[Crunch3nt]

Hi all,

When playing a relay system, there are some problem hands that normal denial cue bidding (DCB) or Spiral Scans methods don't hand well. eg when the person asking has a void, or has two very strong suits but blank in the other 2 suits eg AKQxx x xxx AQxx.

What methods do relayers use to combat these? Some kind of exclusion for the void hands perhaps? I have seen non-relay precision players use Asking bids quite effectively, yet the method seems to be generally looked down upon?

Thoughts please.

[foobar]

Do you have specific examples in mind? It seems that playing a QP scheme with a well defined base, it should be possible to handle most hands, but obviously there will be situations where the relay captain may have to make an educated guess.

In some systems, 4♦ is the canonical terminator puppet after shape resolution, the first step starts the DCB scan and the remaining bids are RKCs for various suits in order of length (Richard's Moscito 2000 document has examples).

Reverse relays might help help with some awkward hands by allowing the(presumably) balanced hand to do the asking. IMO, reverse relays are best suited to limit hand strength, but there may be different schools of thought out there...

[cloa513]

High level bidders don't handle it at all well when partner splinters/cues with hands like

---

They often subside in 4♠ just like yours if partner escrews cueing and you have all the side cues, push forward to the five level must mean I have all the side cues and second level cues, a cue in the side suit means I have most of them.

[mtvesuvius]

Generally the best way to get around this problem is to have a way to Keycard, as well as a way to DCB/Spiral Scan. I've seen a lot of problem hands in relay-type systems, where Spiral Scanning just plain won't work, or leaves a bad guess (I am 1-8 on those now I think, about 5-60 in the bidding room, ask Owen...). If you don't want to, or don't have room for (!) any RKC-type bids, you'll have to break the relay a lot more often to set trumps earlier. It takes a while to get a feel for when to do this, but often it's the only solution.

[blackshoe]

I think of spiral scan as part of the response structure to RKCB. I'd not considered it in other contexts.

That last hand, it seems to me, should be bid (North opening) 1♠ - 2NT (too strong for a direct splinter) - 4♠ (minimum hand, no shortage) - 4NT - 5♥ (2 key cards, no queen) - 6♠. If South opens, then 1♦ - 1♠ - 4♣ (this time, South is not too strong to splinter) - 4♠ - 4NT - 5♥ - 6♠

[dake50]

I suspect you find some parts of an automatic relay
too slow getting info you want.
Try asks in the suits of concern on some "problem hands".
And second, accept bidding in a straight-jacket relay is just as you witness.
Hope the forced relay for all strong hands gets many right -- eventually.

[rbforster]

You can almost always solve this problem with more relays and memorisation. One solution is to allow for RKC and exclusion RKC as alternative asking options, instead of DCB/Spiral, by relayer. For example,

1♣...relays
relays...3♥ showing some shape

3♠ asks for strength, then DCB
3N to play
4♣ exclusion RKC, relay to 4♦ and then show which suit with the next 4 steps (responders' longest comes first, etc)
4♦ end relay, forces 4♥ to sign off
4♥-5♣ regular RKC (responders' longest suit comes first, etc)

You can make various improvements on this by breaking relays earlier, or maybe skipping only 1 step in an earlier relay to focus on a particular suit (which then can be used for stopper asks, RKCs, excl-RKC, etc depending on the subsequent auction).

[Cascade]

rbforster, on 2010-November-15, 02:51, said:

You can almost always solve this problem with more relays and memorisation. One solution is to allow for RKC and exclusion RKC as alternative asking options, instead of DCB/Spiral, by relayer. For example,

1♣...relays
relays...3♥ showing some shape

3♠ asks for strength, then DCB
3N to play
4♣ exclusion RKC, relay to 4♦ and then show which suit with the next 4 steps (responders' longest comes first, etc)
4♦ end relay, forces 4♥ to sign off
4♥-5♣ regular RKC (responders' longest suit comes first, etc)

You can make various improvements on this by breaking relays earlier, or maybe skipping only 1 step in an earlier relay to focus on a particular suit (which then can be used for stopper asks, RKCs, excl-RKC, etc depending on the subsequent auction).

You would need twelve steps for exclusion RKC - four to identify the trump suit times three to identify the exclusion suit.

You could do exclusion control ask with four steps by just identifying the exclusion suit. I suspect it is something like this that you mean.

Certainly by using an end relay or puppet terminator or whatever you want to call it you can free up many bids as specific asks.

In some situations I use discounted controls after the relayer essentially shows a singleton. This focuses on all of the other kings. Compared with RKCB which focuses on the trump queen. Typically we use this in a context in which the trump suit is known so that later we can get information about the trump queen. However if you used 3=2=1 controls you would gather information about queens on the way through.

Obviously these asks will be more efficient when the shape is resolved low. If the shape is resolved at 3♠ then assuming 3NT is a sign-off and 4♦ is a terminator then

4♣ showing a void with a relay as above.
4♦ terminator
4♥ Ask for spades
4♠ Ask for hearts
4NT Ask for diamonds
5♣ Ask for clubs

or some other order based on the suit lengths shown then it is easily possible for the higher asks to be too high to be effective.

Even lower shape resolutions there is the potential to run out of room, say 3♦

3♥ Void with relay
3♠ Ask for spades
3NT to play
4♣ Ask for hearts
4♦ Terminator
4♥ Ask for diamonds
4♠ Ask for clubs

Again the later asks could be more crammed than desirable but at least they are lower than standard RKCB.

Possibly the asks should always be in the order clubs, diamonds, hearts, spades giving a kickback type effect. Or maybe a better efficiency would be clubs, diamonds, hearts, spades but moving any singleton suit to the end so for example if the shape was 4513 then the order would be clubs, hearts, spades, diamonds since the singleton suit is much less likely to be trumps.

Amusing story. I once held void xxx in two suits and partner showed AKQ or none in both suits and I knew she had AKQ in only one of these suits so didn't know how high to bid until my RHO made a lead directing double and I bid the Grand knowing that opponent had some big cards opposite my void.

[mtvesuvius]

Something seems wrong with using S1 as showing a void. This seems inefficient, and won't come up nearly often enough to pay off.

I'd recommend something like:
4C - Begin Spiral-Scan/DCB
4D - End Signal (or Regular Blackwood)
4H - RKC - Longest Suit
4S - RKC - Lowest of the three remaining suits
4N - RKC - Middle of the three remaining suits
5C - RKC - Highest of the three remaining suits
5D - ERKC - Exclude Longest Suit
5H - ERKC - Exclude Lowest of the three remaining suits
...

For the ERKC calls, often it is very cramped, but IMO, it's not worth moving up in any stack. You can respond to ERKC in # of Controls, or just aces... Either way, I think this is better than using a very low bid to start ERKC type auctions.

[Echognome]

I know it's just nomenclature, but I always viewed denial cue-bidding and spiral scanning as different things. To me, spiral scanning is invoked when the trump suit has been agreed. Then the ask is forcing to the next level of trumps and you can ask specific questions, skipping suits as desired. Denial cue bidding is when trumps are not set, so once started, relay asker must always bid the next step in order to continue the ask. Any breaks are to set the contract.