How can GIB be possibly taken seriously...

[MightyMoe]

Really. How can the ACBL offer Masterpoints when GIB is involved and this kind of thing happens?

[hv=pc=n&s=sjt94hkq6dat9ck53&w=sk82h8dk843caj984&n=sq6hat5432dj62cq6&e=sa753hj97dq75ct72&d=s&v=0&b=11&a=1d2c2d(%3F%3F%3F)]399|300[/hv]

 

The standard excuse of "It is just like a human partner making a mistake at the table.", doesn't cut it. Not even the worst LOL partner would come up with a stupid bid like this.

 

I also now how hard it is to program artificial intelligence systems -- I do so for a living.

 

This is atrocious.

[Stephen Tu]

The standard excuse of "It is just like a human partner making a mistake at the table.", doesn't cut it. Not even the worst LOL partner would come up with a stupid bid like this.

 

I don't know about that. I've seen plenty of nutty bids from humans in my life. Maybe they are unlikely to make *this* particular bid, but they can certainly come up with laughers if you examine the travelers closely.

 

I think it's just a bit more jarring coming from GIB since its baseline bidding level on most hands is quite a bit better. One expects weirdness in the bidding when playing with a human beginner. But playing with an advanced partner, it's jarring when they do something unexpectedly bad.

 

This particular bug has been reported a bunch, we've been complaining about it for a long time now. GIB's floor for bidding 2h is way too high (requires GF values), preventing it from making that bid. It doesn't want to neg-x since it has only doubleton spade (I don't like neg-x either), and it feels 2d is better than passing. So it's not totally ludicrous. They just need to allow it to bid 2h on weaker hands. If opener had been say 4-1-5-3, a fairly common distribution, the bidding would have worked out fairly well, I don't think you'd be complaining here quite so vehemently with the "atrocious" and "how can ACBL offer masterpoints" comments.

 

BTW South also contributed with the non-std 1d opening, GIB expects 1c on this shape, not "better minor". When 4432 is the only 3 cd diamond opener, raising on 3 is more attractive.

[Vampyr]

Why can't GIB be programmable -- not totally, but with something like check boxes? Then everyone could configure it (within limits) to their liking.

[Stephen Tu]

Why can't GIB be programmable -- not totally, but with something like check boxes? Then everyone could configure it (within limits) to their liking.

 

The PC version of GIB is programmable with check boxes. Has options for various conventions, NT ranges, etc. Wouldn't fix the bug this post complains about though.

 

For BBO version, I guess they didn't do it to avoid having to create a UI to control the options then send it to the particular GIB instances you are using.

 

I'm more annoyed by the bugs than lack of options. I can live with the set of conventions GIB uses, the only one I really loathe is Capp esp the way GIB employs it bidding 2M with 4cd major freely. There is a tremendous plus side to not giving user options though -- it makes it much, much easier to debug. The best bid in a lot of situations depends on what set of gadgets you are using.

 

To really program GIB one needs to learn its complex database/programming language.

http://orig.gibware.com/bidding/

http://orig.gibware.com/bidding/format

 

It seems BBO is now only releasing the mb.txt output file in its updates for the PC version, and not the original mb.txt2 source file. I only have a very old 1999 mb.txt2 source file; it would be nice if BBO were willing to release an updated one with current fixes so tinkerers with the time & interest could experiment.

[cloa513]

Programmable doesn't work in practice because they need to do the hard yards of careful work on continuations- the history of this forum shows that this is never done.

[cloa513]

This hand is symptom of the high standards for a competitive bid- you need 14+ TP for that so GIB is left with poor choices.

[georgi]

According to GIB, 2♥ requires 11+TP, not 14+. Upper hand has 9TP so can't qualify for 2♥. But even if can, 2♥ is 1R forcing.

 

Anyway, is there a chance that hand to be given in some link or at least to be full?

 

The hanging in the air 2♦ bid is not possible to be replicated in neither of the previous 2 versions.

 

The only possible bid which might comes up is 2♥, although rare. (IMP and MP scoring taken into account).

[Bbradley62]

Anyway, is there a chance that hand to be given in some link or at least to be full?

 

The hanging in the air 2♦ bid is not possible to be replicated in neither of the previous 2 versions.

[hv=lin=pn|mightymoe,~~M36360,~~M36358,~~M36359|st%7C%7Cmd%7C1S49TJH6QKD9TAC35K%2CS28KH8D348KC489JA%2CS6QH2345TAD26JC6Q%2C%7Crh%7C%7Cah%7CBoard%2011%7Csv%7Co%7Cmb%7C1D%7Can%7CMinor%20suit%20opening%20--%203%2B%20D%3B%2011-21%20HCP%3B%2012-22%20total%20points%7Cmb%7C2C%7Can%7CTwo-level%20overcall%20--%205%2B%20C%3B%2010%2B%20HCP%3B%2011-18%20total%20points%20%7Cmb%7C2D%7Can%7C4%2B%20D%3B%206-10%20total%20points%20%7Cmb%7Cp%7Cmb%7Cp%7Cmb%7Cp%7Cpc%7CH8%7Cpc%7CH2%7Cpc%7CH7%7Cpc%7CHK%7Cpc%7CD9%7Cpc%7CD4%7Cpc%7CD2%7Cpc%7CDQ%7Cpc%7CC2%7Cpc%7CC5%7Cpc%7CCA%7Cpc%7CC6%7Cpc%7CC8%7Cpc%7CCQ%7Cpc%7CC7%7Cpc%7CC3%7Cpc%7CD6%7Cpc%7CD7%7Cpc%7CDT%7Cpc%7CDK%7Cpc%7CD3%7Cpc%7CDJ%7Cpc%7CD5%7Cpc%7CDA%7Cpc%7CHQ%7Cpc%7CS2%7Cpc%7CH3%7Cpc%7CH9%7Cpc%7CH6%7Cpc%7CD8%7Cpc%7CHA%7Cpc%7CHJ%7Cpc%7CC9%7Cpc%7CH4%7Cpc%7CCT%7Cpc%7CCK%7Cpc%7CS4%7Cpc%7CSK%7Cpc%7CS6%7Cpc%7CS7%7Cpc%7CCJ%7Cpc%7CH5%7Cpc%7CS5%7Cpc%7CS9%7Cpc%7CC4%7Cpc%7CHT%7Cpc%7CS3%7Cpc%7CST%7Cpc%7CS8%7Cpc%7CSQ%7Cpc%7CSA%7Cpc%7CSJ%7C]360|270[/hv]

with this traveller:

http://online.bridgebase.com/myhands/hands.php?traveller=1253-1345563601-52366246&username=mightymoe

[georgi]

Thanks.

 

Maybe too small percentage of happening for hundreds games. Maybe the simulation should be disabled here.