[straube]

In or strong club auctions, responder show QPs starting with 5 and each successive step shows one more. We get high frequently. what's the best way to economize space? Thinking


S1 5 or 6
.....S1-6
.....other-5 zooming into parity cue bidding
S2 7 or 8
.....S1-8
.....other-7 zooming into parity cue bidding
S3-9
S4-10
Etc

[rbforster]

One way to think about this is in terms of how much opener needs to ask vs sign off, and how strong responder needs to be in order to make a move after a sign off attempt. 18-19 QPs is about the min for a slam, and 18 is going to be pretty optimistic without good shape opposite shortness, a long suit source of tricks, etc.

What level is your typical QP ask by opener? 3♥-3♠?

Some systems use

step 1 - bad hand, fairly wide range
step 2 - zoom into showing something about a slam invite hand

[straube]

rbforster, on 2012-May-19, 10:40, said:

One way to think about this is in terms of how much opener needs to ask vs sign off, and how strong responder needs to be in order to make a move after a sign off attempt. 18-19 QPs is about the min for a slam, and 18 is going to be pretty optimistic without good shape opposite shortness, a long suit source of tricks, etc.

What level is your typical QP ask by opener? 3♥-3♠?

Some systems use

step 1 - bad hand, fairly wide range
step 2 - zoom into showing something about a slam invite hand

[straube]

Yeah. Our 5431 is shown at 3D so typically our ask is 3S. Unfortunately sometimes 4C.

[Zmud99]

What about asking about Neopolitan Control (A=2,K=1) instead of QP's ?
In my modified Moscito version , I start control asking in average in 3♦/3♥ and with a Control reply set up like :
1 step = 4 Controls
2 Step = 2/3 controls
3 step = 4 controls but with just 2 aces
4 step+ = zooming DCB with 5+ Controls

I really think that Neopolitan Controls are far away from QP's in terms of accuracy and space

What do you think ?

[straube]

We like QPs because it gives us a better overall strength of the hand. We use QPs, too, for deciding whether to initiate a GF. I do understand that controls will be more efficient in many circumstances.

[straube]

3H-ask (present method)
.....3S-5
.....3N-6
.....4C-7
.....4D-8
.....4H-9
.....4S-10

3H-ask (proposed)
.....3S-5 or 6
..........4C-ask
...............4D-6 +2
...............etc-5 +1
.....3N-7 or 8
..........4C-ask
...............4D-8 0
....................etc- 0
.....4C-9 -2
.....4D-10 -2

3H-ask
.....3S-5 or 6
..........4C-ask
...............4D-6 +2
...............etc-5 +1
.....3N-7 -1
.....4C-8 -1
.....4D-9 -1

3H-ask
.....3S-5 to 7
..........4D-5 +3
..........4H-6 +3
...............etc-7 +3
.....3N-8 -2
.....4C-9 -2
.....4D-10 -2

[rbforster]

There are a lot of considerations, but let me lay some out for you to think about.

1. You don't need to drop down the 9+ QP responses 1–2 steps since those are almost slam going anyway and should have plenty of space/safety. Remember DCBs are more efficient the more high cards you have since you don't stop and relay as much.

2. You might want to compress/drop down some of the 5-8 QP responses to 3NT or lower, if only because you might want to stop there. That's not the most important factor, but it matters a little. You might even want different schemes depending on the level of you ask, if you were really feeling ambitious.

3. Run-on responses by strong responders after open tries to sign off are much more economical if done over 3N than a suit sign off with 4♦ and having to start showing at 4♠+.

......3a. This suggests that a 3N sign off can be pretty wide ranging, say up to 11 QPs, since with no fit and needing a lot of strength, responder only goes on with 9+. Then there's plenty of space to ask more, maybe find a minor suit slam, etc.

......3b. A suited sign off (4♦ terminator or similar) should be much more limited, maybe <=9 or 10 QPs, since there is lots more space to ask below 4M/5m safely and since when repsonder goes on he's answering at a very high level. A zoom over 4♦ probably wants to be stronger, maybe 10-11 min instead of just 9.

4. Once you decide on your sign off rules, you can start estimating the average opener strength conditional on asking for QPs, and in light of that, what the relative probabilities are for responder's QP strength. I'm pretty sure these are just decreasing with QPs except maybe for 5 vs 6, but there are some variations that you can do when Step 1 > S2 +S3, or similar. Again, here you aren't just optimizing for probability order, but for allocating space to answer 1-2 DCB asks conditional on opener having liked the strength answer.

5. You may also want to consider compressing two steps (say 5-6), but then not separating them and just using DCBs with fewer negative inferences. This makes your 5-6 slam tries worse, but slam was less likely then anyway, and it makes all your 7+ slam tries one step better.