Join Bridge Winners
All comments by Stefan Velja
1 2
You are ignoring the author of this comment. Click to temporarily show the comment.
What do you mean hands are given next to each other? Like consecutive polls, so you can see them both?
In that case, I'd prefer to be in 7c :)
April 24
You are ignoring the author of this comment. Click to temporarily show the comment.
Epilogue: partner has Axx KJ109xx Ax Qx, making 7c an almost laydown contract. One line of play needs either hearts 3-2 or a heart-diamond squeeze, or singleton Q of hearts dropping, which is very highly likely. Another option is to try ruffing a diamond with the partners Q of clubs.
April 24
You are ignoring the author of this comment. Click to temporarily show the comment.
@Steve, partner certainly has both missing aces for his jump to 6 hearts.
April 24
You are ignoring the author of this comment. Click to temporarily show the comment.
Either a regular t/o double with all three other suits, or exceptional strength with other bids being misdescribing. Your answer? :D
April 23
You are ignoring the author of this comment. Click to temporarily show the comment.
What do you expect partner to have for 6?
April 23
You are ignoring the author of this comment. Click to temporarily show the comment.
I added the explanation; it's a continuation of another poll. What would you have bid instead of doubling?
April 23
You are ignoring the author of this comment. Click to temporarily show the comment.
5 would most probably be “to play”, however you may interpret it.
April 22
You are ignoring the author of this comment. Click to temporarily show the comment.
This is not directly related to what the OP asked, but what are exactly the positives of forcing both partners to play the same system, or, say, disallowing them to play different systems at different vulnerabilities? I think allowing both would be very beneficial to the diversity of bridge and allow people to be more creative and to form complementary partnerships, while also forcing them to adapt well to different, unknown situations. It seems to me that it would raise the level of the game.
April 15
You are ignoring the author of this comment. Click to temporarily show the comment.
I find it quite interesting how many players blame north's 1NT opening, but when I posted a poll asking what you would open, the 1NT response dominated.
The poll: https://bridgewinners.com/article/view/bidding-problem-2-lw83cqotl6/
April 15
You are ignoring the author of this comment. Click to temporarily show the comment.
I would say it is very hard to penalize people for playing slowly. One can hardly distinguish between someone spending time looking for some unprobable line of play or wearing their opponents out and someone who is just having a hard time focusing because of noise, stomach problems, or just a catchy song stuck in their head. I agree with most of the other things, though.
April 15
You are ignoring the author of this comment. Click to temporarily show the comment.
https://bridgewinners.com/article/view/atb-aggressive-slam-bidding/ Switched spades and diamonds to see whether people would still downgrade these 15 points, without the “aggressive slam bidding” context.
Jan. 9
You are ignoring the author of this comment. Click to temporarily show the comment.
https://bridgewinners.com/article/view/atb-aggressive-slam-bidding/ Wanted to see how many people would still downgrade these 15 points, without the “aggressive slam bidding” context.
Jan. 9
You are ignoring the author of this comment. Click to temporarily show the comment.
I like to think 2d shouldn't be natural here
Jan. 6
You are ignoring the author of this comment. Click to temporarily show the comment.
I'd start with 2NT as Lebensohl, intending to bid 3h later (4 spades with a stopper). P would break the relay with 3d, after which we would play at least 6d so the road to exploring 7 is open.
Dec. 21, 2017
You are ignoring the author of this comment. Click to temporarily show the comment.
South did think that bidding 4 would be something like Last Train (even though the pair decided not to play it), since the opener's range was so wide, but ultimately decided against it, since he did have only very mild extras for his 5-5 shape.
Dec. 16, 2017
You are ignoring the author of this comment. Click to temporarily show the comment.
Both thought it's a slam try in hearts, but N thought it showed a club cuebid, while S didn't think so
Dec. 16, 2017
You are ignoring the author of this comment. Click to temporarily show the comment.
David, of course, they would have to take into consideration what opponent's bids mean, either natural or artificial (which doesn't even matter since it's all artificial to the AI :) ).

I've seen a research paper on a program developing its own bidding system by itself and it definitely wasn't GCC legal. I definitely didn't mean it as “inventing a system is illegal” :)
You can find an example of such a system in this paper: https://arxiv.org/pdf/1607.03290.pdf

As Wouter said, simply not enough work has been invested into it. Besides, none of the current bridge AI players have any kind of machine learning implemented into them, which is the key to making them superior. This way they kinda depend on human input for the system and some other minor stuff.
Dec. 7, 2017
Stefan Velja edited this comment Dec. 7, 2017
You are ignoring the author of this comment. Click to temporarily show the comment.
AI has been very proficient at poker, which, you must agree, is one of the games where incomplete info has the greatest influence. Thus I don't think that psyches and falsecards would bother it much. It will somehow find a way to find the best percentage, especially after analyzing previous games by the opponents.
Besides, such a program with zero human input wouldn't need to understand things such as “conventions”, since it would develop its own bidding method, which would definitely be GCC illegal. It would have to take competition into consideration, so some human input would need to be incorporated into the opponents, in order for it to have learning material.
If such an effort would be put into bridge AI, I have zero doubts it would be able to win against world's best players.
Dec. 7, 2017
Stefan Velja edited this comment Dec. 7, 2017
You are ignoring the author of this comment. Click to temporarily show the comment.
There's no doubt that a well programmed self-improving program, such as this one, would very quickly adapt to such “bluffs”. If it's doing extraordinarily well in poker as well as chess, I guess it's just a matter of time when it comes to bridge.
Dec. 6, 2017
You are ignoring the author of this comment. Click to temporarily show the comment.
Yes, I meant “total points”, or whatever you might call them, it's just an old habit to write it down as “hcp” instead of “points”. Kxx Qxx xxxx xxx would certainly fall into the 5-8 range, while the other hand you suggested would definitely be bid as 1M-2M.

Thanks for your feedback, it seems like playing 1M-2 is probably a better option.
Dec. 4, 2017
1 2
.

Bottom Home Top