xxhong

Those experts really ignored the success of the prediction of HaiCheng earthquake. IMHO, not all types of earthquakes are predictable, but some certain types of earthquakes are really predictable. http://baike.baidu.com/view/33629.htm

ME! The TV series is pretty good. The pace is a little too fast to those who never read the novel though. Also, it's a pity that "yet not all bastards need be dwarfs" was cut in the TV series.

It doesn't give any signals (no matter counting or attitude). Often it plays the highest card that doesn't cost a trick in its limited sample size(which often costs you a trick in a single dummy analysis). Other times, it simply gives the lowest card no matter it encourages or not. It pays no attention to your signal. It does have some opening lead conventions (K from AK in suits, A from AK in NT, 4th best, top from sequences), although it frequently overrides such leading conventions if the simulation says otherwise.

This is a huge basic problem. Total points and high card points are different animals. With 33 total points, 6NT is not guaranteed at all. 6NT usually has some play with 33 HCP, but not total points. In the program, gib simply ignores the difference. Suppose you open 7S with 13 spades, your gib would correct to 7NT with any cards if this logic applies because you showed 37+ total points.

This is a simple problem to solve. To those who like fast games, bbo can just open tournaments with fast setups. To those who like decent games, bbo can also open some tournaments with slow setups. The basic problems of gib are that often it makes wrong constraints or spends too much time in situations that doesn't make a difference. Also, for defensive plays, one need to construct some kind of specific set-ups to beat the contract, which is often a rare event. In that sense, a random sampling of very limited hands is an ineffective way to attack this kind of problems. For human players, we just try to construct possible hands from partner to defeat the contract then decide which patterns are more likely.

A better treatment is not to play 4th suit gameforcing. Instead, you can do the following: 1D 1H 2C: 2D/H: sign offs. 2S: different invitational hands. 2N: minimum, now responder can show his type of invitation. With extra, opener can bid his hand naturally. 2N: 5+ Hearts, gf. 3C/D: gf, natural. 3H: set up H as trumps, natural. 3S: splinter. The biggest problem for standard treatment is that it only give one bid for gf hands and several bids for invitational hands. It's very ineffective and ridiculous to make 2NT/3C/D/H all to show invitational hands and only 2S as gf.

This comparison may not be fair. Gib can certainly be more effective than human players if both sides partner with gib. However, if you put one gib pair in a normal tournament, their performance may not really be above average, especially if you put gib in a tournament where most players know gib well. Basically, gib declares better than average player ( still, no way close to expert players because of some of the basic design problems). Gib's defense is just not comparable with average players because there aren't any defensive signals. For defense, often, it's not a matter of simulations. In latter rounds, it's often a sure thing after accurate signals. Gib's bidding is above average in some areas (after NT openings) and rather bad in other areas (slam bidding, competitive bidding, sequence after 2/1 and 2C and inverted minors). So I think generally, gib's bridge is comparable with average tournament players and can be weaker. Finally, let me give you a deal to show how bad gib's defense is and how expensive it can be. (Against human players, I wouldn't redouble. Also, I know my human opp is aggressive from previous hands) Here I sit south, the other human player sit west. N and E are gibs. [hv=pc=n&s=sa73hkq64da75cjt5&w=sj952h952d4caq632&n=sk84h8dkqj832c874&e=sqt6hajt73dt96ck9&d=w&v=b&b=4&a=ppp1cp1d1h1n2h3dp3ndpprppp&p=h2h8hah4ckc5c6c4hjhkh5c7dad4d2d6d7c2dkd9dqdtd5c3djh3s3s2d8c9cth9d3s6h6cqs4stsas5hqs9c8h7s7sjsksqs8htcj]399|300[/hv]

The situation got worse on last night as it appears. RPV of reactor 2 might be damaged and there were explosions in both reactor 2 and 4. Also, reactor 5, 6, 7 need to be cool down. Reactor 4, 5, 6 were not operating at the earthquake, but they were said to still have some heat left. The most serious problem is probably the damage of RPV of reactor 2 I guess.

... 2C 3H(set up trumps in H, 7+, solid) 3S(cue) 3N(odd number of aces) 4D(cue, denies C control) 4H(just one ace, so can't bid more than 4H) 4S(SK and we don't miss two aces) 5C(Cue in C) 5H(enough) 6H(CJ can be very important and we have no losers in S + D)

It's usually right to open it and treat it as a minimum but not a dead minimum. It has some potentials in slam zones and may not play well in 3NT.

West is certainly very light for the take out and east is also rather light for the 3NT. I think this is the price you have to pay if you want to make light takeout doubles against preempts, you gotta end up in some unpleasant 3NTs. Also, that's the price you have to pay to play leb because you gotta end up in some bad 3NTs facing a distributional minimum. Nobody really likes the convention leb. However, it is often worse not to play it. I don't think there will be a solution in the future to handle the sequence against preemptive bids very effectively. That's just a life of bridge players. Of course, if you want to avoid more bad 3NTs, you can tight up your take out double a little bit. For the hand listed, the control is rather bad. So it is probably avoidable if you pass 2D. The general idea is still that you have to pay for opp's preemptive bids. So really, one should preempt a lot more than most players do at 2 level IMO, especially when white.

Interesting to see so many went wrong in this normal layout. After partner's original pass, one really shouldn't expect any respectable major suits from partner.

This is probably why we should play transfers. 1S (2H) Double: simple S raise. 2S: transfer to NT. 2N: transfer to C. 3C: transfer to D. 3D: S limit raise. 3H: gf raise to S. 3S: constructive raise with 4 or more S.

The general bidding design of gib puts too much attention into stopper and gib just hates NT. A better logic for 2NT rebid should be "let's bid 2NT when we don't have support in opener's suit, don't have side suit below 2NT and don't have shortness in unbid suits."

I think it is a very common situation. For 5-4 majors, no much extra value, 2S is generally better than 2H. For 5-5, no much extra, 2H is better. For hands with extra values and game is possible, 2S is clearly better than 2H because it keeps the bidding alive. Therefore, 2S is better than pass, especially when opener holds some extra (here pass can be very bad when you may belong to 3NT, 4H or 4S)

Agree. In those old 16-18 1NT days, players rebid 2NT to show 14-15 and bids something else with 12-13. Now more and more play 14-16 1NT. One big headache is how to distinguish good 16 to bad 20. Perhaps if one likes a natural 2 point range system, he perhaps should play strong clubs. like 12-13, 1x then 1NT. 14-15, 1NT: 16-17, 1x then 2NT. 1C: 18+. Of course, those who play 3 point ranges never have problems, they just guess well. 12-14, 1x then 1NT, 15-17 1N. 18-20, 1x then 2NT.

gotta have more ways to raise in C. Now only 1NT makes sense to me. I used to play 1C: 3C(gf) 2D(invitational) 2H(gf raise, void somewhere, gf) 2S (constructive raise with about 7 to bad 10) 2N (balanced invitation) 3C (weak, preempt) Of course, if you want to incorporate fit showing jumps, you can eliminate some levels. Perhaps something like: 2D(fit showing jump in either D/H/S, later 2H: relay, 2S/NT/3C: fit jump in D/H/S. 2H(inv raise) 2S(constructive raise) 2N( balanced inv) 3C (weak, preempt)

This is a horrible convention which is often abused by gib. Usually, it should be replaced by a cuebid or RKC. Basically, the program is very weak at bidding RKC or cue bid. That's generally the major weakness of not having a tricking counting function during the bidding. Also, in many situations, when slams are unlikely and 5m is the right contract, you just can't jump to 5m to play there and have to bid an nonforcing 4m.

Forcing pass should be applied here to distinguish hands with slam interest or just a mere raise I suppose. So hands with slam interest can pass and pull partner's possible double to 4H. Minimum raising hands can just try 4H right away.

1C 1D x p 2D p 3C p 3D p 3S(cuebid) 3N p 4C It's hard to avoid a slam here.

I pass for both hands. When it is unlikely to be our hand, I just don't want to give them distributional information. Of course, when you hit a huge H fit, you may be able to outbid them and make a lot of hearts, which is a rather rare case. Both hands are not close to a opening hand (about one King weaker), so I really don't have to bid them.

It won't be an easy problem to do simulations. First you really need to know your opps' style. Do they preempt too much, which means that they overbid with many many 7 card broken H suit( or even some 6 card suit)? If that's the case, the chance would be that they most likely can't make 4H and your chance to make 4S is also rather slim. So against such opps, pass would be good. If they are really serious with their 4H openings and often preempt on the heavy side of the whole spectrum, you should probably bid 4S, because they may have a tough time to double you when they can probably make 4H. However, you also need to know about your RHO, does he make very light penalty doubles purely based on trump stacks? If that's the case, you should just pass and wait for the nuts. If he is a very conservative doubler ( could be due to his partner's aggressiveness), you should tend to bid 4S again. Without any information, it would be just a guess and any of the three bids can work very well or very badly, well, perhaps not the pass.

Interesting improvements. Still, isn't it just too light to double and cuebid with 18 points(HCP?)? Hands like AKx xx Axxx KQJx would double 1H and cuebid 2H? 17HCP and one point for shape. I think expert standard would be a few HCP stronger here. Somewhere around 20?

4D is too aggressive but 6D is nonsense. Partner's 4D never says that he can make 4M if you can provide only one trick. Now you can provide three tricks, so that really doesn't mean that your side can make 6 M. In this sense, north's 6D is probably more aggressive than south's 4D. Hands like SATxxx HKTxxxx D- CQx looks like a very normal 4D bid, where 4M can really be your limit here.

You misbid in both rounds. With SAKxx, 12 HCP and 4 diamonds, you really should bid something against 2S. If you worry that partner can be quite light, you should at least bid 2NT to invite and I don't mind a 3NT bid at all, because that is what you are going to bid if your opp bids 3S. On the second round, when your partner showed you a decent hand with at least 6 diamonds, you really should bid your game or potential slam if partner's hand is slightly stronger. So it is completely your fault.