[johnu]

benellis58, on 2025-August-24, 21:39, said:

Strengths of the GIB robots:

They don't revoke
They don't lead out of turn

They don't bid out of turn.
They don't make insufficient bids
They don't double their partner or make illegal bids

[manudude03]

benellis58, on 2025-August-24, 21:39, said:

They don't revoke

Or do they? https://www.youtube....l=PeterHollands

[mycroft]

Everyone already knows what I am going to say here, but since this seems to be the thread for broken record playing, why not?

I do not play with pickups on BBO. Haven't for over a decade. Why? Well, apart from listing the "weaknesses of the GIB robots" (which, I guarantee you you will eventually see and more):

• They play a system, but it's not anything I've ever played, but of course it's "just standard";
  
• They fill their profile with conventions, but pass 1♠-1NT; 3♦;
  
• They play Gerber and Blackwood, and assume you play them in the same places and with the same responses they do, even if they don't define them in the mess of conventions in their profile; and
  
• They yell at you if you ever make a mistake (even if it's only a "mistake" because of the above three, or even only a mistake in their minds or double dummy. I still remember the person (live, but didn't have to be) who complained about my play on the first board of the night, where "if I had taken the finesse, I would have make the overtrick". Yes, but if it lost, I went down 2 in a cold contract!).
  
• Oh, and they're likely to disappear without notice (whether or not they yell at you first).

Funny enough, I do play with GIB robots, especially when I have a 25 minute bus ride or am waiting for a doctor's appointment, and I'm likely to be guilty of the last myself. Good, bad, or otherwise, I know what they play and they're not going to surprise me today with something they wouldn't have done yesterday.

And they won't yell at me (and don't care if I yell at them, but the people on the bus or in the waiting room might, so I don't. :-).

What I don't understand about all the GIB-bashing is why, if they think they're so bad compared to "a human player", they haven't made the same decision I have, biased toward their opinions. I do have guesses though. SMerriman - I have different guesses for :-).

[benellis58]

On a positive note: It must be stated that although the GIB robots are staggeringly bad at all facets of the game, their abject incompetence does provide pleasure each and every day, because it is always fun to see the new and creative ways in which they demonstrate their gross ineptitude.

[pescetom]

benellis58, on 2025-August-24, 21:39, said:

Weaknesses of the GIB robots:

System: Very bad
Definitions/Explanations: Very bad, often incorrect, sometimes non-existent
Judgment: Does it even exist? In any case, very bad
Leads: Very bad
Defending: Very bad

Strengths of the GIB robots:

They don't revoke
They don't lead out of turn

I think that is partly unfair.

The system is not very bad, just a bit quixotic and poorly documented.
The leads are not very bad overall, just a bit unusual.
They have an important strength you ignore, above average declarer play.

[benellis58]

I respectfully disagree. Their system is abominable. Their leads are atrocious. I will concede that declarer play is indeed the best part of their game, but if you consider it better than average you must play against a huge number of palookas.

[Huibertus]

This thread is nonsense.

When I still actively played Bridge I was playing at the 2nd best Dutch competition level, I know how to play. When I play here and a human gets replaced by a robot due to disconnects or thinking too long I'm always happy. There are humans that are better then the bots, but the vast majority here is not. And yes Bots do strange things, due to them thinking differently, also bots do brilliant things 95% of humans never find and also humans mostly are clueless about the game.

[benellis58]

My comment to Huibertus: Please read the last 9 words of the post directly above yours.

[benellis58]

I play robot games every day. Obviously I enjoy them and am pleased that BBO offers them. This, however, does not diminish the total contempt I have for the "ability" of the GIB robots. I play every game fully expecting horrendous play from West, North, and East on hand after hand. I frequently observe all three GIB robots contributing atrocities on the same hand. I expect incompetence from them and am surprised whenever it is absent. It is quite interesting that the games remain enjoyable despite the gross incompetence of the GIB robots.

[benellis58]

Here is an example of the mind-blowingly stupid bidding of the robots:

EW are silent. North, a GIB robot, passes. South opens 1S.

North bids a non-forcing 1NT. South bids 2C, showing "5+ spades, 4+ clubs, 12-18 total points" as per the definition.

North has A3, J932, Q98763, 10. 2D and 2S would both be reasonable calls and either might or might not work out best depending on the layout of this random hand. North opts for 2S. Fair enough.

South, with K8765, KQ, A, AJ762, bids 3C, showing, as per the definition, "rebiddable clubs, rebiddable spades, 17-18 total points".

North now stupidly bids 3D. If he wanted to show diamonds, he should have earlier bid 2D. Once he decided on 2S earlier, however, it is now ridiculous and inconsistent to bid 3D, and all the more so when South is now known to have many black cards.

South bids 3S, which is defined exactly the same as the 3C bid was.

North now INCREDIBLY stupidly bids four HEARTS for NO logical reason whatsoever! South bids 4S, which goes one down.

There are countless examples of hopeless bidding by the inept GIB robots. This is but one.

[diana_eva]

benellis58, on 2025-September-09, 02:05, said:

Here is an example of the mind-blowingly stupid bidding of the robots:

EW are silent. North, a GIB robot, passes. South opens 1S.

North bids a non-forcing 1NT. South bids 2C, showing "5+ spades, 4+ clubs, 12-18 total points" as per the definition.

North has A3, J932, Q98763, 10. 2D and 2S would both be reasonable calls and either might or might not work out best depending on the layout of this random hand. North opts for 2S. Fair enough.

South, with K8765, KQ, A, AJ762, bids 3C, showing, as per the definition, "rebiddable clubs, rebiddable spades, 17-18 total points".

North now stupidly bids 3D. If he wanted to show diamonds, he should have earlier bid 2D. Once he decided on 2S earlier, however, it is now ridiculous and inconsistent to bid 3D, and all the more so when South is now known to have many black cards.

South bids 3S, which is defined exactly the same as the 3C bid was.

North now INCREDIBLY stupidly bids four HEARTS for NO logical reason whatsoever! South bids 4S, which goes one down.

There are countless examples of hopeless bidding by the inept GIB robots. This is but one.

---

[benellis58]

Thank you, Diana.

[johnu]

From the Bridgewinners site, a recent post said

Quote

Rob Brady declared in his most recent podcast that he was quitting BBO because of the robots' newfound lack of discipline.

"Bradybot" claimed that it defeats the promise of enjoyable play if you cannot make reasonable inferences while playing bridge. I agree.

[Huibertus]

benellis58, on 2025-September-09, 02:05, said:

Here is an example of the mind-blowingly stupid bidding and declare play of the humans:

This is from the Free Daylong Tournament (IMP) - 2025-09-10.

100 players played this hand as South. ONLY 7 of them reached game. 93 did not realize that;
1. knowing where the bulk of the opponents High cards are improves your odds.
2. when you have decent odds of having 7 quick trick in a NT contract (♦ behave) you are WAY too strong for a 1NT or other overcall.

Of the 7 that reached 3NT ONLY 1 managed to combine deducing West leads a singleton so you should not play ♥Queen, playing on assumptions (West ♠King 5th, ♦must be 3-3) and counting the rest of the points to work out a ♣ or ♥ throw in (depending on what East discards) must bring you the 9th trick. Of others playing NT part scores only 11 made 9 tricks (i haven't checked how).

Oh, and do note the EXCELLENT defense of the bot to duck ♠ King as;
1. The bot is not sure of partner having 3 or 2 spades. Ducking only disables declarers throw in as declarer can not remove easts 3rd ♠ exit in case he has 3.
2. It provides declarer a losing option, to play East for KX by ducking and taking the 2nd round with the Ace.
I highly doubt more then 5% of the humans would have found ducking here.

https://tinyurl.com/2ctoxhkh

This hand is NOT rocket science, just deducing, counting and assumption play.

[pescetom]

Huibertus, on 2025-September-12, 04:49, said:

This is from the Free Daylong Tournament (IMP) - 2025-09-10.

100 players played this hand as South. ONLY 7 of them reached game. 93 did not realize that;
1. knowing where the bulk of the opponents High cards are improves your odds.
2. when you have decent odds of having 7 quick trick in a NT contract (♦ behave) you are WAY too strong for a 1NT or other overcall.

Of the 7 that reached 3NT ONLY 1 managed to combine deducing West leads a singleton so you should not play ♥Queen, playing on assumptions (West ♠King 5th, ♦must be 3-3) and counting the rest of the points to work out a ♣ or ♥ throw in (depending on what East discards) must bring you the 9th trick. Of others playing NT part scores only 11 made 9 tricks (i haven't checked how).

....

This hand is NOT rocket science, just deducing, counting and assumption play.

Surely finding diamonds 3-3 is 36%, maybe even less once East has shown hearts? Hardly decent odds.
It is disconcerting if any human failed to count West with a singleton, agreed.
But doesn't this contract only make at all because West robot fails to take trick 2?

[Huibertus]

pescetom, on 2025-September-12, 10:11, said:

Surely finding diamonds 3-3 is 36%, maybe even less once East has shown hearts? Hardly decent odds.
It is disconcerting if any human failed to count West with a singleton, agreed.
But doesn't this contract only make at all because West robot fails to take trick 2?

"Surely finding diamonds 3-3 is 36%, maybe even less once East has shown hearts?"

Looking at ♥ alone that is a valid point. But ♠ were ALSO bid, and in that case by West. That actually increases the change of a balanced ♦ split 3-3 at worst more likely, or of course 3-2 or better.

"But doesn't this contract only make at all because West robot fails to take trick 2?"
That will never happen, it is double dummy play. Having played a few times with bot we should be able to see, and factor in that unlike humans the Bots will NEVER play for partner to hold more then bidding allows (assuming declarer upgraded). And given that 93 out of 100 human declarers WOULD have had more given the same bidding sequence as evidenced by the traveller sheet, they take the right approach here.

[pescetom]

Huibertus, on 2025-September-12, 12:29, said:

"Surely finding diamonds 3-3 is 36%, maybe even less once East has shown hearts?"

Looking at ♥ alone that is a valid point. But ♠ were ALSO bid, and in that case by West. That actually increases the change of a balanced ♦ split 3-3 at worst more likely, or of course 3-2 or better.

Fair enough that spades balance hearts in deviation from 36%, in any case the deviation is small and we are nowhere near useful odds at MP.

Huibertus, on 2025-September-12, 12:29, said:

"But doesn't this contract only make at all because West robot fails to take trick 2?"
That will never happen, it is double dummy play.

West knows from the fact that partner did not raise spades that he has two cards at most. West holds KJxxx and can see Qxx in the dummy, South promised a stopper. Who will he play for Axx?

Huibertus, on 2025-September-12, 12:29, said:

Having played a few times with bot we should be able to see, and factor in that unlike humans the Bots will NEVER play for partner to hold more then bidding allows (assuming declarer upgraded). And given that 93 out of 100 human declarers WOULD have had more given the same bidding sequence as evidenced by the traveller sheet, they take the right approach here.

Not sure what you are arguing here, not that I am very interested in psyching robots or exploiting their foibles. If West robot trusts his partner (as is my experience) who showed at least opening strength and less than three spades then he still has no good reason to expect the Ace. If it also thinks South should have more for his bid and should have a stopper, that speaks clearly.