junyi_zhu

A typical failure of 4th suit forcing. After so many rounds of bidding, the opener's range is from 10 to 18 HCP.

it appears that gib's rules are not very consistent and there are many holes in the rule, which may lead to the pass in many situations I guess. Still, it's a large amount of work to make the rules consistent IMO.

1C 1D 1N 3D(set up trumps, gf) 3N 4C(cue) 4S(cue) 4N(cue in H) 5C(cue) 6D all pass.

this is a very normal preempt, those who always pass with this kind of hands just missed the best part of preemptive bids. 2H is also acceptable.

It's really a matter of your overall overcall style. If you play a very sound overcall style with about the opening strength, you actually only want to play the weak and very strong version. If you play a very light overcall style, your partner would pass with many 8-9 unfitted hands, you want to play a continuous range or wide range, because you may easily end up in the wrong suit if you don't show your hand. Your up the line treatment doesn't make sense to me, because you would easily miss a lot of 5-3 fit spades. My preference is a sound overcall style with very aggressive preempts, so I actually play a strong and weak version. Still, if you play a rather light overcall style, you may need to guess well in many situations in constructive bidding. Often, you may end up playing in the right suit, but wrong level, which is still a big concern.

It really depends. First, you don't have to play a forcing NT in 2/1. If you play a forcing 1NT response, you can bid 1NT then jump to 3M. Second, if you want to keep low, you can play 2C as a two way bid to show either true clubs gf or 3 card limit raise. You can devise some step response. Third, if you want to play a straightforward 2/1 without forcing 1NT, you can design jumpshifts to show the difference, for example, you can use 1S 3C as a limit raise then 3D to ask how many trumps. My system uses the third approach, mostly because it's rather straightforward.

This is a 2S bid. In my system, 2S shows minimum, 2NT shows 6 spade and 3 levels show extra.

I actually agree that the basic design of gib bidding is very flawed. Bidding should be based on rules and occasionally on simulations for high level competitive auctions. Currently, gib's bidding is very random and often breaks many rules as long as gib forms some unlucky situations in its very limited simulations. That also means that gib would give different bids if the simulation gives different results. This is very difficult to cope with IMO. When I open 2NT, I often see gib raise it to 3NT with very little value, like one Jack one Queen or 4 jacks. Also, when I show a hand in the range of 20-21 points in competitive auctions, I often see gib overbids to games with jacks. These are really very very bad IMO and many players I know just don't play with gib for such kind of reasons including my regular partner. Also, it is really not too difficult to implement a feature to allow the human to declare when in dummy. It's just frustrating and pointless to watch gib goes down in many cold slams and games because some intrinsic problems. Gib's declarer play is about an average club player, and there is really no point to see gib butcher so many contracts playing one's own money. Also, this implementation would make playing with gib way more enjoyable, cause you only need to tolerate the bad bids and bad defense by gib, not the bad declaring plays. This implementation should draw a lot of players into MB IMO. My last point is that a counting signal should really be implemented into gib's defense. It's really trivial to do such kind of programming. An attitude signal is not that easy to implement, but count is really not that difficult. With the simple counting defensive signals, many defensive situations can be solved and defense can be much more enjoyable.

That's interesting. I actually don't know whether Hun has a strong relationship with "Xiongnu" in China. They certainly sound very close. Hun was so powerful in the early Han dynasty and the Han emperors was forced to send the princess to them for peace. Later, they were defeated by Han and divided into two parts. The southern division merged into Chinese in a few hundred years. Actually Jin Dynasty was destroyed by them. Still, they actually appreciated Chinese culture and later completely merged into Chinese. Now I think some north Chinese with the family name Liu can easily have Xiongnu's blood, because many of them adopted Liu as the Han name because they also think they are descendants of Han emperor (because of the marriage of Xiongnu's King and Han princess). The northern division ran to central Asia and later, nobody really knows where they went. A similar story happened again in Tang dynasty. Tang defeated Turks and Turks divided into two parts. The western part also ran to central Asia and nobody really know where they went later. Nowadays, some think Turkey can have a strong relation with the Ancient Turk.

What's the relationship between Hungarian and Hun?

Not that difficult. First, there are about 5K to 6K daily used characters. Second, at elementary schools, kids just learn about 10 to 20 characters in each lessons. So kids can pretty much read newspaper after grade 2 or 3. Of course, there are a lot of quizes by the teacher to test whether the kids still remember the new characters. To me, Chinese characters are simpler because I can memorize the shape better, not the order of English characters to form a word. Chinese speaking is relatively simple. Many foreigner can speak good Chinese after staying there for a few years.

Playing IMPs, one should frequently bid 1NT with 5422, because what it misses is some partials and when your side can force to game, it's always not very difficult to find the potential minor suit fit by two way checkback. This would also make the overall strength of 1H 1S 2m stronger and the subsequent bidding slightly easier.

China is a poor country and doesn't have that much money though.

that's paying the due, no big deal.

double, HA and the balanced shape certainly suggests doubling.

I'd bid 3NT, if I get doubled, I may run to 4D though.

I think in words when figuring out what to say or write. But I'm not sure how words could even be involved in many of the things we think about -- and I'm not just talking about erotic reveries. My first wife and I liked to play chess without a board, for example. Don't know how you'd even approach doing that using words. Same with visualizing the opponents' hands at bridge. (Well, maybe that would be possible, but quite awkward.) I played Chinese chess vocally with my friend when we biked home. Then I realized that I rarely think in words.

It's a systemic issue. If you don't have designs for responder to sign off easily after 2NT or reverse, you should pass most marginal hands with 5 HCP over partner's 1 level openings. Still, those sign offs may affect your accuracies in constructive bidding. I actually don't have much feelings either way. When you respond with 4 or 5 card major 4-5 HCPs over partner's opening, you may certainly hit goldmines if he has a super maximum. Still, very often, you may find yourself in an unmakable game or 3M, or 2NT.

2H, or 2D if play suction.

You can probably just bid 2C to show both majors then correct 2H to 2S to show your shape and overall strength. This also may place the contract right. I actually don't quite get why you play transfers here. You certainly want the opener to lead...

Chinese. Could be Shanghaiese because I don't speak Mandarin before I was 6. Still, I can't speak Shanghaiese well now.

Looks very close. Those two tens are very important IMO. I'd bid 3NT for sure when red. When white, there can be other considerations.

I think that is a good idea. Good to see you here again, Tysen <_< Han is of course right that 2 rounds is much more complex than one round. But even if the scope of the project is to develop a system that could cope with an unlimited number of rounds, it may still be a reasonable starting point to construct the opening scheme on the assumption that responder will have to place the contract immediately. Then responder may not necessarily place the contract, but make a descriptive bid that allowed opener to place the contract as effectively as possible. Etc. HotShot: Basically, a good system is a system that wins a lot of MPs (or IMPs or w/e) in the tourneys that we let them play. So after then hands have been bid, they will be played by a DD solver or by looking them up in BridgeBrowser or in some other way. This is straight forward. It is a little more tricky if we restrict us to uncontested auctions, since for example if opps are cold for 7NT while we can make size tricks in hearts, passing the board out would lead to the maximum score but we would say that opening 4♥ or something would be better. So one could restrict the training DB to deals where it is in some sense plausible that opps won't interfere. Richard/Han: Why would you penalize complex systems? Akaike's criterion is meant to avoid overfitting. This may or may not be relevant depending on how we construct construct the systems. Or is it that you think that simplicity is a virtue in itself, regardless of the overfitting issue? I think Man-made systems are actually quite good in many basic areas. The key problem for computer to bid well is not only the hand evaluation problems computers may face. One basic problem is how computer can devise a good sequence and achieve the right and useful information to make important decisions. This is not a problem of language, but more close to a problem of writing a novel or painting, which needs some careful planning and insights. A lot of people can speak English, but few can use it accurately to achieve their goals. In this sense, bidding is not only information sharing and hand evaluation, but also careful planning, insights of rare situations where you can make money to make a difference and others may ignore and bid a second best or third best contract. Of course, neural network may help discover some insights of hand evaluations at early stage of bidding that we may ignore, but so far, I don't find many from the limited literature research I did.

If you want to bid something other than pass here, you should open the hand instead of passing.

With extra values and some shape, suit quality doesn't matter much IMO. Still, I am not quite convinced by his statement not to overcall with a 4 card suit in opener's suit with a top honor. If I have some extra value or good suits, I still think overcalling here is clear, because you may either hit a nice flop or find 3NT against aggressive openers, or find a reasonable partial at relatively low levels. Of course, if you play this style, you can't sac too much at high levels, which is the trade off.