[Cyberyeti]

A sim is only as good as its assumptions, for the single suited style of SJS IMO the spade suit should be:

4 of the top 5 if 6 cards long.

AKQ or 4 of the top 5 if 7 cards long

AK or 3 of the top 4 if 8 cards long

I suspect you have included much weaker spade suits than this.

[smerriman]

 thepossum, on 2019-October-25, 15:06, said:

Currently in hiding for a while but here are some Sim results which kind of back up any side of the argument (using bdeal). Not really good enough for me to bid past game but each to their own

Assumptions by North of South's hand

Sim 1: 15+ HCPs and splinter clubs (Gib doesnt know), 6+ spade points
Sim 2: 15+ HCPs and no splinter assumption, 6+ spade points
Sim 3: 17+ HCPs and no splinter assumption, 6+ spade points

You're forgetting you bid 3♠.

You're right in that bidding 2♠ shows a good suit, with at most one missing honor (amongst other things).

But after partner doesn't raise spades to show the other honor, bidding 3♠ confirms a solid suit, as per all of the posts + links above.

If you had a splinter, GIB would know, because you would have bid it!

Eg, Karen Walker's site:

Quote

A simple rebid of your jump-shift suit (1C-2S-3C-3S) shows a long, solid suit with more high-card strength than the jump to four.

So all your sims need to be adjusted to give South a solid spade suit to give accurate results.

[edit] I may be wrong, but I have a feeling you may be confusing 'splinter' with 'shortness'. You said you originally planned to show a splinter - but you don't have a splinter. (Splinter means shortness in the suit bid, AND support for your partner's suit.) If you had a splinter, you'd be in a diamond slam, not a spade slam, so sim 1 shouldn't be counting tricks in spades.

[smerriman]

The definition GIB uses for solid suit can be found here:

https://www.bridgeba...escriptions.php

Quote

solid 6-card
AKQTxx or AKQJxx
solid 7-card
AKQxxxx
solid 8-card
AKJxxxxx or AKxxxxxxx or better

I therefore ran a sim with these conditions:

Quote

predeal north HAJ2, DKQ963, CAJT92
predeal south SAK

condition hcp(south)>=16 and (
(spades(south)==6 and (hcp(south,spades)==10 or hcp(south,spades)==9 and hascard(south,TS)))
or
(spades(south)==7 and hcp(south,spades)>=9)
or
(spades(south)==8 and hcp(south,spades)>=8)
or
(spades(south)>=9)
)

The results over 500 deals:

13 tricks: 48.8%
12 tricks: 34.8%
11 tricks: 12%
10 tricks: 4.4%

6♠ is a terrible bid - but only because of the high chance of making 7. After hearing 3♠, North should never be stopping at less than a slam - GIB just isn't good enough to find 7.

Edit - I like Cyberyeti's definition more, which could have one loser - adjusting the sim to use those figures still clearly gives 6♠ as the minimum contract you want to reach.

[smerriman]

While I'm here, I actually have my own question about Soloway and Karen Walker's interpretation of it that I've always been meaning to ask.

She says one option is:

Quote

A slam-try or better hand (16+ pts.) with a good 5+-card suit and length in opener's suit.

She also says:

Quote

A raise of responder's suit promises one of the top three honors and at least doubleton support (Qx or better, although some players advocate raising with even a singleton honor).
A failure to raise responder'ssuit DENIES one of the top three honors in his suit.

With the hand described above, after:

1♦ - 2♠ - 3♠

Is a new suit still a splinter for diamonds, or a cue in spades? If the latter, doesn't that mean we could be playing in a bad spade fit (possibly 5-2 - or even 5-1!) instead of a great diamond fit? If the former, how are spades set as trumps? If I need to bid 4♦ to set diamonds, what do I bid when wanting to play in spades but have xx in clubs?

I know there's an alternative form of Soloway with relay responses, but the version she mentions must be playable.

[hrothgar]

The 3♠ rebid sets the trump suit

I believe that new suit is traditionally used as a cue bid

[Stephen Tu]

 smerriman, on 2019-October-25, 16:56, said:

Is a new suit still a splinter for diamonds, or a cue in spades? If the latter, doesn't that mean we could be playing in a bad spade fit (possibly 5-2 - or even 5-1!) instead of a great diamond fit? If the former, how are spades set as trumps?

There's like zero reason for opener to raise immediately with only 1 or 2 trumps. Assume spades are trumps for now.

If you want to retain 4d to show diamond fit, then perhaps utilize 3nt as a "serious 3nt" type waiting bid demanding cue?

[smerriman]

 hrothgar, on 2019-October-26, 08:31, said:

The 3♠ rebid sets the trump suit

 Stephen Tu, on 2019-October-26, 12:19, said:

There's like zero reason for opener to raise immediately with only 1 or 2 trumps. Assume spades are trumps for now.

So the system on that page is plain wrong?

As is the later example:

Quote

A jump-shift is often the best way to find out whether partner can fill in your long, almost-solid suit (a 6+ card suit missing only one of the top three honors). With Hand #5, a grand slam (or even a small) probably depends on partner holding the diamond king. Since it's unlikely partner will be able to raise a 1D response, the best way to find out about his support is to set the suit by jump-shifting to 2D. If partner has at least Kx, he'll raise to 3D. You can then Blackwood and bid the laydown 7NT if partner shows two aces and another king. If partner doesn't raise diamonds, you can still Blackwood to find out about a possible singleton king.

[thepossum]

 hrothgar, on 2019-October-25, 15:24, said:

I'm not sure what hands match the constraints in your sims. For example,

What does "6 spade points" mean?
How many spades does the bid promise?

Once again, the 3!S rebid shows a pretty specific hand type

The article that Chasetb link suggested the following as a prototypical example...

♠AKQJ843 ♥6 ♦A84 ♣64

Does it really matter hrothgar. You are going to challenge my sims now (I will provide my basic back of the envelope sims if you like that back up common sense). Maybe one day you will learn other people have valid opinions other than yours. WOuld you prefer I run a one hand sim with your prototype to back up your limited views

And I really don't understand why you spend so much time defending a bot's bids when you sepnt so much time over the last year and ahalf bagging it.

Maybe there is a reason

Maybe you forgot this thread wasnt even about the stupid bot until you in your usual overdefenseive attacking way siezed on something and launched an attack. The way to bully anyone who dares to question anything on this site

Are you their attack dog

Or maybe its just that you are feeling insecure that there maybe somebody around here who can actually challenge your self-proclaimed expertise on everything, and is prepared to stand up to you - mr expert on everything

Updated results for Hrothgar's ideal hand

S by S, number of tricks: 25328 / 2000 = 12.664 Ań 0.0314122 (95% conf.) min: 8 max: 13 sdev: 0.716313
chance to win 1S by S: 2000 / 2000 = 1 Ań 0 (95% conf.) min: 1 max: 1 sdev: 0
chance to win 2S by S: 2000 / 2000 = 1 Ań 0 (95% conf.) min: 1 max: 1 sdev: 0
chance to win 3S by S: 1999 / 2000 = 0.9995 Ań 0.000980331 (95% conf.) min: 0 max: 1 sdev: 0.0223551
chance to win 4S by S: 1996 / 2000 = 0.998 Ań 0.00195919 (95% conf.) min: 0 max: 1 sdev: 0.0446766
chance to win 5S by S: 1977 / 2000 = 0.9885 Ań 0.00467556 (95% conf.) min: 0 max: 1 sdev: 0.10662
chance to win 6S by S: 1772 / 2000 = 0.886 Ań 0.0139369 (95% conf.) min: 0 max: 1 sdev: 0.317811
Number of hands dealt: 2000. CPU time usage [sec]: 6.12 (x8 threads).


Note, in response to your request for info on my earlier assumptions with rather broader range of hands than a single prototype here they are

return S:spades()>5 and S:S():hcp()>5 and S:hcp()>14/16 (depending on hcp/total point assumptions)
-- and S:clubs()<2 (if I put a splinter in)

6+ spades and 6+ HCPs in spades (eg AQ, KQJ min etc)

I could of course further constrain my assumptions if you like in order to confirm (or at least provide evidence to back up) your expert opinion

Failing that I will just save time and not bother playing with (my humble toy) sims and just put your assessment into my results

begin sim
if hrothgar.hasOpinionOn(hand) then
return hrothgar.getassessment(hand)
else
return runSim(hand)
end if
end sim

[sfi]

 smerriman, on 2019-October-25, 16:56, said:

With the hand described above, after:

1♦ - 2♠ - 3♠

Is a new suit still a splinter for diamonds, or a cue in spades? If the latter, doesn't that mean we could be playing in a bad spade fit (possibly 5-2 - or even 5-1!) instead of a great diamond fit? If the former, how are spades set as trumps? If I need to bid 4♦ to set diamonds, what do I bid when wanting to play in spades but have xx in clubs?

Back when I played this, we always raised on a top honour - even a singleton honour. Responder then could bid NT to show the balanced hand or opener's suit to show a fit. Opener can then show an actual fit for partner by rebidding responder's suit. Any other bids were cuebids for the last naturally bid suit. To answer your last question, both 4C and 4H would be cues for spades. If I didn't have a control in either suit, I probably should not have bid 2S in the first place.

We never did show splinters, but there is merit in that.

The big problem with playing Soloway Jump Shifts is that they are really low frequency. That means that people start to stretch the hands on which they are used, which reduces their value when they come up. I haven't played them for decades and I can't remember a hand where I wished we were playing them.

[hrothgar]

 thepossum, on 2019-October-26, 16:53, said:

Does it really matter hrothgar. You are going to challenge my sims now (I will provide my basic back of the envelope sims if you like that back up common sense). Maybe one day you will learn other people have valid opinions other than yours. WOuld you prefer I run a one hand sim with your prototype to back up your limited views

Possum

It is not enough to "run a sim".

You need need to make sure that the people that you are communicating with have confidence that your simulation accurately describes the question that is being addressed. This is why I (and a lot of other people on these types of threads) make a deliberate practice of posting our code or pseudo code so other folks can see precisely how we are framing the problem and what assumptions we are making.

I will note in passing that I was not the only one who thought that your sim was orthogonal to the topic being discussed

Quote

And I really don't understand why you spend so much time defending a bot's bids when you spent so much time over the last year and a half bagging it.

There is a difference between "defending a bots bidding" and pointing out that your analysis is wrong.
I think that both of you have room for a world of improvement.

Quote

Are you their attack dog

I am hardly BBO's attack dog. I have heaped all sorts of criticism on BBO over the years (An awful lot of which had focused on their decision to carry forward with the GIB code based rather than partnering with Jack or WinBridge or most anything)

Quote

Or maybe its just that you are feeling insecure that there maybe somebody around here who can actually
challenge your self-proclaimed expertise on everything, and is prepared to stand up to you - mr expert on everything

Perhaps it's time for you to up your meds (cause they don't seem to be working too well right now)

[smerriman]

 thepossum, on 2019-October-26, 16:53, said:

return S:spades()>5 and S:S():hcp()>5 and S:hcp()>14/16 (depending on hcp/total point assumptions)
-- and S:clubs()<2 (if I put a splinter in)

See my posts.

clubs()<2 is not a splinter; clubs()<2 and diamonds()>=4 is a splinter (and would only apply if you had bid 3♣). It looks like that misunderstanding is the primary reason you got unstuck by bidding 2♠, thinking you had a splinter hand in the first place, and your confusion over why you then couldn't show your club shortness (because it promises a splinter for diamonds).

And that sim still doesn't match the 3♠ definition.

[smerriman]

 sfi, on 2019-October-26, 17:32, said:

Back when I played this, we always raised on a top honour - even a singleton honour. Responder then could bid NT to show the balanced hand or opener's suit to show a fit. Opener can then show an actual fit for partner by rebidding responder's suit. Any other bids were cuebids for the last naturally bid suit. To answer your last question, both 4C and 4H would be cues for spades. If I didn't have a control in either suit, I probably should not have bid 2S in the first place.

We never did show splinters, but there is merit in that.

The big problem with playing Soloway Jump Shifts is that they are really low frequency. That means that people start to stretch the hands on which they are used, which reduces their value when they come up.

Thanks, that all makes sense.

I agree regarding your last paragraph - but I also think that weak jump shifts are even lower frequency, and people start to stretch the hands they bid them on even more. So if the choice is between those two, I'd rather go strong. (For jump shifts to the three level, invitational is clearly better than either.)

[sfi]

 smerriman, on 2019-October-26, 23:04, said:

I agree regarding your last paragraph - but I also think that weak jump shifts are even lower frequency, and people start to stretch the hands they bid them on even more. So if the choice is between those two, I'd rather go strong. (For jump shifts to the three level, invitational is clearly better than either.)

I'm not sure if they are lower frequency, but they may be. I agree they are poorly defined and it's often hard to know when to go on with a good hand.

There is a lot of merit in invitational jump shifts at the two-level as well - for example 1C-2H. Aside from describing a common type of hand well, it helps other auctions. Consider 1C-1H; 1S-2H. Now opener knows that responder isn't being conservative with a potential invitation. Also, you now get to use 1C-1H; 1S-3H as a strong single-suited hand. And you know it's strongly one-suited because you didn't go through 4th suit-forcing (or whatever XYZ mechanism you have here).

Fundamentally, being able to divide the range into three (weak, invitational and game-forcing here) and immediately show the middle one often clarifies the later auction. The negative inferences when you don't show that range can be quite valuable.

Invitational jump-shifts are somewhat "flavour of the month" at the moment. But they make sense in theory and work well in practice, so I'm happy when I get to add them to a system.

[Zelandakh]

This thread started with:

 thepossum, on 2019-October-23, 23:20, said:

Hi everyone

Despite having played GiB 2/1 (including Soloways) for almost 18 months, they are still quite a mystery to me. I'm still trying to get a handle on the rebids and sequences and whether or not I have used them appropriately - starting to discover there are occasions I have used one with an inappropriate shape/strength

The following hand caused some problems finding the right contract - me and GiB playing from an archived Bermuda bowl match. (Edited to avoid confusion, although one BB team ended up in Spades too - but mine and GiB's contract was worse than the BB team )

I would be grateful for any comments

I had thought I was almost strong enough and planned to show the splinter in clubs. However you need support for diamonds. This left showing my spades again or bidding 3NT. I rebid spades which didn't work out well.

Thanks for any comments and guidance on Soloways.

 hrothgar, on 2019-October-23, 23:37, said:

Playing a Soloway SJS you want to have one of three different hand types (which will be clarified by your rebid)

1. A self sufficient Trump suit (good chance for zero losers opposite a stiff in partner's hand)
2. A fit showing SJS
3. A NT oriented hand

Note that in each of these cases, the player making the SJS should know what the strain is going to be before making the SJS/seeing responder's rebid.
If you don't know what the strain is going to be, then your hand does not qualify for the SJS.

I think we could have stopped the thread here and everything would have been perfect. The description of SJSs is clear, accurate and educational. Unfortunately the follow-up was:

 thepossum, on 2019-October-25, 03:54, said:

As others have indicated (thanks for all the advice) this hand was not strong enough or well suited for a Soloway SJS since you end up in a slam (Spades, Diamonds, No trumps) all of which are a disaster. Best contract was 3NT but at least I wasn't alone ending in Spades - I would have been happy with 4S - nothelped playing with GiB which figures a void in trumps is a good slam chance

 thepossum, on 2019-October-25, 03:54, said:

PS I must say I am disappointed that a very interesting thread about Soloways has been ruined yetagain by a stupid comment from someone. Par for the course in tho place. Almost every thread I have posted over the last 18 months has been ruined by some ignorant disrepsectful moronic comment from someone

This unfortunately seems to have led to the thread turning nasty, both because the former comment shows a lack of understanding of hrothgar's initial explanation - when you make the SJS you know the strain and that is specifically spades after a subsequent 3♠ rebid - and because the latter was obviously uncalled for (which comment is even meant by this?). Possum, I honestly think you would do well to go back and re-read the thread from the beginning dispassionately. I think you would then see that Richard's intention was to be helpful. If you consistently react aggressively after such answers then you might well find that those constructive answers disappear and you only receive critical ones or simply get ignored. That would probably not be good for you or BBF. I will certainly be watching now to decide if I want to post in your threads constructively or not.

[thepossum]

 Zelandakh, on 2019-October-28, 10:19, said:

This thread started with:



I think we could have stopped the thread here and everything would have been perfect. The description of SJSs is clear, accurate and educational. Unfortunately the follow-up was:




This unfortunately seems to have led to the thread turning nasty, both because the former comment shows a lack of understanding of hrothgar's initial explanation - when you make the SJS you know the strain and that is specifically spades after a subsequent 3♠ rebid - and because the latter was obviously uncalled for (which comment is even meant by this?). Possum, I honestly think you would do well to go back and re-read the thread from the beginning dispassionately. I think you would then see that Richard's intention was to be helpful. If you consistently react aggressively after such answers then you might well find that those constructive answers disappear and you only receive critical ones or simply get ignored. That would probably not be good for you or BBF. I will certainly be watching now to decide if I want to post in your threads constructively or not.

Dear Zelandakh

There really was no need to dredge it up since the thread went back on track. However be careful not to point the finger at the wrong people.
Its easy to jump to conclusions about where aggression starts. However I have been trying harder to avoid reacting to real and/or perceived slights.
It takes a while for compounded aggressions over many (18??) months to calm down but I'm getting there. And often those who caused the worst aggression and provocation initially are forgotten and others end up in conflict. But I'm not going to dredge through 18months of threads and table behaviour. Haven't we all got better things to do

Back to bridge and Soloways

regards P

PS People are allowed to have different views on how to bid a hand. It would be a very dull game if everyone agreed all the time and bid exactly the same way