[Bermy]

No sooner have I noticed the welcome announcement of Gib 37 then this happens.

Daylong imps $0.25

http://tinyurl.com/y8c6w6uh

What kind of 2/1 bidding is this? Surely after I was showing 2 black suits I am entitled to believe that North will have both red suits covered? That there should be no possibility of a 5-3 fit in hearts?

I can only describe this as total abuse of the 2/1 system and its forcing NT.

I have long been an advocate of an alternative bidding systems and have made no secret of my dislike for 2/1 and Gib version of 2/1. Roll on 38, 39 40 and so on, but you will never get this rotten system right.

[steve2005]

I have no idea if Gib bidding is advanced enough to use this 2/1 agreement but I have seen it in print and in action
1S-1N-2C-3H shows 6H INV
so with 5H INV you go
1S-1N-2C-2N and with 3H opener goes 3H on way to 3N to find potential 5-3H fit.
Notice 3N bid is listed as having les than 3H.

[Bermy]

 steve2005, on 2017-August-22, 17:25, said:

I have no idea if Gib bidding is advanced enough to use this 2/1 agreement but I have seen it in print and in action
1S-1N-2C-3H shows 6H INV
so with 5H INV you go
1S-1N-2C-2N and with 3H opener goes 3H on way to 3N to find potential 5-3H fit.
Notice 3N bid is listed as having les than 3H.

Yes I figured as much after, however why do we have to endure such ridiculous bidding?

We now have a system where we have to trust North (as you say) to handle these complicated and unnecessary bids. Surely the object of the players is to find the right fit and contract? Now declarer has to bid its 3 card suit to find some 5-3 fit? While trying to avoid a 4-3 (maybe a 3-3) contract in hearts or a bad 4 clubs contract is not out of the question. Declarer has to make the decision without spoiling the bidding, game is by no means certain after the 1NT relay. I can only advocate that the heart suit is shown as soon as possible, you cannot have 2 bids without showing the 5 carder and you cannot introduce NT without some diamond protection.

"Notice 3N bid is listed as having les than 3H." the 2NT bid is not forcing its invitational, how can this be relevant?

[Bermy]

The system places too much pressure on the blind bidder. After the 1NT relay opener is really quite blind to Norths hand and has to rely on North to make the contractual decisions. Something we all know Gib can't do well.

Its almost as if we must now play poker to get a result, is that robot bridge?

We have a few options,

1) Pass the forcing NT at the first round of bidding. (I love doing that, do you know how many 100% scores it gets me)

2) Pass the 2NT bid and get the diamond run against us to go down. (Gib will probably mislead anyway)

3) Bid 3 Hearts and risk a 4-3 or 3-3 contract. Gib will most likely bid 4 clubs to this, then what do I do?

4) Bid 3NT........well if you going to play poker bridge, no point giving away any more information.

Is this the system we really want for robots?

[Stephen Tu]

Bidding 3h with the south hand is absolutely routine if accepting. The main danger is reaching 3nt going down, it really shouldn't ever get to 3-3 fit or 4-3 fit (except when you WANT to be in 4-3 fit, with strong hearts nothing in diamonds). As long as robot is programmed with the right continuations over 3h, bidding 3nt with diamonds stopped and not great hearts, it should be fine. The system is ok as it is going to get, it is just a matter of correct implementation.

There really is no perfect bid with the north hand. Passing 2C or bidding 2s can miss game as that can be done with much weaker hands. Raising to 3c sucks on only 4 when so often south has 3 only. 2H is weaker with longer hearts normally. 2H over 1s works best if not gf, but the vast majority of serious bridge players consciously decided to play this as gf a long time ago. It makes this set of hands harder to bid, but makes others easier.

I do think the robots should use 2/1 as it is by far the predominate system in the live tournament scene. I think you are wasting your breath advocating for non 2/1 or the implementation of other systems. They should focus on cleaning up the one they have first before trying to do multiple systems. Also since you sound like you have never written a computer program in your life, you should probably refrain from making judgments on how easy it ought to be for the programmers to implement your wishes. You sound like a non rocket scientist complaining that we don't have normal $500 shuttle service to the moon yet and saying it ought to be easy.

[steve2005]

Well Bermy is right about 1 thing, this would be easy auction not playing 2/1.
Also, have seen expert partnerships ,miss finding 5-3 ♥ fit on similar hands.
But that is 2/1 system.

[steve2005]

 Bermy, on 2017-August-22, 17:45, said:

"Notice 3N bid is listed as having les than 3H." the 2NT bid is not forcing its invitational, how can this be relevant?

If Gib had 3H it would have bid 3H forcing rather than 3N. Then you get to choose game. So 3N is 0-2H. Of course Gib is know to fib about lengths.

2N is listed as unbalanced. Unfortunately 2N is often unbalanced. In this sequence is likely to be balanced. The only other possible bid would be 3♣ INV. As opener may have 3♣ this is unappealing, also have 5♥ so stuck with 2N.

[Bermy]

 Stephen Tu, on 2017-August-22, 18:44, said:

Bidding 3h with the south hand is absolutely routine if accepting. The main danger is reaching 3nt going down, it really shouldn't ever get to 3-3 fit or 4-3 fit (except when you WANT to be in 4-3 fit, with strong hearts nothing in diamonds). As long as robot is programmed with the right continuations over 3h, bidding 3nt with diamonds stopped and not great hearts, it should be fine. The system is ok as it is going to get, it is just a matter of correct implementation.

There really is no perfect bid with the north hand. Passing 2C or bidding 2s can miss game as that can be done with much weaker hands. Raising to 3c sucks on only 4 when so often south has 3 only. 2H is weaker with longer hearts normally. 2H over 1s works best if not gf, but the vast majority of serious bridge players consciously decided to play this as gf a long time ago. It makes this set of hands harder to bid, but makes others easier.

I do think the robots should use 2/1 as it is by far the predominate system in the live tournament scene. I think you are wasting your breath advocating for non 2/1 or the implementation of other systems. They should focus on cleaning up the one they have first before trying to do multiple systems. Also since you sound like you have never written a computer program in your life, you should probably refrain from making judgments on how easy it ought to be for the programmers to implement your wishes. You sound like a non rocket scientist complaining that we don't have normal $500 shuttle service to the moon yet and saying it ought to be easy.

This sounds like more excuses. I don't see any point in insulting the end user for not wanting to pay for rubbish. You sound like if you were assigned giving this shuttle service it would never leave the ground or drawing room.

Robots that bid properly with multiple bidding systems DO exist and have done so for over 25 years now. 25 years ago I had a robot made by Saitek that had multiple bidding systems. 2/1 is mainly an American system and is not nearly as popular or as good as you may think, most of us only play 2/1 because we have to. The world robot bridge championships starts soon, go see what they are capable of. Take note Gib has not even bothered an entry.

[Bermy]

OK so the results of this hand are out now and so I can comment with full hindsight.

1 player played 3NT+3 by opening 1 diamond and bidding 3NT when robot respnoded with 1 heart. Robot misdefended of course as it would not touch the psyched suit.

Only 1 player bid 3 hearts from 32 players, making 4 hearts. 1 from 32 Understands this rubbish? For such a popular system only 1 from 32 could find the right contract? As I explained earlier 3 hearts from south at this point is a ridiculous bid, and is only going to work under special circumstances and double dummy. (he probably read this article first)

As I predicted, the best prize went to the only player to pass the forcing 1NT earning 3 imps http://tinyurl.com/yc65aw3b Is this really how we should play 2/1?

6 players made 2NT or more by incorrectly passing the 2NT bid and having robot misdefend. (Also predicted)

18 players went down in 2NT or 3NT contracts, so Im not alone with this complaint, it seems like other want to bid properly too. I dont think any other bidding system would play this contract outside the 5-3 heart fit, finding game should not be difficult either.

What about having Gib sitting south at one table at every tournament, so we can see how Gib would bid it.

Carry on making excuses, but I seem to have the majority of support, judging by the bidding of 32 players.

[Stephen Tu]

 Bermy, on 2017-August-22, 23:14, said:

This sounds like more excuses. I don't see any point in insulting the end user for not wanting to pay for rubbish. You sound like if you were assigned giving this shuttle service it would never leave the ground or drawing room.

Robots that bid properly with multiple bidding systems DO exist and have done so for over 25 years now. 25 years ago I had a robot made by Saitek that had multiple bidding systems. 2/1 is mainly an American system and is not nearly as popular or as good as you may think, most of us only play 2/1 because we have to. The world robot bridge championships starts soon, go see what they are capable of. Take note Gib has not even bothered an entry.

If you don't think you are getting your money's worth then vote with your wallet, don't play on BBO. There are other bridge sites. []

Sure lots of software have options to play other systems, and an array of gadgets. Even GIB has this capability when released 20 some years ago. But I doubt that many software play multiple systems and gadgets WELL. They may get the initial call right, but handling all followup bids in competitive sequences, defending against all these gadgets/systems? They are like beginner/int humans who try to learn too many conventions, instead of learning a few conventions thoroughly backwards and forwards. I'd rather the bots concentrate on bidding/defending one system very well. They have enough trouble updating the one system fast enough, now you demand them make their job 20x harder by adding additional system support? You have NO IDEA how dumb computers are and how tedious and time consuming it is to program them. At least take an intro CS course before you make specious comments about how easy it ought to be.

As for your survey about this hand ... well most players don't know how to bid, the average player on BBO kind of sucks. Supposedly GIB advanced bots average 56-57% or so vs. humans in tourneys when tested. Do we want GIB to mimic bad players, or program it to mimic good players? I for one want it to bid properly.

Switching to non-2/1 methods might make it easier for this one hand type, which rarely occurs (which is why so many of the BBO randos never learned the sequence and didn't bid 3h along the way to 3nt, though to me passing 2nt is not at all unreasonable with an aceless 13 count), easier to handle, but then it opens up a huge can of worms because there are just so many variations in non-2/1 styles. Is 1s-2h-2nt forcing or nf? Show extra or not? Is 1s-2h-3h NF. If NF how does opener show a strong heart raise, too strong for 4h, no splinter available? Is 1s-2h-2s-3s forcing or not, again if not how do you handle if too strong for 4s? Does 2/1 always promise a rebid or only on certain sequences? There are just no real standards for these sort of questions, which is why people like Larry Cohen are advocating just teaching beginners 2/1 to begin with. To me it's natural to do so, long ago 4 cd majors were standard, at some point everyone started teaching beginners 5 cd majors even though opening 3 cd suits hardly feels natural. At some point we just can teach people 2/1 from the beginning, and hopefully the teacher/book mentions this sequence. Duplicate bridge players, BBO's audience, in the live game heavily favor 2/1 for all but the very beginning level players. 2/1 would not have taken over if people didn't generally prefer it.

2/1 has spread quite a lot in duplicate circles even outside of America, if you look at the world championship pairs playing natural 5cd major systems, very few are playing non-gf 2/1s. In Acol land Britain it's not very popular, but those are mostly 4 cd majorites and many of the top British pairs did switch to 5cM and 2/1.

This post has been edited by diana_eva: 2017-August-23, 15:00

[Bermy]

Roll on Version 38, 39, 40 and so on and so on. Keep making excuses, let people hear. And thanks for the advise on other bridge sites, do they give ACBL masterpoints too?

I was right in the first place, that Gib is the agenda of its programmers as to how they think 2/1 should be played. My survey proves that most disagree. The person who complains that everybody is riding the "wrong" way on the motorway?

Gib cannot be relied on as a partner,a teaching tool or a gauge of true bridge ability, when psyches and poker bids and are the only way to get decent results.

[The_Badger]

Whilst sympathising with what happened on this board, Bermy, another recent post by you on the Discussion Forums was: "I'm so tired of humans. Why I prefer robot bridge." Mmmm....

Wonder if you feel the same way about robots after this bidding nonsense? Though I have to say, in my opinion, the real reason for ending in a hopeless contract on this hand is my bugbear of 2/1: a forcing 1NT response covering a multitude of hands, and a 2♣ rebid by opener that could potentially be a 3 card suit.

[Bermy]

"Supposedly GIB advanced bots average 56-57% or so vs. humans"

Come on accept my challenge. Play one Gib South at every robot tourney. This will have the effect of adding only 1 extra table, and the South Gib result can be published alongside us humans. I really do not believe that Gib would perform anywhere near 55%, perhaps once on a good day, but really on an average no way. Let us see how Gib would "brilliantly" bid and play these hands.

[johnu]

 Bermy, on 2017-September-01, 06:38, said:

I really do not believe that Gib would perform anywhere near 55%, perhaps once on a good day, but really on an average no way. Let us see how Gib would "brilliantly" bid and play these hands.

You overestimate the quality of bridge of the average and below average human players. How many times have you been in a simple and routine 4 of a major or 3NT contract and with nothing interesting in the play, ended up with 60-75% of the matchpoints. Looking at this from the other players point of view, they are getting 25-40% matchpoint scores on a lot of routine hands. That's one of the reasons they are scoring below 55%. You don't have to play brilliantly to end up with 55%, you just have to make fewer mistakes on average than 54% of the competition.

[Bermy]

 johnu, on 2017-September-01, 12:37, said:

You overestimate the quality of bridge of the average and below average human players. How many times have you been in a simple and routine 4 of a major or 3NT contract and with nothing interesting in the play, ended up with 60-75% of the matchpoints. Looking at this from the other players point of view, they are getting 25-40% matchpoint scores on a lot of routine hands. That's one of the reasons they are scoring below 55%. You don't have to play brilliantly to end up with 55%, you just have to make fewer mistakes on average than 54% of the competition.

So what does that say about Gib if 55% is not much better than most novices? Remember North Gib contributes to these results too, not to mention the rubbish I often see from E/W

[johnu]

 Bermy, on 2017-September-01, 14:15, said:

So what does that say about Gib if 55% is not much better than most novices? Remember North Gib contributes to these results too, not to mention the rubbish I often see from E/W

You've seen the awful bids and plays that GIB routinely makes that have no apparent relationship to good bridge and yet according to BBO, that's good enough for 55%+ on average. What does it say about novices and intermediates that GIB scores better than them? To me, 55% is the new 50%, about average strength.

[jdonn]

GIB made a reasonable choice with no good option. You clearly misbid (by choosing 3NT instead of 3♥), and thus missed the best contract that was easily reachable. Perhaps GIB has more to complain about on this hand than you do.

And GIB absolutely averaged over 55% when run later in past human tournaments. If you don't believe it then what can I say, ignorance is bliss.

You might get more satisfying responses if you were a bit more polite, and if quit questioning the 'agenda' of the programmers.

[Bermy]

 jdonn, on 2017-September-01, 16:44, said:

GIB made a reasonable choice with no good option. You clearly misbid (by choosing 3NT instead of 3♥), and thus missed the best contract that was easily reachable. Perhaps GIB has more to complain about on this hand than you do.

And GIB absolutely averaged over 55% when run later in past human tournaments. If you don't believe it then what can I say, ignorance is bliss.

You might get more satisfying responses if you were a bit more polite, and if quit questioning the 'agenda' of the programmers.

OMG you really don't get it, do you. Its you that is making a fool of yourselves. All can see, and the results show for themselves. Yes, carry on demanding we bid 3H "your" way and play 2/1 "your" way and wonder why we accuse you of an agenda. Next I will be writing about how Gib left me in a 3-3 contract because I made the other bid, or didn't solve the mysteries of your ridiculous explanations. I am not the only one here and this is not the only hand worth complaining about. Are you not watching how everybody is playing or reading what everyone is saying?

And for the record its 2NT that was the WRONG bid.