BBO Discussion Forums: Benellis58 GIB bashing on repeat - BBO Discussion Forums

Jump to content

  • 17 Pages +
  • « First
  • 15
  • 16
  • 17
  • You cannot start a new topic
  • You cannot reply to this topic

Benellis58 GIB bashing on repeat Groundhog Day

#321 User is offline   benellis58 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 344
  • Joined: 2022-July-07

Posted Today, 05:39

Declarer play was what the GIB robots were "best" at - at least compared to everything else in their bridge game...and it is also what the GIBBO robots are "best" at.

However, observe the declarer play of the GIBBO robot sitting West in the hand linked below, and you might conclude (as I have concluded on numerous occasions) that he is not likely to win any brilliancy prizes for his declaring ability!

https://www.bridgeba...H5%7Cpc%7CHK%7C
0

#322 User is offline   benellis58 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 344
  • Joined: 2022-July-07

Posted Today, 05:55

West, North, and East are all GIBBO robots. On this board, North proves that he is as utterly hopeless at defending as his East and West robot brethren are.

At trick nine, North makes the monumentally stupid decision to play his ACE of clubs rather than his jack, thereby allowing EW to score up a game that had ZERO play until the GIBBO robot sitting North botched an obvious and elementary play that no human defender who was even halfway competent ever would after the auction that occurred.

https://www.bridgeba...H7%7Cmc%7C10%7C
0

#323 User is offline   benellis58 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 344
  • Joined: 2022-July-07

Posted Today, 06:10

West, North, and East are all GIBBO robots. North-South collect 100 for beating East's 3C one trick, and North-South thereby score a decent 88.46% on the board. So why am I writing up the hand?

Because I think we need to expose and publicize North's ridiculous and illogical DIAMOND shift at trick three. Looking at THAT dummy after THAT auction, why would anyone ever think a diamond shift was right? Well, anyone EXCEPT a GIB or GIBBO robot, anyway!

https://www.bridgeba...CHQ%7Cmc%7C8%7C
0

#324 User is offline   benellis58 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 344
  • Joined: 2022-July-07

Posted Today, 06:31

West, North, and East are all GIBBO robots. West opens 1H (promising at least a five-card suit) and later leads his fourth best heart (the five) against South's 2S contract. East can EASILY determine from the auction, the lead (which was correctly fourth best, as per the GIBBO system), and the Rule of Eleven, that South has exactly one heart higher than the five, and that ONE higher heart will be the STIFF ACE (because West would NOT be underleading the ace on this auction). Yet the GIBBBO robot sitting East foolishly, wastefully, and most of all mindlessly, squanders his heart queen, rather than correctly playing his four.

At trick two, West pops his club ace, turning East-West's two natural club winners into only one.

Given that EW trumps break 5-1 and the AQ of diamonds are offside, these defensive misplays might or might not be relevant on the hand, but that does not diminish the issue that both plays were subpar. NS scored 73.81% for going down only one.

https://www.bridgeba...C7%7Cpc%7CS8%7C
0

#325 User is offline   Huibertus 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 430
  • Joined: 2020-June-26

Posted Today, 08:20

View Postbenellis58, on 2025-December-30, 06:31, said:


Given that EW trumps break 5-1 and the AQ of diamonds are offside, these defensive misplays might or might not be relevant on the hand, but that does not diminish the issue that both plays were subpar. NS scored 73.81% for going down only one.



Yet as long as West does NOT lead Q (which makes it +1) this contract ALWAYS exactly makes, simply by scoring 1, 1, 1, AK and eloping 3 small spades by ruffing from dummy and (if needed when West ducked K) forcing West to lead 1 round of .

So how did the bots convince the human to go down?
0

#326 User is offline   benellis58 

  • PipPipPipPip
  • Group: Full Members
  • Posts: 344
  • Joined: 2022-July-07

Posted Today, 12:31

Thanks for your return to the thread, Huibertus. Long time no see! There have been many hands discussed here since you last commented way back in September, so I'm surprised that you had nothing to say until now, but I'm very pleased to have you back.

As for the question you asked, my guess is that "the bots convinced the human to go down" by the simple but effective strategy of not allowing him to look at all four hands before declaring. Perhaps that slight advantage of seeing all 52 cards on the random layout of this random hand aided you slightly in your analysis? For sure it would have aided me!

In any case, rather than simply giving the quick and cursory summary that you did, I think it would be much more helpful to all readers (and of course to myself) if you gave a detailed, card by card presentation (with accompanying explanations, of course) of how exactly the play should go from the very beginning and why that specific line should immediately be chosen prior to seeing all 52 cards. Perhaps when one looks at only the North and South hands there is a possibility that different layouts might exist for the East-West hands, so I'd greatly appreciate your learned analysis of how and why the hand should be played from the moment dummy is exposed and trick one is completed - on a single dummy rather than double dummy perspective. That would probably be helpful not only to the human who the bots convinced to go down but also to all the other humans who played the board and scored even less on it than the meager 73.81 % that the human in question did.

When you submit your analysis, please make sure to mention all alternate possibilities for both the declarer and the defenders as each trick is played. That will be highly beneficial. I eagerly look forward to reading and learning from your complete detailed report, so thanks in advance.

By the way, just to make certain that you understand the motivation behind the posts in this thread: The purpose is never to gloat about a good result or to cry about a bad one. The purpose is always to expose some of the many atrocities committed by the GIB (or now GIBBO) robots, in the hope that the powers that be will realize the crucial need to vastly improve the performance of the robots. After all, those very robots are probably the single most important element in the day-to-day activities (and ultimately the success or failure) of BBO. Lorserker recently did excellent work in improving some of the robotic actions (such as leads, for example), but many, many features (system, definitions, defending, etc.) still require considerable upgrading and I sincerely hope that this extremely popular thread that Diana created is helping to point some of those things out.

If you read the 17 pages of this thread, I think you will find a fair number of hands where the human might have done something better, so if he somehow erred on this board, it would certainly not be the first time and will undoubtedly not be the last. However, with all hands posted here, including this rare one that actually seems to have captured your attention, the GIB or GIBBO robots did something bad that needed to be exposed - whether the human was flawless...OR whether he was imperfect.

Thanks again for your valued input. I greatly look forward to your detailed analysis of this hand, covering all possibilities, and clearly explaining why and how the exact plan should be formulated at the point when declarer initially sees ONLY his hand, dummy's, and the two hearts played respectively by West and East at trick one (namely the five and the queen). And, Huibertus, please don't be a stranger! I'm sure most readers would welcome further input from you as much as I would. Welcome back, happy holidays, and a very happy new year to you!
0

  • 17 Pages +
  • « First
  • 15
  • 16
  • 17
  • You cannot start a new topic
  • You cannot reply to this topic

7 User(s) are reading this topic
0 members, 7 guests, 0 anonymous users

  1. Google