[straube]

We've traditional played PES (Polish End Signal) where after shape resolution 4D requests 4H for pass or correct. I think instead that it makes sense for S2 (step 2) to be the end signal. Very often this equates to the same thing as pattern may be resolved at 3S or 3N, but occasionally we might pick up some useful room.

Now I know that if we use S2 for one thing that we have to surrender it for something else. Many folks play (for example) that S2 below 3N asks for a stopper in partner's fragment. For others it might be RKC for partner's longest suit. Etc. But here are some upsides in mostly decreasing orders of usefulness...

1) What interested me firstly in this idea was the ability to stop specifically in 4D on certain hands. No, I'm not sacrificing machinery to get out of game. I think instead that captain may be able to relay more hands when he really doesn't have the strength for it...as long as he has diamonds and as long as responder won't get too high too fast without diamonds. Say for example that we play a system in which 1C-2C is a semipositive with spades and a minor. Now holding diamonds, opener can relay knowing that he can break relay if partner has clubs or he can potentially sign off in 4D (or even 3D) when responder has diamonds. Without this ability to relay, opener may have to rebid 2N over 1C-2C and a potential diamond fit is missed. It also frees up opener's relay breaks. He no longer has to show a minimum hand with diamonds.

2) When the slave hand wants to super-accept, the super-acceptance starts earlier. Say the bidding goes 1C-3D where 3D is GF 3145. 3S announces the desire to sign off somewhere and responder super-accepts starting with 4C as opposed to 4S. Three steps is a lot of room for DCB etc.

3) Signing off in 4m. In that same example hand 1C-3D, 3S-3N, 4C would offer to play 4C. Sometimes 3N looks bad, right? Responder may carry on to 5C.

4) Responder gets to know more often when the contract has been decided. Say the bidding goes 1C-3D (3145), 3S-3N, 4H and responder is looking at a hand just shy of a super-accept and perhaps he has the heart queen or king. He can show a max non-super-accept if he wants.

5) Captain can sometimes force the slave hand to play 3N.

[rbforster]

Aside from stopping in 4m (which I can't really speak to, seems specific to your semi positive responses), it seems the main benefit is getting responder to show his super accept QPs at a lower level over openers attempted sign off. As you say, this matters most when shape resolves at 3♦, so step 2 is 3♠ rather than higher where things mostly equivalent with the 4♦ end relay.

I would contend this is not that much of a benefit. Consider two cases when you resolve shape cheaply:

1. If opener is minimum and wants to play NT, he'll bid 3N rather than asking and then responder super accepts by bidding on with 4♣=9 QPs, etc. If opener bid 3♠ in your scheme, 4♣ is the first super accept step (instead of puppeting to 3N) and you get the same outcome.

2. If opener is minimum and wants to play a suit game, he can still ask pretty aggressively for QPs since he's doing so at a very low level. Suppose the sign off system after a QP ask is that any non-relay is to play. So if instead of using an end signal, if opener asks for QPs with 3♥ and hears a response of 4♣ or less, he can still sign off anywhere. If he hears 4♦, responder is showing 9 QPs (at least in my system that starts with 6 minimum) and would have forced higher anyway. Alternatively, if he had tried to sign off with 3♠ in your system, the first super accept is 4♣ and is only one step cheaper for showing 9 QPs.

In short, it seems on the QP resolution front, you're giving up your second step for something that never helps in NT and only saves one step in suits when responder is strong.

[akhare]

rbforster, on 2013-July-25, 15:41, said:

Aside from stopping in 4m (which I can't really speak to, seems specific to your semi positive responses), it seems the main benefit is getting responder to show his super accept QPs at a lower level over openers attempted sign off. As you say, this matters most when shape resolves at 3♦, so step 2 is 3♠ rather than higher where things mostly equivalent with the 4♦ end relay.

I would contend this is not that much of a benefit. Consider two cases when you resolve shape cheaply:

Another downside is consider to consider that it pushes the RKC asks by 1 or 2 steps (except for auctions that terminate in 3♠).

[benlessard]

The idea of using the 3S as a signoff is interesting and similar to using a "non-serious" 3S/3NT. However I do believe that its only when NT as already been bid that its worth while, becasue you dont want the slave hand to play 3nt if you can avoid it. Also the extra bids that you gain are higher in the food chain and less valuable.

Ex you in 3D and 5521 is showned,

3H = QP
3S = H rkc
3Nt to play
4C = S rkc
4D = D rkc
4M to play
4Nt quant

if you keep 4D as PES

4H =D keycard
you gain 4H/4S/4NT to show voids instead of being soff/quant.

but being able to keycard in D at 4D instead of 4H is more frequent & valuable I think, this depend on the opportunity to show a void earlier.
--------------------------------------
If you used 3S as relay to 3nt to PES

3NT become H rkc
4C = S rkc
4D = D rkc
and you have the voids.

You are using the fact that 3Nt as a non forcing bid is non optimal.

However NT must been grab otherwise you are going to wrongside a lot of contracts.

[straube]

rbforster, on 2013-July-25, 15:41, said:

Aside from stopping in 4m (which I can't really speak to, seems specific to your semi positive responses), it seems the main benefit is getting responder to show his super accept QPs at a lower level over openers attempted sign off. As you say, this matters most when shape resolves at 3♦, so step 2 is 3♠ rather than higher where things mostly equivalent with the 4♦ end relay.

I would contend this is not that much of a benefit. Consider two cases when you resolve shape cheaply:

1. If opener is minimum and wants to play NT, he'll bid 3N rather than asking and then responder super accepts by bidding on with 4♣=9 QPs, etc. If opener bid 3♠ in your scheme, 4♣ is the first super accept step (instead of puppeting to 3N) and you get the same outcome.

2. If opener is minimum and wants to play a suit game, he can still ask pretty aggressively for QPs since he's doing so at a very low level. Suppose the sign off system after a QP ask is that any non-relay is to play. So if instead of using an end signal, if opener asks for QPs with 3♥ and hears a response of 4♣ or less, he can still sign off anywhere. If he hears 4♦, responder is showing 9 QPs (at least in my system that starts with 6 minimum) and would have forced higher anyway. Alternatively, if he had tried to sign off with 3♠ in your system, the first super accept is 4♣ and is only one step cheaper for showing 9 QPs.

In short, it seems on the QP resolution front, you're giving up your second step for something that never helps in NT and only saves one step in suits when responder is strong.

I think this is good criticism. Thank you for it. I'm still keen on the idea for my top-listed reason, but your analysis mostly invalidated reason #2.

[straube]

benlessard, on 2013-July-25, 21:02, said:

The idea of using the 3S as a signoff is interesting and similar to using a "non-serious" 3S/3NT. However I do believe that its only when NT as already been bid that its worth while, becasue you dont want the slave hand to play 3nt if you can avoid it. Also the extra bids that you gain are higher in the food chain and less valuable.

Ex you in 3D and 5521 is showned,

3H = QP
3S = H rkc
3Nt to play
4C = S rkc
4D = D rkc
4M to play
4Nt quant

if you keep 4D as PES

4H =D keycard
you gain 4H/4S/4NT to show voids instead of being soff/quant.

but being able to keycard in D at 4D instead of 4H is more frequent & valuable I think, this depend on the opportunity to show a void earlier.
--------------------------------------
If you used 3S as relay to 3nt to PES

3NT become H rkc
4C = S rkc
4D = D rkc
and you have the voids.

You are using the fact that 3Nt as a non forcing bid is non optimal.

However NT must been grab otherwise you are going to wrongside a lot of contracts.

I had thought to use S2 for mostly hands that belonged in suits. Perhaps rare hands where captain thought RR should play the hand. I had thought to keep 3N by captain natural although it occurred to me that it could be an asking bid when S2 was available. Perhaps 3N should be an asking bid when S2 is available and NT has already been bid. At this stage, I'd rather not add that complication.

[Free]

Seems like an interesting idea, especially when step 2 is below 3NT. In that case, you only need 1 end signal and you can use 3NT as a relay.

For example, 3♦ showing a 5431:
Original scheme:
3♥ = QP ask
3♠ = RKC 1
3NT = signoff
4♣ = RKC 2
4♦ = Terminator
4♥ = RKC 3
4♠ = RKC 4

New scheme:
3♥ = QP ask
3♠ = Terminator
3NT = RKC 1
4♣ = RKC 2
4♦ = RKC 3
4♥ = RKC 4

RKC 1 is 1 step higher, but RKC 3 and 4 are 1 step lower, so overall more efficient.

An additional advantage is when RR has to zoom, he can do it lower improving our denial cuebidding sequences.

When step 2 is 3NT or higher, your RKC gets worse (because you still want 3NT to be signoff, so 4♣ is the lowest Terminator) but you keep the zoom + denial cuebid advantage, plus you can signoff in 4♦. Still slightly better imo.

Note: when you're very low, at 3♣ for example, then I'd use 3♠ as an end signal, not step 2.

[hrothgar]

I'm not sure how this works out from a useful space perspective

Consider the following: Whenever feasible, a bid of 3♠ is the end signal.
RR will rebid 3N unless he holds 3+ QPs more than base (in which case he zooms)

Essentially, we're losing the normal use of the "3♠" bid, but we're gaining a 3N bid and a 4♦.

From the a purely constructive perspective, this seems like a bad trade

Moreover, auctions like

3♦ - 3♠
3N - 4♠

will contain a lot of dead space

I'd be leery about using this

[Free]

hrothgar, on 2013-July-26, 06:33, said:

Essentially, we're losing the normal use of the "3♠" bid, but we're gaining a 3N bid and a 4♦.

From the a purely constructive perspective, this seems like a bad trade

An additional advantage I just realized: opps have one fewer opportunity to Dbl for the lead (3♠ opposite 4♦+4♥). The disadvantage though is that when you want to signoff in 3NT you need to go through the end signal, which gives away an opportunity to Dbl. So I guess that cancels out.

The so called dead space isn't a big issue imo, and it's very helpfull when RR zooms (3 steps lower than after 4♦ terminator, same as after 3NT signoff). And we're not that fond of RKC anyway...

I wouldn't use 3♠ whenever available btw, I'd play:
step 1 (not 3NT) is ALWAYS QP ask
IF available: 3♠ = Terminator
ELSE: 3NT = signoff, step 2 = Terminator
other bids are RKC's

[rwbarton]

hrothgar, on 2013-July-26, 06:33, said:

Essentially, we're losing the normal use of the "3♠" bid, but we're gaining a 3N bid and a 4♦.

You also gain 3♠-3N-4♣ and 3♠-3N-4♦ to use as signoffs or more asking bids. Assuming you want 4 RKC asking bids, you could arrange them like this:

3♦ shape resolution then

3♥ = QP ask
3♠ then pass or bid game = signoff
3♠ then 4♣ = RKC 1
3♠ then 4♦ = RKC 2
3N = RKC 3
4♣ = RKC 4

so your asking bids come at 3N, 4♣, 4♣, 4♦. Comparing to the original

3♥ = QP ask
3♠ = RKC 1
3N = signoff
4♣ = RKC 2
4♦ = PES
4♥ = RKC 3
4♠ = RKC 4

with asking bids at 3♠, 4♣, 4♥, 4♠, it looks like an improvement.

(This is ignoring potential wrong-siding considerations, of course...)

[straube]

I'd like to focus a bit of attention on the #1 reason I'm interested in this...which is handling semipositives or hands that might be semipositives. For example, we play IMPrecision responses to 1C.

1C-2C=semipositive or light GF with an unbalanced hand with 5 clubs or 4+ diamonds. At the moment our relays continue thusly...

1C-2C
.....2D-GF relay
..........2H-6C
..........2S-5D
..........etc-reverser diamonds

such that...

1C-2C
.....2N-could be 4351 or 4441 or 4252 etc misfitting and limited.

Now if we reorder the relays such that...

1C-2C
.....2D-GF or 5+ diamonds
..........2H-6C
..........2S-4D, higher or no shortness
..........2N-4D, lower shortness
..........etc-5D

we can find a variety of club and diamond fits.

1C-2C
.....2D-2H
..........3C-could be that 4252. Now we play 3C instead of 2N when pd has Axx x xxx Jxxxxx.

1C-2C
.....2D-2H
..........2N-3D opener shows 5D by breaking relay and responder chooses a 5-3 diamond fit

1C-2C
.....2D-2N
..........3D-could be 4351 opposite 3145

That sort of thing. So re-ordering how responder relays out is the main thing. But S2 handles those situations where responder has 5+ diamonds and we've discovered a 10-cd diamond fit but still aren't sure about the 5-level. Infrequent but possible.

1C-2C
.....2D-3C
..........3H-3S
...............4D

I'm not sure what semipositives Moscito is using at the moment, but I suspect the same sort of thing can be done for it.

[benlessard]

In my system opener has a split range 11-14 or 18-22. So when a responder that is GF (12+) make a 3NT or 4M signoff opener holding a 18-22 (but he is assumed to have a 11-14) need to make another move. The gain of a PES 3S is really when responder would make a 4M signoff. bidding 4C,4D,4H,4S instead of 4Nt+ is pretty good even if trumps isnt set. But weve rarely got bad board when GF vs 18-22 so we didnt care that much.

I already play a complex system and 3Nt that is sometimes a keycard and sometimes to play is a bit scary.

Gadget like these you forget them 5% of the times and you a loser in the end.So maybe if I start a new partnership I would play taht way but once your used to an old method changing can be costly.

[straube]

benlessard, on 2013-July-26, 14:59, said:

I already play a complex system and 3Nt that is sometimes a keycard and sometimes to play is a bit scary.

It concerns me a bit, too. I wasn't really thinking of playing that, but I understand how others could gain a step out of it.

[rbforster]

I think you need to decide what your priorities are, since specifically on auctions resolving shape at 3♦, there are lots of sequences one can use - probably more than you can find sensible meanings for. Consider one where you want cheap RKC asks:

3♥ QP ask
3♠ forces 3N unless extras
....P sign off
....4♣ sign off
....4♦ sign off
....various exclusion asks
3N RKC ♥
4♣ RKC ♠
4♦ RKC ♣
4♥ RKC ♦
etc

This gets you a 3 steps of space to respond to the RKC ask at or below game in each suit (note the order of RKCs are fixed, not relative to responder's suit lengths, 2+Q goes past). If you wanted less space for RKCs and more space for some thing else (I.e. RKC asking was willing to force to the 5 level opposite some replies), you could make all the bids starting with 3♠ puppets to the next step, after which opener would ask a range of questions with each of his bids.