under at least in ultimate). While Rodwell and Meckstroth no longer play transfer preempts, they use to at teh very highest levels. If transfer preempts were a great deteriment they would not have (no doubt they changed back because it might be, as I have always admitted, a milld detriment). Clearly they were playing them at the very highest level. A hand quoted by LArry Cohen in Bridge World shows that Marty Bergen opened wtih transfer preempts (it was 2D transfering to 2S), and of course he is a champion of his two under preempts. (NOTE, amazingly multi 2♦ itself is a transfer preempt, yet we don't see everyone playing that running for the hills because of the weakness of transfer preempts).

You're ignoring a few points here:

1. Many of the pairs that you cite adopted transfer preempts for very specific reasons. Case in point, Marty Bergen adopted a 2 under preempt style because his preempts were extremely undisciplined. 87643 was a good enough suit for a 2♠ opening. The two under style allowed partner the luxury of asking about the hand without bypassing 2♠.

2. Several pairs adopted transfer preempt styles because they thought that xfers were a winner in and of themselves. In theory, the gains of rightsiding contracts and allowing the undescribed hand to declare could outweigh the losses from providing the opponents with extra bidding space. For the most part, these pairs have abandoned this theory.

3. I don't konw many players who adopt a multi because they love opening 2♦ with a weak 2 in hearts or Spades. Rather, they they adopt a multi because there are are more valuable uses for the 2♥/2♠ opening bids than "traditional: weak single suit hands and they'd rather give up a weak 2♦ opening that sacrifice the 2♥/2♠ openings.