Echognome Posted June 26, 2009 Report Share Posted June 26, 2009 The idea that I'm espousing is to have a separate calculation done for TopTricks in addition to whatever calculations/simulations GIB does for its normal play. I have no idea that Matt Ginsberg is a brilliant person. That does not me he has (or has not) thought about running a separate calculation in parallel. Of course this separate calculation will be additional to any processing time GIB needs. If it was given an order to run this separate calculation first, I think it might be able to offset the processing time, because it wouldn't need to get to the point of simulations whenever the top tricks function kicked in. As Helene mentioned, Jack allows you to claim, but it does not allow you to claim any number of tricks. So I don't believe this is reinventing the wheel (except perhaps that the code for Jack is intellectual property, so I'm sure is not readily available). I personally think it would be an improvement for GIB, especially if it means GIB can claim and accept claims. However, it's easy for me to say, since I won't bear any of the direct costs needed to make any such upgrades. So I just file it under something that would be nice. Quote Link to comment Share on other sites More sharing options...
jeffford76 Posted June 26, 2009 Report Share Posted June 26, 2009 Does anyone know how Jack's claim checker works? For example if all I have left is A10x opposite KJx, can I claim since double dummy I can't lose? Quote Link to comment Share on other sites More sharing options...
helene_t Posted June 26, 2009 Report Share Posted June 26, 2009 No you need a 100% SD line Quote Link to comment Share on other sites More sharing options...
NickRW Posted June 27, 2009 Report Share Posted June 27, 2009 I'm sure one of the DD engines has a "top tricks" function built into it (as I recall reading in a version release report that a bug in it had been fixed). I don't recall which one it is, nor do I know for sure whether GIB has such a function (though it would seem not). Whether it gains much in overall efficiency I don't know - probably. Nick Quote Link to comment Share on other sites More sharing options...
cherdanno Posted June 27, 2009 Report Share Posted June 27, 2009 To bad he didn't discard the diamond instead of the spade on the last club and find the squeeze on north. Probably this could be programmed into GIB by sacrificing speed. I would rather an anamoly like this happen on very rare occasions than that GIB was slowed down on every hand. By a mile. Just checking for "have all the rest in top tricks, and the communication to take them" would be so fast that you wouldn't notice the difference. Other times it would speed things up a lot and you would notice it. If GIB doesn't have this I would be surprised. If GIB has this, but this deal triggered a bug in that routine - I would not be surprised (no offense to GIB, all programs have bugs). Quote Link to comment Share on other sites More sharing options...
hotShot Posted June 27, 2009 Report Share Posted June 27, 2009 Another thing GIB doesn't know anything about is safety plays. But GIB maximizes expected total points, doesn't it? So if a deal where the unsafe line goes down occurs once in its simulations, it should go for the safe line. I think GIB optimizes the number of tricks he can take. Quote Link to comment Share on other sites More sharing options...
golfacer Posted June 28, 2009 Report Share Posted June 28, 2009 For the GIB tournaments, I see a disadvantage to allowing claims. [hv=n=saj7hdc&w=shdc&e=shdc&s=sk4h5dc]399|300|[/hv] Declarer lost count of hearts and is unsure of whether the 5♥ is a winner...so declarer presses the claim button. If rejected, declarer can now try the spade finesse. Quote Link to comment Share on other sites More sharing options...
barmar Posted June 28, 2009 Report Share Posted June 28, 2009 Declarer lost count of hearts and is unsure of whether the 5♥ is a winner...so declarer presses the claim button. If rejected, declarer can now try the spade finesse. The discussion of claims in this thread has been about allowing the robot to claim. As I mentioned above, losing count of a suit is not the kind of mistake GIB is likely to make. There's another thread discussing claims by the human. The best suggestion I saw there was that this should simply allow the bots to defend double dummy, so they'll play faster. It would work well for your example. Quote Link to comment Share on other sites More sharing options...
cherdanno Posted June 28, 2009 Report Share Posted June 28, 2009 Another thing GIB doesn't know anything about is safety plays. But GIB maximizes expected total points, doesn't it? So if a deal where the unsafe line goes down occurs once in its simulations, it should go for the safe line. I think GIB optimizes the number of tricks he can take. Why do you think so? That would be either idiotic or lazy, and I don't think Matt Ginsberg is either. Quote Link to comment Share on other sites More sharing options...
matmat Posted June 28, 2009 Report Share Posted June 28, 2009 The best suggestion I saw there was that this should simply allow the bots to defend double dummy, so they'll play faster. I would actually like to see this as an option to practice declarer play in various contracts, but probably not as part of what they do in tournaments. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.