[straube]

Hughes is a method similar to Denial Cue Bidding or Parity Cue Bidding. Crunch3NT told me about it and it looks like it saves room. Instead of relaying exactly which cards you have, it relies on relay captain being able to make inferences from his own holding...similar to RKC. Anyone have an opinion on the method? If I understand it correctly, it accepts some uncertainty to gain economy. Here's the method below...


This is hopefully our most common (alas, it isn’t) slam method. In Hughes, we don’t ask for controls. We still use controls to justify openings, Min/Max decisions etc. But with Hughes we go direct to an Ace Ask.

Responses to Ace Ask: Differs if shower is wide-range or not. If limited then:
Step 1....Odd number of Aces
Step 2....Even number of Aces, 1st step of Hughes DCB
Step 3....Even number of Aces, 2nd step of Hughes DCB etc

Responses to Ace Ask: Differs if shower is wide-range or not. If wide-range then:
Step 1....1 Ace
Step 2....0 or 3 Aces
Step 3....2 Aces, 1st step of Hughes DCB etc


Relay Bidding - Continued


Denial Cue Bidding - Hughes Method
a) Show DCB Suits in this order: Longest suit first, all ties highest ranking first. Next suit
continues. (Do show Singletons)

There are 12 honour cards to show in DCB, the four K, Q, Js. These are divided into 4 groups
.....of 3 as follows:
.......1st Group.....1st, 2nd, 3rd DCB King
.......2nd Group....4th DCB King, 1st DCB Queen, 2nd DCB Queen
.......3rd Group.....3rd DCB Queen, 4th DCB Queen, 1st DCB Jack
.......4th Group.....2nd, 3rd, 4th DCB Jack

c) Each relay asks about Each Group (1st three groups only) with following step responses:
.......1st Step:.......0 or All 3
.......2nd Step:......1st or both 2nd & 3rd
.......3rd Step:.......2nd or both 1st & 3rd
.......4th Step:.......3rd or both 1st & 2nd

The 4th group (Jacks) are shown as per Symmetric DCB – ie in order, no or yes.

d) Do Zoom on last step eg: If answering 1st Group:
.......1st Step:.......0 or All 3 of 1st group
.......2nd Step:......1st or both 2nd & 3rd of 1st group
.......3rd Step:.......2nd or both 1st & 3rd or 1st group
.......4th Step:.......3rd or both 1st & 2nd of 1st group AND 0 or All 3 of 2nd group
.......5th Step:.......3rd or both 1st & 2nd of 1st group AND 1st or both 2nd and 3rd of
2nd group etc

.......But, only zoom once to next group in Hughes (ie 2 sweeps)


e) Special Sequence When responding to Hughes DCB with known Singleton:
.......1st Group = KKK
.......2nd Group = QQQ
.......Then Symmetric DCB – Stop = Singleton K or Q, Pass = J or worse
.......Then Symmetric DCB Jacks

[Crunch3nt]

I find it pretty surprising this post has no replies as in my experience, "Hughes" is easily the most efficient DCB/Spiral Scan method I've come across.

I found it in "Building a Bidding System" book by Canadian Roy Hughes. He didn't give the method a name, so since my regular partner & I now use it, we named it after him. The major example in the book is a hand apparently from the final of the 2004 Olympiad in Istanbul between Italy and the Netherlands - but it fails to say who used this relay method - I guess one of the dutch pairs.

Anyway. Here is an example:

AQ10942 63 K54 A5

KJ AQ5 AQ983 864


North opens 1H showing spades, South GF relays and North shows 6=2=3=2 with a last bid of 3D.


1) Playing controls with symmetric DCB, the auction would continue:

3H 4D 5 controls
4H 5C A or K in Spades, and Diamonds, but not in Hearts
5D 5NT A or K Clubs, Queen of spades, but no Queen of Diamonds


2) Playing relay points with parity PCB:

3H 4D 9 Queen Points
4H 4S Even number of controls in Spades
4NT 5C Even number of controls in Diamonds
5D 5H Even number of controls in Hearts (Can't tell between KH & KD?)


3) Hughes:

3H 4D Even number of Aces, and the KD or both of KS and KH
4H 5C QS or both KC and QD
5D 5H 0 or all 3 of QH, QC, and JS
5S 5NT No JD (not sure you care - bid the grand anyway)

Hopefully I have these auctions right. Just trying to show Hughes efficiency not insult or mis-speak any one's methods.


Now for the bad news about Hughes. It is most efficient but we have found that when a weak hand is relaying a strong hand, we often can't decode the responses. Major problem. It does do much better when a strong hand relays a weak one. However we still like Hughes and use it in conjunction with other methods - the key being, the asker (captain) can tell when he/she can decode the responses - so you use Hughes when you can tell, and you use other methods when you cant.


Anyway, comments, thoughts?

[Cascade]

Crunch3nt, on 2011-January-14, 15:23, said:

Anyway. Here is an example:

AQ10942 63 K54 A5

KJ AQ5 AQ983 864


North opens 1H showing spades, South GF relays and North shows 6=2=3=2 with a last bid of 3D.


1) Playing controls with symmetric DCB, the auction would continue:

3H 4D 5 controls
4H 5C A or K in Spades, and Diamonds, but not in Hearts
5D 5NT A or K Clubs, Queen of spades, but no Queen of Diamonds


2) Playing relay points with parity PCB:

3H 4D 9 Queen Points
4H 4S Even number of controls in Spades
4NT 5C Even number of controls in Diamonds
5D 5H Even number of controls in Hearts (Can't tell between KH & KD?)


3) Hughes:

3H 4D Even number of Aces, and the KD or both of KS and KH
4H 5C QS or both KC and QD
5D 5H 0 or all 3 of QH, QC, and JS
5S 5NT No JD (not sure you care - bid the grand anyway)

Hopefully I have these auctions right. Just trying to show Hughes efficiency not insult or mis-speak any one's methods.

My normal method after controls is to include the queens in the first round of denial cue-bids and to pass through a six card suit twice immediately.

3H 4D five controls
4H 5D two spade honours, diamond honour and no heart honour

which happens to work remarkably well for this particular hand.

A major problem with denial cue-bid is that often in the early round(s) you are giving partner information that he already knew from the control response or didn't need to know precisely.

Relay points attempt to get around this by showing queens in the control (relay point) response.

We include the ambiguity of queens in the first round of denial cues in an attempt to give partner more information.

Hughes just doesn't ask for controls gathering raw information immediately with some possibility of ambiguity.

It must be possible to quantify the chance of ambiguity and efficiency of these methods.

[Fluffy]

Instead of asking for everything I have a method that identyfies wich suit, so in your example, after 3♦, instead of just bidding 3♥ with any hand, south sets up a trump suit (longest first) with 3♥=♠, 3♠=♦, 4♣=♣ 4♦ = ♥ and is a keycard blackwood on that suit.

I haven't tried any spiral scan systems yet, only the CRASH convention, and we soon realiced that stablishing trumps was a big improvement over it. Cos it lets you stablish wich bids are relays and wich bids are to play

In your exaple 3♥ would be keycard setting spades as trumps, answer is 4♦ (2 and the queen), next comes 4♥ wich asks further on the way you want, but will soon discover that ♦K is there but not ♣K.

My methods after the keycard blackwood are probably not optimum, maybe a version of hughes after it is better.

[straube]

Crunch3nt, on 2011-January-14, 15:23, said:

Anyway. Here is an example:

AQ10942 63 K54 A5

KJ AQ5 AQ983 864


North opens 1H showing spades, South GF relays and North shows 6=2=3=2 with a last bid of 3D.


2) Playing relay points with parity PCB:

3H 4D 9 Queen Points
4H 4S Even number of controls in Spades
4NT 5C Even number of controls in Diamonds
5D 5H Even number of controls in Hearts (Can't tell between KH & KD?)


3) Hughes:

3H 4D Even number of Aces, and the KD or both of KS and KH
4H 5C QS or both KC and QD
5D 5H 0 or all 3 of QH, QC, and JS
5S 5NT No JD (not sure you care - bid the grand anyway)

I think the auction for PCB is
3H 4D 9 Queen Points
4H 4S even number of controls in Spades
4N 5D odd number of controls in Diamonds, Even number of controls in hearts
5H 5S even number of controls in clubs
5N 6D the diamond control is the king

PCB usually gives me a complete picture of partner's hand at a lower level than does DCB. On the downside, with PCB, the picture is sometimes very fuzzy until after the king parity ask; especially if we are light on QPs, partner may have several possibilities. I sometimes find myself running out of room and guessing that partner has the right cards. OTOH, it seems very strong for grand slam bidding.

DCB tends to give more concrete information earlier than PCB, but it seems to take too many steps in total...especially if it has to scan the singleton.

Hughes seems to blend giving concrete information while making more use of the cards that the captain is looking at...thus it seems efficient.

I wonder though if captain can become pessimistic too early in the process of Hughes. For example, if we're off an ace and captain doesn't have sufficient strength, he may give up too early....just when RR has all the rest of the controls/QPs. I like knowing QPs because sometimes even if I don't know the exact cards, I know the partner will have sufficient strength for a small slam to have play. The other thing about Hughes that I wonder about is if captain gets an answer he likes (first round) and then finds RR has no other cards, he is potentially too high.

In the example hand, PCB gets up very high because responder's 9 QPs take up a lot of room (4D). The good news is that with 9 QPs there is very likely a slam in the offing. If opener had more of those QPs though (move a few from responder's to opener's hand), then the control bidding would end lower.

I've been wondering if Hughes and PCB have different advantages in different situations, but not sure that they do. Both seem to work best when captain is bigger. I'm also wondering if one might be better than the other depending on when pattern resolution occurs. For instance, if pattern is known by say 3D, then maybe PCB is better, but if by 3H, then maybe Hughes is better. I.e. the less room one has, the more uncertainty one might be willing to accept.

What's Adam's opinion on Hughes?

[awm]

The way you've described the Hughes method seems to involve a lot of "run-ons" when you have the highest step. This is potentially quite problematic because the highest step doesn't necessarily show the best hand (in fact, describer's hand was never really limited in any way). So you'd probably need to insert more "stops" in the method.

In any case, I think Hughes is more efficient for locating specific cards. However, this ignores the fact that the overall strength + shape information is frequently enough to determine the level without locating any specific cards at all. Honestly I haven't had much trouble with my QP method, and even the control-based method works well if you play a slightly older version (where A+K and out is not a GF opposite 16 hcp, "minimum" can conceal five controls if A+A+K and out, and "maximum" can be three controls).

[straube]

Thanks. Side question regarding your PCB method. I initially thought that you zoomed from the king parity ask into jacks and that consequently you stopped with the "odd" king and skipped with the "even" ace or queen in order to save space. Having been corrected that you don't zoom into jacks, then I take it that you stop with the ace or queen and skip one step with the king (to save room 2/3rds of the time). Is that right? Do you include the AKQ combination with the AQ or with the K?

[Free]

The way I always played AKQ-asks with DCB (TH = top honour):
... - 4♦ (... ; 9 AKQ-points)
4♥ - 5♣ (r ; 1/2 TH ♠ and ♦, 0 in ♥)
5♦ - 5NT (r ; 1/2 TH ♣, 2 TH ♠, 1 TH ♦)
6♣ - 6♦ (r ; not ♥J)
6♥ - 6♠ (r ; 1 TH ♣)
...

If opener had ♣KQ, he'll bid 6NT denying the ♠J (which we hold, luckily), so it's safe to continue relaying - but just barely!

This method scans for all 3 tophonours immediately, not A+K only during the first scan.

[Zelandakh]

Fluffy, on 2011-January-15, 09:05, said:

Instead of asking for everything I have a method that identyfies wich suit, so in your example, after 3♦, instead of just bidding 3♥ with any hand, south sets up a trump suit (longest first) with 3♥=♠, 3♠=♦, 4♣=♣ 4♦ = ♥ and is a keycard blackwood on that suit.

I haven't tried any spiral scan systems yet, only the CRASH convention, and we soon realiced that stablishing trumps was a big improvement over it. Cos it lets you stablish wich bids are relays and wich bids are to play

I think most relayers have a method of relay breaks whereby they can set the trump suit and begin normal keycard responses. The most common is 4-way key card Blackwood (bids of 4D thru' 4NT set suit below) but there are plenty of alternatives. It is an even bigger improvement here to have 2 (or more) ways of conducting the slam auctions depending on which specific cards you are searching for.