iandayre

I prefer the overcall to an immediate double with this hand. I would plan on doubling if a 2S raise is passed back to me. North's pass of 2C is just too bizarre for words.

Not at all. Post anything you like. My main point is, GIB doesn't do well responding to takeout doubles in general. Making them with marginal hands is losing strategy. In my opinion, this forum exists not only to point out GIB's weaknesses, but to identify successful - and losing - strategies for playing with and against them.

I thought my comments were pretty clear. I acknowledged your point that GIB's bidding was foolish. No, yours wasn't worse, but doubling vulnerable on that hand in direct seat would not have occurred to me.

GIB doesn't play Wolff Signoff over a 2NT rebid, so you are also somewhat stuck with a very weak hand with a long major (say KJxxxx and out) in that auction. I love Soloway strong jump shifts, but you also need a way to show this sort of hand. I honestly have never known anyone who played this 1NT rebid as 18-19, nor have I seen it in print, so it's not surprising that continuations from there are unclear.

Your point is of course valid. GIB should never be bidding 3S, much less 4S. But I have noticed lately that a large majority of the posts reflect very poor bidding by the human player, forcing GIB into decisions it never should have had to make. I wonder if GIB would have balanced with a double if given the opportunity.

Agree with 1NT. At matchpoints Pass is a reasonable shot. At IMPS I bid 4C

If you agree with a minimum advance, then I see no problem with a strength-showing action on the next round. The problem I do have with N's bidding is the failure to make the prepared 2S initial advance, leaving Hearts in reserve.

Your double would have been my choice as well, and your point is right on. GIB has a full opening bid and promised exactly ZERO with the actions taken. Another of the myriad examples showing that GIB's biggest weakness is in acting over partner's takeout double. Although in this case GIB has plenty for a free bid of 3D.

This was the method advocated in Max Hardy's books on 2/1. I always found it to be worth playing.

The problem here is the description of 1NT, which you obviously didn't read. I don't agree with it, but with GIB your 1NT rebid over their double shows 18-19. Given that, the auction is similar to this non-competitive one. 1m-1S-2NT (same range you showed on this hand) 3S, which is unlimited and forcing. It's still perfectly OK to PASS to show minimum, balanced values in context with previous bidding. I would pass in any case, my 1NT rebld would be a good 13 or 14 HCP.

True but why not just raise to 3S?

I am sure that Jack is correct. And I have heard that it is difficult, if not impossible, to change GIB's card play parameters. But it seems to me that GIB can never be taken seriously as a bridge-playing program if it makes insane errors like this. Is it really impossible to program GIB to take its certain tricks when those amount to ALL the remaining tricks? Why should it ever take the slightest unnecessary chance? I imagine it is more than just GIB choosing between "equal" plays. Did all GIB declarers in this contract play the same way? This implies that the programming makes it somehow more attractive to take the unnecessary finesse, as if to prove GIB can really count points accurately.

It's true that the hand is a bit light for the 2S game forcing call, but given the C weakness it's my choice. And yes it's a jump shift, not a jump reverse. Then after 3H - 4H, responder can BW and once all Key Cards are confirmed, with the great strength W has shown and the fitting cards in opener's first suit, 7NT stands out.

I completely fail to understand the allure of such gross overbidding. How about Axx, KQxxx, AT, KQx? GIB will not play you to have 4 Spades since you didn't reverse, though showing the strength to do so.

Agree. You make an aggressive bid, get lucky and find a fit, and GIB hangs you. This despite the fact that it heard you open 1NT and thus belong in a part score.

I would start by balancing with a double. Too good support for C to show a red 2 suiter. E should bid 3C, playing Lebensohl this shows some values but is non forcing. 3H, 4C and probably 4S next. Can E bid the slam? I don't really see it. I'd sign off in game as E and leave it there as W.

Balancing with 11+ balanced points against a strong NT is utter nonsense. The NT bidder's partners knows his partner's hand and will often know exactly what to do. You will do much better passing and defending.

Truly GIB at its worst. It does seem that we don't see quite as many of this sort of nonsense auctions as we used to, but they do still happen.

GIB follows standard practice in this area. Pass of the double shows a doubleton in partner's suit, and says nothing about the holding in the transfer suit. With a strong holding that you believe is likely to take 8 tricks with that trump suit, you should redouble.

True. I understand less why E then passes 3C. Or for that matter, why your double of 2NT is essentially undefined, giving no additional information to partner beyond the opening bid.

I have played Soloway strong jump shifts for many years, and they are a great slam tool - both reaching slam and staying out of poor ones. The followups are quite clearly defined. I am sure you could find a write-up, and I'd strongly recommend giving them a try.

Well, so much for that theory. I also thought your point was valid. http://tinyurl.com/q9yr37x Note E did not split the C honors.

I'll comment on a couple of points here. I will admit that the final 3NT call by the advanced robot is simply infuriating.(I should be used to it by now). NO sane human would make this call and the fact that the robot does so is very disturbing. An double of 3S with the maximum 1NT response would be quite sensible, and pass would be at least within reason. It appears I owe the basic robots something of an apology! Jack's next-to-last paragraph brings up a point I have been seeing more often. One partner's point range has been established, then the other makes a call that ostensibly shows a range which, at the higher end, would certainly belong in game, while at the lower end clearly does not. Another simple example - I doubled opener's 1D in direct seat with 1444 shape and 20 HCP including AT9x of D. After partner's expected 1S I bid 1NT, and partner took out to 2C. The description was 1-9 points. How can one possibly be expected to know what to do opposite that range? Fortunately RHO saved me by doubling 2C and after P by me and 2H by LHO, my GIB partner came to life with 3NT, making easily. Then there was the recent time I doubled 3 of a major, and partner took out to 4D. The description? 20 or less points. LOL. Gee GIB, you don't think you might make a stronger bid with 19? The hilarious result? With 2 small of opener's suit, 44 in the others, AKx of D and 16 HCP, I passed for down 2. Those who went to 5D were treated to a generous defense and were down only 1!! But then we already know that the bid descriptions are, in many instances, totally inadequate, and that BBO has no intention to expend the necessary resources for the complete overhaul that is needed. A shame. It's true that most of them are OK for simple constructive auctions, but once there is competition we are often flying blind.

It certainly seems as though the hand qualifies even under the GIB definition to make the sane bid, 2H. I wonder why it doesn't. It has only 10 HCP, but both 2H and 2NT promise 11-14. You lose a lot of the gain in opening light distributional hands when you don't show the distribution in subsequent rounds of bidding.

It's true that nobody posts hands where the bidding has totally gone off the rails and GIB is able to get additional defensive trick(s) by not cashing out. But still I, like Steve, question the quality of the simulations. Is GIB simply simulating what lie of the cards is most likely, or is it (as it must) also considering the cost in MP's or IMP's of being wrong (failing to cash when it was needed) as opposed to the gain in extra undertrick(s) for being right?