[FrankerZ]

---

Arena challenge. Not sure what to make of that pass, system indicates that I'm only promising three diamonds, so I don't know why it doesn't show hearts/clubs/anything other than pass? I can't really rebid spades and indicate I have six...

[steve2005]

I think Gib should be 3♣ which is weak.
Perhaps in Gib database shows a better hand. It should 2N is the INV and can be off-shape.

[virgosrock]

steve2005, on 2018-May-27, 08:52, said:

I think Gib should be 3♣ which is weak.
Perhaps in Gib database shows a better hand. It should 2N is the INV and can be off-shape.

agree s2005. 3C is the bid. Maybe it went astray because of stiff spade, the 2D bid which might indicate c shortness. Still, all-in-all, a terrible pass (if 3C is drop bid).

But then "everybody is in the same boat". Today you lose MP/IMP/$ because of GUBBO's strangeness, tomorrow you will gain when it does same to your opponents.

The worst arrogant rationalization I have been hearing for 10+ years. All that is needed is "We are working on it" and appease all.

Sorry to rant.

vrock

[Stephen Tu]

I'm pretty sure this is related to the problems discussed earlier in: http://www.bridgebase.com/forums/topic/78690-100-gibug/page__view__findpost__p__950938
about lack of invitational 6+m range.

3c is like 10-, so this hand is "too strong" to trigger the rule for that, since it counts out as "11 total points". 3C can only be reached via simulation thus probably only with advanced bots and not basic?
The hand falls into the 11-13 total point "hole" (Slightly too aggressive an evaluation IMO because it shouldn't treat spade shortness as asset, 10 total points maybe not 11), also with a stronger 4H-6m that isn't a GF you pretty much always have to bid 1nt even if an invitational jump shift were available to not miss hearts, and do some guessing whether to underbid with 3c or try 2nt if partner bids 2d.
With the basic bots prob it needs some rules added to steer it into 3c with this hand shape and this amount of HCP even though the bid won't exactly match the description of 10- TP.

Quote

All that is needed is "We are working on it" and appease all.

Vrock, there have already been many posts by BBO personnel who state that. Jdonn has said it like dozens of times recently. You aren't appeased because the progress is slow and difficult.
And I think you concurred in another post that from a competition standpoint it is still a fair situation since GIB isn't singling out people to mess up with; it spreads the wealth so it evens out.
So make your criticisms more realistic and accurate! Quit complaining that it's losing you money, because really you aren't since it evens out. Gripe about the quality of product vs. the rake, fine, not your random money bridge losses which are cancelled by GIB's random gifts to you.
I don't see how it's an "arrogant rationalization". It's a rational stance. The games are fair. They aren't expensive. People are still playing despite the crazy bot moves. There's no real clearly overall better competitor site to flee to (some may have better robots, but not offer the user base, # and variety of tournaments, etc.). So really it's a take it or leave it situation for you. It's not rational for you to get so angry about it, post your bugs but tone down your gripes or stop playing/find somewhere else to play, IMO. BBO is slow to improve GIB because they don't have a competitor pushing them and the games are cheap enough that people don't care too much about the bots screwing up. It's kind of like McDonald's; if you are cheap enough it doesn't matter that much that the food is mediocre, just needs to meet the min quality to keep customers coming in. Until you get competitors near the same price range with better food there's not as much incentive to improve the product.

[steve2005]

Stephen Tu, on 2018-May-27, 13:37, said:

The hand falls into the 11-13 total point "hole" (Slightly too aggressive an evaluation IMO because it shouldn't treat spade shortness as asset, 10 total points maybe not 11),

Yes, well if this hand falls in the 11-13 whole then i'm shocked that Gib can't evaluate it's hand. <sarcasm>

My pet peeve. Gib has to be giving pts for singleton spade and doubleton diamond as Stephen indicates. If anything Gib should be subtracting points. Anyone who has been playing 2/1 for any bit of time knows your best chance is to play in 3♣. and i don't care how many points you think you have it isn't as much as you think.

[Stephen Tu]

It's kind of OK to count some short suit points since that's cancelling out some of the long suit points that it ought to be counting but isn't. Just the formula is crude and not as accurate as a good human player's point count.

After investigating this further, it looks like my initial guess was wrong; it's not that it's too strong for 3c. The computer thinks the suit is too weak for 3c; it wants to have the T of clubs instead of a small spot for the suit to qualify as "twice rebiddable", then it bids it. Maybe a description bug also since that suit should already qualify as "twice rebiddable". That's way too stringent with 2 only diamonds. Any 6 is maybe fine with 1-2 in opener's suits. Maybe you want better if doubleton spade or 3+ D.

But it needs the simulation to overrule its book programming to see that 3c will work out better, simulation is what gives the computer a dose of what appears like common sense. Simulation is mostly (entirely?) disabled for the basic bots though. Just illustrates how dumb computers are unless you specifically force it to think things over before passing 2d, they just blindly follow the rules they are given and won't pick anything that doesn't fit a rule exactly unless you give it additional code to simulate between multiple close but not exact fit options.

[pescetom]

Stephen Tu, on 2018-May-27, 13:37, said:

There's no real clearly overall better competitor site to flee to (some may have better robots, but not offer the user base, # and variety of tournaments, etc.).

Are you sure of this?
I've only played a few trial games on Funbridge for instance, but I was impressed by how easy it was to configure a 2/1 bidding system close to my own preferences and it seems to have variety of tournaments plus significant participation. If I was interested in playing against robots for money I would check it out and would not be surprised it it was better.