[benellis58]

West, North, and East are all GIB robots. This explains South's strange bidding. He makes an atypical "weak" (ha, ha!) jump overcall in spades and later commits the cardinal sin of bidding AGAIN with 3S. As he commits these actions, he is well aware of his sins, but he has the firm and long-held opinion that GIB robots are extremely weak at bridge, an opinion that is reinforced by what he sees every time he plays with or against GIB robots. Because of his strong opinion about the abilities of the GIB robots (be they his partner or his opponents), he will often take liberties that he would never dream of taking in a game with decent human players. His experience has shown him that his "stepping out" or "operating" in games with GIB robots will occasionally be costly but will more often pay dividends.

On this board, South buys the contract in 3S. This should easily go down two (via two top clubs with dummy's king in front of East's ace, two top hearts, the diamond ace, and an inescapable trump loser). Down two would be minus 200 for NS, but declarer miraculously gets out for down "only" down one. He ends up minus 100 and scores mildly above average on the board with 57.1 %. He achieves this "miracle" PURELY because of horrendously inept defending by the EW GIB robots. West led the club queen but when it held he foolishly, pointlessly, and inexplicably shifted - and to a diamond of all things. EW could STILL have taken their five basically TOP tricks, but the absolute ROT of their defensive plays continued and declarer never lost a second club! This was a TRULY dismal defensive performance by the EW GIB robots, not to mention that West might well have doubled the final contract.

South was heard to mutter "And THIS hand is a perfect illustration of why I am so often happy to 'step out' in the bidding when playing against...GIB ROB07S."

https://tinyurl.brid...se.com/bdfm26dr

[benellis58]

The GIB robot sitting West doubles 3S. His double is defined as "4+ hearts; 2+ spades; 12+ HCP; 13+ total points, forcing".

As is often the case, the GIB robot sitting West has made a call that does not match the definition. He does not have "2+ spades" - he has a stiff ace. He does not have "12+ HCP" - he has a mere NINE - a full three points fewer than the bare MINIMUM of 12 promised by the definition!

And by the way, is his double REALLY "forcing", as the definition states? What if his partner wanted to convert the double for penalties by PASSING?

EW end up in 4H and go down one, for minus 100 and a lowly EW score of 17.9 % on the board. Perhaps the GIB robots might consider that the philosophy "honesty is the best policy" actually has merit!

https://tinyurl.brid...se.com/bpdh2upp

[benellis58]

The GIB robots are deserving of our sympathy. Why? Because the poor souls are hampered by (among many other issues) their very poor (and often antiquated) GIB system, their frequently awful GIB definitions, and their apparent lack of a desire to signal while defending. Consequently, and not at all surprisingly, they tend to be extremely bad at many facets of bridge - bidding, leads, defending, and even declarer play (which is the best part of their game, but which still cannot be called "good"). One of the numerous reasons that they are so weak at defending is their apparent and totally illogical love for returning whatever suit their opposing DECLARER plays. For example, in the hand linked below, notice that the GIB robot sitting East wins his club king at trick three...and then at trick four returns his remaining club (the jack). Never mind just this particular hand, where a club play at this point rather than the normal and obvious spade play seems bizarre: it's their usual tendency to do this silly type of stuff that is one of the many characteristics causing them to be such hopeless and wretched defenders on a frequent basis.

https://tinyurl.brid...se.com/2p85xa2v

[benellis58]

From the limited number of times I have seen GIBBO in action so far, my impression is that the leads have greatly improved and are no longer as repulsively horrendous as they used to be. Assuming that my first impressions are correct, this is a very positive step towards eventually making the GIBBO robots at least moderately competent.

I think the credit for this long-needed improvement in leads should go primarily to Lorserker and secondarily to the various people who were honest enough to frequently expose some of the massive weaknesses of the GIB robots.

Now it would be great if Lorserker could again work his magic and correct some of the other huge problems with the GIB robots (such as, to name but a few examples, the wretched system handcuffing robot auctions, the extremely poor GIB definitions, and the absence of any intelligent system of defensive signaling).

Lorserker has seemingly performed one miracle, and now - like Sisyphus - he has many more awaiting him. It will be a Herculean task to fix all the other gross and offensive GIB weaknesses, but I have faith in Lorserker. Well done, Sir!

[benellis58]

According to BBO's announcement of the long-awaited and hugely necessary upgrade from GIB to GIBBO, "One of the most exciting improvements comes during card play. In situations where DD analysis concludes that 'it does not matter' which card to play, GIBBO will opt for the common-sense choice, based on typical bridge patterns (e.g., 2nd-hand-low, third-hand-high)."

To this, I say "Bravo", while at the same time asking "Why the Hell wasn't COMMON SENSE (!!!) used before?" I am undoubtedly not the only BBO player who was astounded by the frequency with which the inept GIB robots would stupidly squander tricks left, right, and centre because of their perverse and totally illogical habit of frequently playing second hand high and third hand low. Yes, both second hand high and third hand low can indeed sometimes be the right play, but much more often they are not, so the random card-tossing of the GIB robots was on far too many occasions completely lacking in common sense.

Assuming that Lorserker has indeed remedied this previous ridiculous situation, then he is deserving of even more praise.

[benellis58]

Despite Lorserker's great achievements in correcting some of the most heinous habits of GIB, there is clearly still a lot of work to be done. For example:

The definition of North's 2C is "Free bid - 5+ clubs; 11- HCP; 11-12 total points; forcing".

As is annoyingly so often the case with GIB definitions, the beginning ("Free bid") is entirely unnecessary. I'm calling the definition "GIB" because I think it would be unfair to tarnish GIBBO with the blame, since the huge problem with GIB definitions has not yet been addressed with an upgrade. Thus, I consider GIBBO innocent so far.

A more important criticism, however, is that the GIB definition foolishly states that 2C is "forcing". This is obviously silly and incorrect. North is a passed hand, South opened in third seat, and the GIB definition itself correctly states that North has "11- HCP", so it would make absolutely no sense for this natural bid of 2C to be forcing here. The human South realized this (as did 18 other human Souths) and had a shared top of 83.02% for wisely IGNORING the incorrect GIB definition and passing the supposedly "forcing" 2C, making 8 tricks for plus 90. Fourteen other human Souths also wisely passed, but failed to make the contract. Twenty human Souths, perhaps led astray by the egregious GIB definition, bid over 2C and instantly had no prayer of doing well on the board. One human had passed in third seat and scored 64.15% when the hand was passed out, so he did not have to suffer from reading the misleading GIB definition.

https://www.bridgeba...CD6%7Cmc%7C8%7C

[benellis58]

Here's an example of how Lorserker has succeeded in bringing some SANITY to the robot leads. I have had similar auctions to the one here many times in the past, and the GIB robots used to LOVE leading the suit in which their LHO (their OPPONENT!) had opened a weak two. In those days, I would have expected the GIB robot sitting West to make the ridiculous lead of his stiff 10 of spades. Now, though, with the dramatically improved GIBBO leads, that monumental stupidity seems to be a thing of the past. Here, West quite reasonably leads a club. Once again I salute Lorserker for this huge improvement.

https://www.bridgeba...HQ%7Cpc%7CHT%7C

[benellis58]

GIBBO seems to be a marked improvement over GIB but there is still a LOT of room for improvement. In addition to the obvious things needing huge fixes (such as the really bad GIB system, the poorly written and often incomplete or useless GIB definitions, and the virtually non-existent signaling), there is also the sad reality that the robots still frequently exhibit a lack of logic with their bidding decisions.

[benellis58]

Here we go again with GIB definitions that are poorly worded and/or inaccurate and/or even possibly unwittingly dishonest.

The result, which was fine, is not the problem. The GIB definitions are, as is so often the case, the problem.

North's 3NT is defined as "6+ diamonds; 13-15 HCP". You'll notice that North has a mere TEN (!) HCP...a full THREE less than his announced MINIMUM total! An ACCURATE definition would be: "6+ diamonds, mild slam try" or perhaps even "6+ diamonds, 13-15 TOTAL points". Instead, GIB gives us a definition with blatantly incorrect information.

The definition of North's 5D later repeats the misleading, grossly inaccurate statement that North has 13-15 HCP.

There is no excuse for this, as the transfer to a minor followed by bidding 3NT is not some rare once in a lifetime occurrence.

https://www.bridgeba...S2%7Cpc%7CCK%7C

[benellis58]

The Good News: GIBBO appears to be better than GIB in several ways. Nice to see the progress so far!

The Bad News: There is clearly still a HUGE amount of improvement required, because even the new, vastly superior GIBBO robots still produce some of the worst, most appalling bidding that bridge has ever seen and some of the most atrocious defensive play in recorded history.

Even a miracle worker like Lorserker has a massive task ahead, so I wish him the best of luck.

[benellis58]

On the excellent Bridge Winners site, there is a thread called "Early Returns on GIBBO". In it, someone showed a hand that perfectly illustrates how monumentally stupid the bidding of the GIBBO robots can be. I do not want to step on anyone's toes by posting the hand here, but I will state that the GIBBO robots sitting East-West had an incredibly insane auction to a zero-play grand slam that was doubled and went down two. As horrendous as this auction was, I cannot say that it surprised me in the least, since the perpetrators of the atrocity were GIBBO robots.

[benellis58]

To say the least, I (South) could have done a LOT better on this hand. At IMPs I would have won trick two to protect against a fatal heart shift, and would have banked everything on the hopes of taking six diamond tricks. Here, though, I took the (losing) spade hook for two reasons: First, it was matchpoints, so I decided to be greedy. Second, I was playing against GIBBO robots and even though they're much better than the horrid GIB robots, I figured they still wouldn't be good enough to find the killing heart shift even if my jack lost to the queen.

I was right about that, because my jack did lose to West's queen, and he wasn't good enough to find the heart shift. He returned his last spade, so I was still alive as long as the diamonds came home. Unfortunately (for me, anyway!), after cashing my diamond king at trick four and seeing East's jack, I decided to play for restricted choice rather than a 2-2 split. When I lost to East's queen, the hand caved in and I ended up down...FIVE! Had I guessed to play for diamonds 2-2, I would have taken at least ten tricks (six diamonds, three clubs, and the king of spades), so my misguess cost me SIX (!!!) tricks. But I've misguessed or misplayed other hands in the past and will undoubtedly do so again in the future, so that's not the reason for this report. West's failure to find the heart shift that would have GUARANTEED beating me is not the reason either. The reason is:

the bad GIB definition of North's 3S bid, in combination with North's stupid choice of that same 3S bid.

3S is (hopelessly) defined as "9-11 HCP; 10-12 total points; forcing". Okay, it (SUPPOSEDLY!) tells us how many points he has, but it tells us NOTHING else. If the only relevant thing about his hand is that he has "9-11 HCP", then why isn't he simply either inviting 3NT or bidding 3NT himself? Conversely, if his hand has a suit or suits that might possibly be alternative strains, then why isn't he providing information about it or them instead of ONLY telling about his "9-11 HCP"?

Then, added to the poorly-planned concept of the definition, we see that the GIBBO robot sitting North seems to have either ignored or lied about his own GIB definition, because he does not have the promised "9-11 HCP" - he has a mere seven, a full TWO below his promised MINIMUM.

Now, if instead of either incorrectly or dishonestly bidding 3S, North had either shown his diamonds...OR invited in NT (if GIB has a way to even do that?)...OR bid 3NT for play (using whatever methods GIB has to do that), I would not have complained at all and would not have written this post...and if I had gone down the same FIVE tricks, I would have just said, "I'll have to start guessing better", but because of the poor GIB definition and North's useless and inaccurate bid of 3S, I would have still written this post even if I had guessed diamonds correctly and scored a top, because the GIB atrocities would still have DEMANDED exposure. The posts I write here are not based on whether the results are good or bad - they are based on exposing the many, many flaws of GIB and/or the improved but still flawed GIBBO.

https://www.bridgeba...CC6%7Cmc%7C4%7C

[benellis58]

The GIBBO robot sitting East pulls his partner's double of 4S for absolutely no reason whatsoever. He has no extra heart length, no extra shape, and has not at all misrepresented his hand to his partner until now. Furthermore, he DOES have potentially useful defensive values with the heart ace and club queen (and sometimes even the heart queen). Unless EW grossly misdefended, 4S would have gone down...but EVEN if the GIBBO robots had totally blown the defence, NS would STILL have scored "only" 790 - as opposed to the 800 they got for doubling East's stupid and selfish solo flight to 5H and beating it three tricks.

Just another example of the laughably poor bidding judgment of the GIBBO robots.

https://www.bridgeba...CT%7Cpc%7CC4%7C

[benellis58]

North's double of 4S is perhaps not the greatest of all time. A secondary observation is that he once again ignores his own GIB definition, since the double is supposed to show "9+ HCP" and he has only 7. That's actually only a minor quibble here, however, since it's not HCP per se that defeat contracts, it's tricks - and where does he see FOUR tricks coming for NS, especially with the heart length he has? South saw no reason to overrule North, as South's 2NT call had adhered perfectly to the GIB definition, and South did not wish to be guilty of the same SIN that the inept GIBBO robot was on the hand in the post directly above this!

https://www.bridgeba...CK%7Cpc%7CSJ%7C

[benellis58]

North's bidding is inconsistent and his final bid is illogical.

In an uninterrupted auction, South opens a normal 1D and North bids a normal 2D (inverted minor raise, 4+ diamonds, 10+ HCP, forcing to 2NT). South bids a normal (for the GIBBO system) 2S, showing a spade stopper (and presumably denying a heart stopper, although the GIBBO definition lazily doesn't even BOTHER clarifying that).

If North now bid 2NT, I would not complain, and if he bid 3NT I would also not complain. I think a case could be made for either, and on any given random hand, either might or might not work out.

On this hand, North chose to take the low road by bidding 2NT. Fine, no problem.

Now South, knowing that this is a "best hand" game and also knowing from the GIB definition that 2NT was (only) invitational and showed a maximum of 12 HCP and possibly as few as 10, decided to ALSO take the low road...but in DIAMONDS, not in NT. Thus, he bid 3D, which SURELY should end the auction! North had presumably shown his hand, had invited in NT, and his partner South had converted to 3D, which CANNOT be forcing (or else it would be impossible to STOP in 3D). South could have passed 2NT but didn't, could have raised to 3NT but didn't, and could have made a FORCING bid (something other than 3D!) but didn't. CLEARLY, he wanted to play in...DUH...three diamonds!

North then made the inconsistent call of 3NT. He had earlier decided his hand was only worth an invitation and based on that decision he had bid a PASSABLE 2NT, so it is ridiculous that now, with the exact same hand with which he earlier invited he is now overruling his partner's signoff and unilaterally raising to 3NT! If he had initially bid 3NT over South's 2S, I would not have had a word of complaint even though the final contract and result would have been the same. But what he did here is ludicrous, as GIBBO bidding so often is. And I must again point out that EVEN if 3NT had made, North's bidding would STILL have been inconsistent, illogical, and just plain bad!

https://www.bridgeba...CQ%7Cpc%7CST%7C

[benellis58]

North's double in the passout seat is not defined by GIB. Perhaps it should be, perhaps not, but South figured it should be either penalty or "cards", and therefore passed.

The REAL question, though, is why the GIBBO robot sitting North decided to double with his two HCP in a hand that rated to take ZERO defensive tricks and that almost certainly did not want his side to declare at the four level. His double seems to defy logic (not the first time a GIB or GIBBO robot has done that). It worked out fine for NS, however, because the contract went down one for plus 200 NS and a score of 87.04% on the board for NS.

But why did it go down one? Simple: It went down one because a GIBBO robot was declaring! Although declarer play is by far the best part of the game for the GIBBO robots, it is still not good!

With the very friendly layout, the hand should have made, but the GIBBO robot sitting West, for reasons known only to HIMSELF, stuck in his club 10 at trick three, when he could have played low and slid it to dummy's nine. But even THAT didn't matter, and the hand was STILL cold. His club 10 won (remember that...because it seems that maybe he didn't!). At trick four, he led a spade (trump), losing to South's king.

At trick five, South played clubs for the second time, and...the GIBBO robot declaring from the West seat...POPPED HIS CLUB ACE (!!!)...even though his TEN had WON the previous round of clubs! South then won trick six with his spade ace and cashed the setting trick with his club king!!!!

Yes, declarer play is indeed the best part of the GIBBO robot game, the part at which they are the MOST competent...and even in this they are obviously capable of being grossly deficient, as this hand illustrates!

https://www.bridgeba...S5%7Cpc%7CDT%7C

[benellis58]

Watch how a GIBBO robot magically turns his side's two guaranteed trump tricks into only ONE trump trick:

Declarer South, in 1H, wins the opening spade lead in hand and leads a heart (trump) from his hand towards dummy's K2 doubleton. The GIBBO robot sitting West pops his ace (from) A94 for absolutely NO reason...as exemplified by the fact that at trick three he simply returned a second heart (probably because playing or continuing the suit declarer just played is another of the numerous LOSING habits that the GIBBO robots have). With EW hearts 3-3, declarer now had FOUR heart winners and ONLY one heart loser, the ace.

All West had to do was make the obvious, NORMAL play of withholding his heart ace at trick two and EW would have had the GUARANTEED two heart tricks that they had been born with, but even that outstandingly obvious and SIMPLE play was too difficult for the GIBBO robot sitting West!

https://www.bridgeba...C3%7Cmc%7C10%7C

[benellis58]

EW are GIBBO robots defending against South's 4H. West leads his stiff club. Declarer wins in hand and, since it's matchpoints, decides to risk the heart (trump) finesse. An added reason is that he knows GIBBO robots are defending, and even if the lead was a stiff club, and even if the heart hook loses, the robots are fully capable of not getting their ruff anyway.

The hook loses and East does not give West a club ruff at trick three. He cashes his diamond ace and then at trick four plays...another diamond (!)...so South wins and claims eleven tricks.

So it was all East's fault for not giving West his club ruff, right?

Wrong! When East cashed his diamond ace at trick three, the GIBBO robot sitting West, the same GIBBO robot who had led his STIFF club obviously hoping for a ruff, and holding Q9643 of diamonds, played the...NINE of diamonds (!!!) when his partner cashed the ace! Have I ever mentioned that one of the many reasons the GIBBO robots are such incredibly bad defenders is that they seem to play all their spot cards randomly and appear to have complete disdain for signals? Here it was almost as if West PURPOSELY went out of his way to mislead his partner!

https://www.bridgeba...D3%7Cmc%7C11%7C

[benellis58]

South, the only human in the game, decides to open 1NT with (only) 14 HCP, and NS then end up in 3NT. West makes the reasonable lead of a spade, which South wins. At trick two, South leads a heart, which East wins.

Does East return a SPADE, the suit his PARTNER led? No, of COURSE he doesn't, because he's a GIBBO robot, so of COURSE he plays a HEART, the same suit that DECLARER just played! For Lord knows what reason, the GIBBO robots love to do this sort of thing!

Declarer still goes down one...but ONLY one, rather than the two down that would have occurred if East had returned his PARTNER's suit instead of DECLARER's. But it seems these GIBBO robots will never learn.

NS scored 60.61% on the board.

https://www.bridgeba...D4%7Cpc%7CDQ%7C

[benellis58]

West, North, and East are all GIBBO robots. On this board, there were several different contracts (7D, 6NT, 6D, 5NT, 5D, 4H, 3NT, and 1NT), all of which were played by either North or South.

18 pairs played 6NT, sometimes by South, sometimes by North. Of those 18 pairs, 4 made 6NT, 5 went down one, and 9 went down 2.

6NT is certainly not a bad contract. It will be ice cold any time the hearts play for five tricks, and it will even make an overtrick if the club king is onside. On the actual deal, the club king was offside, but more importantly, the hearts were 4-2, with the long hand having the jack. I went down one, playing the hearts from the top rather than finessing for East's jack.

Okay, you win some, you lose some, so I'm not complaining about being in 6NT and I'm also not complaining about going down one (for which I scored 35% on the board - not the end of the world).

I am, however, complaining about GIBBO bidding - as is often the case, whether the offending robot is a partner or an opponent.

Here, South (myself) actually underbid slightly, as my 3NT was defined as "Strong hearts; semi-balanced hand; 18-19 HCP; rebiddable hearts"...and I actually had a good 20 HCP and the possibly valuable 109 of hearts in my five-bagger.

North's 4NT was defined as "Quantitative invite to 6NT; 3+ diamonds; 14 HCP". Since my 3NT call had been a mild UNDERBID, I had an obvious acceptance and leapt to 6NT, as I assume the other humans also did. BUT...if we look at North's actual hand, we see that he had a mere TWELVE HCP, two fewer than the FOURTEEN that he had promised, so: First, this is yet ANOTHER case of either the GIBBO robot ignoring his OWN definition...OR it is another case of the GIB definition being inaccurate. Take your pick, but EITHER of those is a BLEMISH on GIB! Second, given that South's UNDERBID of 3NT had shown a MAXIMUM of 19 HCP, one might question North's decision to invite slam with "only" 12 HCP...and all the more so since he is telling a LIE and CLAIMING to have 14!

Now, as I already said, there is really nothing wrong with the contract, and it even COULD have been made had I guessed to hook the heart, so I'm not claiming sour grapes. The point is that even if I had made 12...or even 13 tricks, North's bidding would still have been questionable and would still not have adhered to his own definition.

https://www.bridgeba...C4%7Cmc%7C12%7C