[Flame]

I wish we could compare systems in a semi mathmatic way.
First of all , i dont care about gadgets and complex her, yes its a big plus not to have to use gadgets and get a full nad simple system, but i prefer to focus on the design of the system.
I have seen some tries to compare systems based on the bids of the systems, like 1M in pc is more limited the 1M in 2/1, 1D show more cards etc, but i think a better way to compare is to compare hand types, if we can define groups of hand types, and ranks systems of the way they handle those groups, later maybe we can do some math with the lackwood of those groups , and the score on the line with those hands (sorry for my poor english). This will give us a good feeling of those system, and maybe some idea of which system is better.
The groups should be defined to support the differents between te systems, so if Im comparing pc (wj2005) and italian style 2/1, it will might look like.
* 11-17 5+ cards major
* 18-22 5+ cards major
* 11-14 4+ cards diamond (no 5 card major)
* 15-17 unbalanced 4+D (no 5 card major)
* 18-22 4+D
* 11-14 5+C 4M
* 11-14 6+C
* 11-14 balance no 4D
* 15-22 5+C 4M
* 15-22 6+C
* 23+ hcp
This is just a scratch group definition which i wrote as it came to my mind, but this is the idea, we could extend this to be usufull to compare more systems, but i believe its better to begin with only 2 systems and not to far from each other. (for this resson i choose wj2005 over wj2000 and 4d/2c style over better major, so 1d openings are very close.
Please comment on the idea, on the groups and lets try to work this out.

[the hog]

Dan Neill has a document which compares the mean opening of various systems, this is an interesting exercise to look at.

[Flame]

The_Hog, on Aug 23 2005, 07:55 AM, said:

Dan Neill has a document which compares the mean opening of various systems, this is an interesting exercise to look at.

Do you have it ? can you send it to me pls ?

[PMetsch]

This is the document about systems

http://www.geocities.com/daniel_neill_2000...-collection.txt

[inquiry]

I think your proposed approach is intereresting, and I look forward to what might be posted h ere. But I would have probably group this slightly differently. I would suggest you define hands by slightly different terms. I would use something like, (were Game force is just that, "strong" is game force after partenr shows any life, interemdiate is highly encouraging of game after partner responds, and weak is a limited opening bid).

Game force one suiter
Game force two suiter
Game force three suiter
Game force balanced
Strong one suiter - major
Strong one suiter - minor
Strong two suiter - major
Strong two suiter - minor
Strong two suiter - major/minor
Strong three suiter
Strong balanced
Intermediate one suiter - major
Intermediate one suiter - minor
Intemediate two suiter
Intermediate three suit
intermediate balanced
weak one suiters
weak two suiters
weak three suiters
Weak balanced.

Maybe even this is too many groups, but then I am not sure what mathematical metric are going to use when yuo stick all these together.

Game force hands are handled with the game forcing opening bid in most systems (2C for 2/1, 1C for others). The strong hands (not game force) start 1C in 1C systems and start with the longer or higher suit in 2/1. The interemediate hands begin with best suit in 2/1 at the one level and with the best suit at the one level (or 2C) with the forcing club systems. As do the weak hands.

An advantage to 2/1 with 2C is that it is GF (or maybe GF on only some rebids by opener) making the really strong hands clear from the start.

The advantage to the polish club systems is they can readily distinquish between intemediate hands and strong hands. They open the intermediate hands one of something (other than clubs) and then rebid strongly, while they open the strong hands 1C and then contnue from there. So that in these systems, when they open one suit and jump in a new suit, they hold both suits and the hand falls into the intermediate range (strong enough for the jump, not strong enough to open 1C).

In 2/1, you open weak, interemediate and strong hand 1 of suit and have to try to distinquish between overall stregnth (strong or intermediate) and type of hand. This means people either can't jump shift or jump rebid with the intermediate hand, or if they can jump rebid their suit with an intermediate hand one suiter, they have to invent some other forcing bid with a strong one suiter. There is a second problem, sometimes in 2/1 you open a strong hand one of suit, and your partner with no remarkable fit and a weak hand passes, and a game is missed. After the forcing club opening, many of these strong hands opposite very bad hand games are found.

On the other hand, a weakness of 1♣ systems is that with an strong hand (not gf) and a two suiter (or even a one suiter), they open 1♣ inviting competition at a low level, and when the bidding comes back to them, they may have to start trying to seperate strong two suiters from strong one suiters from GF hand of any of the types, when so far they have provided no information about the suits they hold. In 2/1 if you open one of a suit, and then on second round are forced to bid a new suit at the three or four level, your partner will have a much better picture of your hand.

Now, since we have group these, as above, and where I think they go in the opening scheme of the two systems in question, I will say how I prefer to open each of these hand types in a pseudo 2/1 system. Note, I have pushed some "strong" hands (not gf) into 2♣ and 2♦, and I have removed an important group of GF hands (two suiters) out of 2♣ and combined them with strong 2 suiters in a compleletly different kind of opening bid.

Game force one suiter - 2♣, what else?
Game force two suiter - 2NT, 3C or 3D - Misiry transfer preempt
Game force three suiter - back to 2♣
Game force balanced - 2♣
Strong one suiter - major 2♣ - forcing only to 2♥, with 2♥/2♠ response instant 2nd negative
Strong one suiter - minor 2♦ (multi) with 9.5 tricks in minor
Strong two suiter - major 2NT misiry transfer preempt
Strong two suiter - minor 3♦ -Misiry transfer preempt
Strong two suiter - major/minor - 2NT, 3C or 3D - Misiry transfer preempt
Strong three suiter - 2♣with conventional follow up
Strong balanced - 2♦ Multi 2D
Intermediate one suiter - major - 1M
Intermediate one suiter - minor - 1m
Intemediate two suiter - longer or higher suit, jump rebid or reverse --except with minor two suiter, 1D - 1any - 3C is weak two suiter, not strong.
Intermediate three suit - 1M if five card suit, otherwise, 1m
intermediate balanced int
weak one suiters - five card major, else better minor
weak two suiters - longer suiter or higher suit at one level
weak three suiters - [b]1M if five card suit, otherwise, 1m
Weak balanced - [b]1m rebid 1NT or make minimal raise

The modification to 2♣, 2♦. 2NT, 3♣, and 3♦ allow me (playing this) to describe distributional interemdiate hands (jump rebids, jump shifts, reverses) without partner worrying I am sitting on a moose -- similar to what happens in strong club systems after a natural opening. The use of 2♥/2♠ response to 2♣ as instance double negative allows for strong one suiters in major to be opened 2♣ and you can stop in 2♥ or 2♠ if partner is broke. For me, coming from a precision/romex background (I actually played shenken club [An artificial 1 Club system devised by Mr. Howard Schenken] before precisoin was introduced to the world), the ability to show intermediate good hands with jump rebids was a very important consideration, and the need for jump shifts and reverses to show REAL suits is also key. Others don't seem to need this kind of assurances with new suit second bids).

A second reason I like my approach is that the strong hand versus magic yuck hand fit games are easily found after 2♣ or MisIry transfer opening bid. That is, you are not resting peacefully in 1M when 4M (or some other game) is cold. So the approach I take is to limit 1 of suit to 20 or 21 hcp when balanced, and significantly less when you are one suited (more than five cards) or two suited (10 cards or more in two suits) or three suited. The extra distribution allows more playing stregth, pushing such hands up into the "strong" category.

[Free]

Nice idea, but I wonder: if you have lists of 2 systems, what can you conclude from them??? It's just an opening scheme, no follow ups, nothing...

[P_Marlowe]

Hi,

look at the following thread from RGB

http://groups.google.de/group/rec.games.br...c06e26250096348

With kind regards
Uwe Gebhardt

[tysen2k]

P_Marlowe, on Aug 23 2005, 09:09 AM, said:

look at the following thread from RGB

http://groups.google.de/group/rec.games.br...c06e26250096348

Yeah, this was a little contest I ran about a year ago in an effort to actually quantify how good one system is over another. It was a lot of fun and a lot of the participants really enjoyed the experience.

Tysen

[Flame]

Ok i thought the list I think for most groups its easy to say which system is better like its obvious that with 12-17 hcp and 5 cards major or diamond polish is better then 2/1.
Now there are few groups that arent clear and i think its better to focus on them first.
I decide to make it a 3 systems question
WJ2005, 2/1 4D 2C , and a 5 cards light opening precision like system. (i added the precision to make it more intresting although different NT range make it a bit problematic here)
Again most is clear i'll focus only on the non-clear groups.

First I look only on major suit hands (5+ card major), assuming 2/1 1M to be 12-22 hcp, polish to be 12-17, and precision 10-15.

I can devide it to 4 groups and try to rank them, the first two groups are clear
* 10-11 hcp : This is easy precision is best, polish and 2/1 are behind
* 12-15 hcp : Here polish and precision are the same and better then 2/1
The second two groups arent as clear
* 16-17 hcp : I think polish is best here, 2/1 second and precision last, but not sure about it.
* 18-22 hcp : This is even less clear, i think polish is best then precision then 2/1 but again not sure.

Another group that isnt clear is the GF 23 hcp+ hand. Here i think polish is supirior to precision, but is the 2/1 2c better then the 1C opening with 23+ ?
One resson its better to open 2c is that its create a forcing pass right away, this might help if opponents interfere, on the other hand opening 1C opener might be better on slow bidding to slam with asking bid like 1c-1M-2D in polish, or 1C-1M-2M in precision. My feeling is 2/1 is best here then polish then precison.

Another question is the 5 club 4M hand 11-15(14) hcp, is preciosn and polish 2c better here then 2/1 1c ?

There are also the diamond hands but ill leave them for now because this is gettign to long.

Pleast comment.

[inquiry]

Ok... three systems, Precision, Polish, 2/1.

* 10-11 hcp : Precision is best, as long as the light opener is not punished.

12-15 hcp : I disagree that polish and precision are equal, I think slight advantage goes to precision, because the opeining is more limited. And I doubt that 2/1 is disadvantaged here to polish. I would rate them equal.

* 16-17 hcp : I agree, clear win for polish with 2/1 second and precision last. The 1♣ bid in precision does little to help these (playing a higher range 1NT opening helps a little bit for precision here).

* 18-22 hcp : I think polish and precision are equal here, but then I am not all that familar with polish. Both have advantage on 2/1.

*23+ hcp My view is 2/1 has the advantage here.

[Flame]

inquiry, on Aug 23 2005, 11:55 AM, said:

12-15 hcp : I disagree that polish and precision are equal, I think slight advantage goes to precision, because the opeining is more limited. And I doubt that 2/1 is disadvantaged here to polish. I would rate them equal.

I intentially defined the precision 1 level opening as 10-15 so it will be a 6 points range just like the polish. I cant understand how you can compare polish to 2/1 here when one is limit and the other one is not. I think this is the most clear of all.
12-17 is better then 12-22.

[inquiry]

Flame, on Aug 23 2005, 01:10 PM, said:

inquiry, on Aug 23 2005, 11:55 AM, said:

12-15 hcp : I disagree that polish and precision are equal, I think slight advantage goes to precision, because the opeining is more limited. And I doubt that 2/1 is disadvantaged here to polish. I would rate them equal.

I intentially defined the precision 1 level opening as 10-15 so it will be a 6 points range just like the polish. I cant understand how you can compare polish to 2/1 here when one is limit and the other one is not. I think this is the most clear of all.
12-17 is better then 12-22.

Becuase with 12, 13, 14, and most 15's 2/1 is going to simply open and make the simpliest rebid. While ith this same range, polsih will do the same (again, I freely admit to not being fully knowledgeable on polish club). There could be some light "splinter bids" by opener with 15 or 16 (maybe even 14) I guess that 2/1 can't afford, but those don't always turn out alright. My assumption is that polish big rebids after a non-1♣ are saved for the maximum hands (lets call them 16-17). So typically the rebids will match those in 2/1.

Precision on the other hand can jump with 14-15 without overly exciting partner. So I give the advantage to precision. On the 16-17 hands, the advantage swings to polish in my view over 2/1 for it is those hands they can show as "strong" but limited. If 2/1 show them as strong, they risk gettign partner too excited, and if they choose to stay low, they risk missing close games. Of course precison opens these 1♣ but then they have to scream minimum, minimum after that to avoid getting too high, and some 8 versus 16 games and no fit are horrible disasters for precision.

[Flame]

Ben even if you are right and 1S-1NT-2D in polish shows a 12-15 hcp, the range in 2/1 is bigger usually up to 17 hcp. (yes you can solve it with ur 2c gadget but that has its price too) also having limited opener helps in some other ways for example 1M-4M can be less defined. Beside on interference things the limited range effect is bigger (here polish is even a little better then precision). Anyway i take ur point on precision being better then polish, but you cant hide the different between the ranges of 2/1 and polish, somehow they will make a different and here i dont want to get into structures, so in ipotetical optimal structure, a 12-17 opening will be better then 12-22.
But ok what next.

[mike777]

Flame, on Aug 23 2005, 06:30 AM, said:

I wish we could compare systems in a semi mathmatic way.
*,

OUCH!

Math students around the world wince in pain.

I guess if we use semi Math with semi Logic in a semi way that could work.

What does compare mean?
What does system mean?

What does it mean to compare a 12-14 opening with a 15-17 nt opening?
Do you need to consider the follow ups? What does consider mean?

Do you need to consider the rest of the system?
If so how and why?

HInt you need to start with some theory and some methodology. Otherwise the exercise is useless.
Hint 2 LISTING stuff is not comparing.
I must admit most undergrad and grad students think it is.

[hrothgar]

for what its worth, I did a bunch of work about 5 years back designed to generate objective metrics that could be used to evaluate different bidding systems. I proposed two different metrics, both focused opening bids for different systems

Metric 1 was designed to define the accuracy of the system in providing information about "Shape". For each opening bid in the system I evaluated the frequency with which responder knew with certainty that the partnership had a 7+ card fit in a given suit. For example, if you are playing a 5 card major based system and partner opens 1♠, you are guaruteed a 7 card fit any time that you hold either 2+ Spades or a 7+ card suit. If you are playing a 4 card major based system, you are guarunteed a 7+ card fit if you have 3+ Spades or a 7+ card fit. I weighted each opening bid based on its freqeucny and summed across all oepnings to arrive at an overall "Shape" metric. [The same analysis can be repeated for 8+ card fits and 9+ card fits]

Metric 2 was designed to providen similar information regard Strength

I argued that individual biddings systems could be plotted in a two dimensional Shape/Strength space. Many bidding systems traded off in their relative accuracy (one was "better" at providing information about strength while another was better at describing shape). The interesting cases occured when one system "dominated" another and was superior on both metrics...

Given enough time, I might be able to find some of the articles (I think that I sent a copy to Luis)

[Flame]

hrothgar, on Aug 23 2005, 03:45 PM, said:

Given enough time, I might be able to find some of the articles (I think that I sent a copy to Luis)

Hope you will find it, sounds very intresting.

[tysen2k]

hrothgar, on Aug 23 2005, 03:45 PM, said:

for what its worth, I did a bunch of work about 5 years back designed to generate objective metrics that could be used to evaluate different bidding systems.  I proposed two different metrics, both focused opening bids for different systems

That sounds very interesting. I'd love to hear more if you can dig it up.

Tysen

[Gerben42]

Quote

*23+ hcp My view is 2/1 has the advantage here.

Defenitely Polish here too, then 2/1, then Precision.
The Polish 1♣ - 1♦ - 2♦ sequence shows a 2/1 2♣ opening bid, and here you have the extra information that responder has a 1♦ response. The real fun begins when responder can dig up a positive response like 1♥. Now Polish is way ahead of 2/1. Needless to say, when opponents have the distribution and vulnerability to interfere, 2/1 is ahead again.

[tysen2k]

hrothgar, on Aug 23 2005, 03:45 PM, said:

for what its worth, I did a bunch of work about 5 years back designed to generate objective metrics that could be used to evaluate different bidding systems.  I proposed two different metrics, both focused opening bids for different systems

I've been thinking about this in more detail and I think there are a couple problems with it.

• It ignores preemption and any ability to make it tough for the opponents. It only looks at the constructive elements.
  
• It ignores the level of the opening. If I define 7NT = 10-15, 6+ spades, it would score well.
  
• It ignores the fact that finding a fit in a major is more useful than a minor fit.
  
• It ignores negative inferences. I like using a 11-15 NT that denies a 4cM. The fact that responder knows for certain that there is no fit is useful information.
  
• Some "either or" bids would be unfairly penalized. An artificial 1♦ opening that promises a 4cM is an example. Responder couldn't guarantee a fit unless he has support for both. However, further bidding will reveal a lot more information very efficiently. A multi 2♦ would have the same problem to a lesser extent.

Maybe it could work with a couple tweaks. I'd still like to see what you came up with if you can ever find it.

Tysen

[han]

Thanks for posting that link to the thread by Tysen on rbg. That was a pleasure to read.