Climbing Mt. St. Ogustine:

Modified Ogust -- then and now

Gordon Bower

Then

In my longtime partnership with Michael Schmahl, we initially simply agreed to play all our responses to weak twos as described in Andersen and Zenkel's Preempts from A to Z: simple Ogust; new suits forcing; jump shifts control-asking bids. At some point we used a 3 response as asking for opener's singleton, but it hardly ever came up, so I can't honestly remember when we started playing it.

Eventually we got to thinking about Ogust. We quickly formalized the notion of "good suit" as "2 of the top 3 honours," as recommended in many of the standard sources. That got us wondering how best to formalize the notion of "good hand." Ultimately we agreed that our first- and second-seat weak twos would promise QJxxxx or better in our bid suit, and at most one stopper outside trump. We would use the "hand" part of Ogust to show or deny the presence of that outside stopper.

We quickly realized that we might care which stopper opener had, so we reorganized the Ogust responses to 2 and 2 to give us room to find out:

3 showed 1 of the top 3, and promised an outside stopper. Responder could ask where it was with 3, and opener would answer with 3 of the unbid major (natural), 3 of his own major (diamonds), or 3NT (clubs).

3 showed 2 of the top 3, and promised an outside stopper. There wasn't room for a full asking sequence, so we just focused on the unbid major for the followups: responder's rebid of 3 of the unbid major said "bid 3NT if you have this suit stopped."

In a nutshell, 3 was the most frequent and complicated hand, and 3 of opener's major was the weakest response.

This was a more comforable arrangement over 2 than over 2. Over 3 the same principles applied, sort of: 3 was as above (with responder's 3 and 3 rebids both saying "bid 3NT if you have a stopper"), 3 was the weakest bid, 3 was a good suit with a stopper, and 3 was a good suit without a stopper. But it was hard to get much use out of the rearrangement of the steps over 2, and we both found it hard to remember the order of the steps.

Michael and I tried to introduce this to a mutual third partner in 2011, but ultimately abandoned the above modification of Ogust. For a time, he went on to some other relay-like experiments, and I (shock! horror!) went back to playing Features.

Now

In summer 2013, I got interested in Ogust refinements again. I still felt, as I did 10 years prior that 3M, not 3, should be the weakest rebid, and 3 should be the hand-type most likely to need followup questions.

In the meantime, I had come to another realization: my weak two style varied dramatically with the vulnerability-- as preempts should. In effect, using classical Ogust, I would only ever use the weakest two or three responses at favorable, and would only ever use the two strongest at unfavorable. Perhaps it would make more sense to redefine the Ogust rebids on a sliding scale that matched the vulnerability.

On further reflection, I realized that the kind of hands that I opened with weak twos was affected by the vulnerability in an interesting way. For pure preempt hands, I was applying the "rule of 2, 3, and 4." With a hand like KQJxxx xx xx xxx I almost always open at the 3-level favorable; at equal vulnerability this is a spot-on sound weak two, at unfavorable it is close to the worst pure-preempt hand I would still open 2. But there were hands with outside values that I would open with the same weak two at any vulnerability-- say KJxxxx Axx xx xx. I adhere strictly to the "never have two defensive tricks for a preempt" rule, but am willing to open a weak two with "two aces" hands. That is -- it appears that it makes sense to vary some, but not all, of my Ogust rebid definitions with the vulnerability.

Now, the final ingredient in creating a better Ogust: as responder, what questions do I want answered, when I use Ogust? If I care about how good your support for my suit is, I should bid my suit and see if you raise. But I might use Ogust if —

• I have the other three suits stopped, and I want to know if your suit will run for six tricks in notrump (or play for no losers in a slam);
• I have the other three suits stopped, and I want to know if I can set up five tricks in your suit, then reach your hand to cash them;
• I am missing a stopper, and only want to play 3NT if you have that side suit stopped;
• I am about to raise you, and simply want to know how many tricks you have, to know how high to raise.

As before, our two areas of focus will be trump quality and side stoppers. How best to handle the trump quality question? "Do you have 2 of the top 3 honors?" and "Can I run your suit opposite (say) Kx" are almost but not quite the same question. And I sometimes open a weak two on a 5-card suit, at favorable vulnerability: knowing the suit will run might not always mean having six winners.

I will present my chosen solution in a table, then elaborate on how it is used and its rationale. After 2-2NT:

• 3: promises an outside stopper.
• 3 now asks where the stopper is: 3 hearts, 3 diamonds, 3NT clubs (or whatever permutation you find easiest to remember. To my mind, "new suit natural, clubs are touching notrump" was most intuitive.)
• 3 asks whether you have a good or bad trump suit: 3 bad, 3NT good.
• 3: shows an "unusual hand", either a weak two with too much outside of trumps, or a hand that would be too strong to open only 2, but has too much defense to preempt.
  Responder's 3 asks which: returning to 3 shows the "two aces" hand, 3NT the "1½ tricks outside" hand.
• 3: shows a pure preempt, maximum for the vulnerability.
• 3: shows a pure preempt, minimum for the vulnerability.
• 3NT (rare): as in classical Ogust, AKQxxx in trumps. Probably only at unfavorable.

After any of the above, responder's bid of 3 or any game or slam is to play; one can either use 4 as an artifical shortness-ask, or have all new-suit bids between 3M and 4M be control-asking bids (in effect asking for shortness in that suit.)

Over a 2 opening, 2 is the asking bid; the entire structure is slid down one step (and 3 rather than 4 can be the follow-up shortness ask.) 2-2NT shows a spade suit. Since 2-2(natural and forcing)-2NT never made much sense as a natural bid anyway (opener can't have both minors stopped), loss of the extra step seems not to matter. The simplest rebid scheme after 2-2NT seems to work: 3 or 3, feature, no spade fit; 3, minimum, no spade fit; 3, minimum, spade fit; jump shifts, splinters promising Hxx in spades.

Over a 2 opening, there is some question whether to use this structure at all; 2M natural and forcing and 2NT asking something simpler is reasonable. Alternatively, one can use 2-2 as Ogust, 2-2 as hearts, and 2-2NT as spades. It depends how much artificiality is to your taste.

Varying with vulnerability

The 3 and 3 "pure preempt" rebids vary strongly with vulnerability.

At favorable, "minimum" pure preempts ought to provide 4 tricks, or close to it: QJTxxx, or KTxxxx, or similar -- or KQJxx or AQT9x, if you open 5-card suits. "Maximum" pure preempts should provide about 4½ or more. KQxxxx, AJTxxx, AQxxxx, or AKJTx. Responder needs two of the top three to expect six tricks opposite a minimum; at least Kx, opposite a maximum.
For six-card suits, "maximum" is nearly synonymous with "2 of the top 3," except for AJT-high. Pure preempts as strong as KQJTxx or AQJxxx will likely open at the 3-level vulnerable. (KJTxxx can probably safely count itself as maximum; responder looking at Qx will know AK-high is unlikely and a responder with Ax has an even chance at running the suit.)

At equal, anything less than 5 tricks is minimum: KJT-, KQT-, AJT-, and AQ-empty suits. A maximum requires very good intermediates, KQJTxx, or at least AQTxxx or AKxxxx. Partner can be nearly certain of six tricks with Hx opposite a maximum; may need a finesse or a bad break opposite a minimum.

At unfavorable, 5-trick hands are now minimum; a maximum means AQJ or AKJ-high, with play for six tricks opposite xx or xxx. Hands like KQT9xx or AJT9xx are distinctly subminimum.

The 3 rebid, on the other hand, varies only slightly with vulnerability (except that, at favorable, it will occasionally include a 5-card suit with a side stopper, AQT9x Kxx Dxx Cxxx or similar.) The 3 rebid only slightly more, as the definition of "too strong for 2" changes. At favorable, KQxxxx HAxx Dxx Cxx would be strong enough to justify 3; 3C could be as weak as KT9xxx Qxx Dxx Cxx. At equal or unfavorable, the former would be a normal 3 bid and the latter not worth opening. Equal 3 bids will feature many 6-tricks hands that aren't well suited to a 3 opening, like AQTxxx xx Dxx CKx. Unfavorable 3 rebids will overwhelmingly include a reasonable ace-high trump suit and a side ace.

Opener's third call after 2-2NT-3-3, however, is again very sensitive to the vulnerability: a "good" suit here can only be about half a trick worse than 2-2NT-3 would promise-- good enough to produce five tricks opposite one fitting honor and six tricks opposite two, at favorable, and scaling up appropriately at equal and unfavorable.

Examples

Opener A: A K x x x x x x x x x x x	Opener B: K J T x x x x x x A x x x	Opener C: K 9 x x x x x x x A x x x	None vul	Responder: Q x x A K x x x x A K x x

Playing classical Ogust, East would go on over any rebid but 3. Playing this version, Opener A will show a good suit but no side card (2-2NT, 3-4); Opener B will first show a side stopper, then show (under the circumstances) good trumps (2-2NT, 3-3, 3NT-4); Opener C will show a side stopper but then show bad trumps (2-2NT, 3-3, 3.)

Opener: K Q T x x x x x x x x x x		Responder A: x x A x x x A K Q A x x x	Responder B: x x A K x x A x x A x x x

At favorable, Responder A should ask with 2NT; opener shows a maximum pure preempt with 3 (he has KQTxxx where he could have had K-empty or QJTxxx) and a good game is reached. Responder B should seriously consider passing: he should expect 8 tricks against most "bad" favorable weak twos, and 9 tricks opposite most good hands. Only the very best favorable weak twos have play for game.

At equal, Responder A can reasonably blast to game. Responder B should ask with 2NT, and now opener shows a minimum with 3 (he has only a 4½ trick suit where he could have held six to the AK, AQT, or AQJ.) Only at unfavorable should Responder B consider blasting to game (and this opener is subminimum for an unfavorable preempt -- opening at unfavorable is a calculated risk, an effective preempt when it is the opponents' hand, but he cant be too surprised if his partner expects more and goes off one.)

K J 9 x x x x x x K 7 x x	None vul	A Q J 9 x A K Q J x x J T

A Q 9 x x x x x x x x x x	NS vul	K x A x x x A x x x A x x

Classical Ogust can achieve the same result; classical features do not.

K Q T x x x x 9 x x A x x	Both vul	A J K x x A Q x x 9 7 x x

At favorable, East should abandon hopes of a tip-top maximum and pass; I would open 3 at favorable with this 5½ trick hand. (And perhaps responder should consider passing at equal too, if he isn't willing to try the 9-trick game.)

x x A 8 7 x x x A J x x x	None vul	A x K Q Q x A K Q J x x x

Playing the system described above, the auction might be 2-2(asking), 2NT(side stopper)-3(where?), 3(here), followed by Blackwood and a swift trip to the grand.

Q J x x x x K Q x x x x x	None vul	K x J x x A x x x A K x x

Opener A: K Q T 9 x x x A x x x x x	Opener B: Q J x x x x x x A x x x x	Opener C: K Q J x x x x x x x x x x	Opener D: K Q T x x x x x A x x x x	None vul	Responder: A x x x x x x K Q x x

Playing features, Opener A will surely show his A, and so might opener B, resulting in getting too high.

Playing the system described above, East asks with 2NT. Opener C rebids 3 showing a good pure-preempt hand but nothing outside and responder signs off. Openers A, B, and D all rebid 3 showing a side feature. Responder wishes he knew both where the feature was and how good opener's trumps were, but there isn't room to ask about both: responder must prioritize, and use the 3 trump-quality-ask rebid. Openers A and B with 4-trick trump suits rebid 3; Opener D rebids 3NT to show his better suit.

A final thought

If you are happy with standard Ogust, incidentally, but wish you could save a bidding step, notice that all you need are the 3, 3, and 3 responses, and the rebids after the 3-3 re-ask. If you don't care about ever finding specific stoppers, you only need three steps. (Indeed, even in my version, 2-2NT-3 is rather under-utilized, and a prime place for you to stick whatever "nonstandard" hands you like to open with a weak two.) Maybe you want to use 2NT for something else, and want to fit a "mini-Ogust" into the 3 response:

• 3: good hand. Now 3 asks about suit: 3 bad suit, anything higher, good suit.
• 3: bad hand, good suit
• 3: bad hand, bad suit (or the other way round if you prefer the traditional order)