[iandayre]

After 2C-2D-2NT-3C, since GIB does not play Puppet Stayman, the normal response to deny a 4 card or longer major is 3D. On that auction with the following hand:

KQ, xxxx, J, 9xxxxx

GIB bid 3NT. The meaning of a 3NT rebid by an opener is unclear to me, since you would want to leave room for responder to show a 4-5 or 5-4 hand if held. Perhaps it should show 2-2 in the majors, but that is not the description. The descriptions of 3D and 3NT, while not written identically, appear to amount to the same thing - 22-24 HCP and no 4 card or longer major. But yet, with the hand above, GIB bid 4C over 3NT.

So what IS the difference between 3D and 3NT?

[mike777]

As I understand the GIB, it may bid the exact same hand in the exact same auction differently. IN this case sometimes it will rebid 3d and sometimes 3nt. It is how its program is built. For example perhaps 75% of the time it will rebid 3d and 25% of the time it will rebid 3nt.

I think what posters find confusing is they expect GIB to make the exact same bid all the time with the exact hand and auction.

[steve2005]

Why would Gib not bid 3♦ with no 4-card major always? Gib uses Smolen which cant be used over a 3N rebid, so if responder has 5-4 in the majors you may miss a fit.

OP made this point but I think it needs to be emphasised as 3N is functionally inferior unless it carried some special meaning which it appears not to.

Besides, the system notes (which apparently are out of date) do not give 3N as a possible response.

[Bbradley62]

mike777, on 2014-December-02, 14:52, said:

As I understand the GIB, it may bid the exact same hand in the exact same auction differently. IN this case sometimes it will rebid 3d and sometimes 3nt. It is how its program is built. For example perhaps 75% of the time it will rebid 3d and 25% of the time it will rebid 3nt.

I think what posters find confusing is they expect GIB to make the exact same bid all the time with the exact hand and auction.

I don't think this is the case. I think OP is giving a hand where some human openers rebid 3♦ and other human openers rebid 3N, but the descriptions of those two alternatives don't make it clear which one a human is suppose to select when bidding.

In placing a contract, GIB will sometimes do simulations and will make decisions based on a small sample size, but this is not supposed to happen earlier in auctions, and particularly not when GIB is responding to a conventional asking bid.

[Bbradley62]

3♦ is explained as "No major suits -- 2-5♣; 2-5♦; 2-3♥; 2-3♠; 22-24 HCP; 23+ total points",
whereas the explanation of 3N is "3-5♣; 3-5♦; 2-3♥; 2-3♠; 22-24 HCP..."

Ignoring the slight difference in minor suits, it would be my guess that the latter is not supposed to be part of the GIB bidding system, but it was put into the explanation grid simply so that GIB wouldn't freeze if his dumbass human partner made this bid.

[mike777]

Bbradley62, on 2014-December-02, 18:01, said:

I don't think this is the case. I think OP is giving a hand where some human openers rebid 3♦ and other human openers rebid 3N, but the descriptions of those two alternatives don't make it clear which one a human is suppose to select when bidding.

In placing a contract, GIB will sometimes do simulations and will make decisions based on a small sample size, but this is not supposed to happen earlier in auctions, and particularly not when GIB is responding to a conventional asking bid.

Why doesn't the robot's hand always match the description?

When humans play bridge, they don't just follow rote rules for bidding; they often use their judgement to find better bids, or fill in holes in their system. We would love it if we could program judgement into GIB, but that would be pretty advanced artificial intelligence. As with many game-playing computer programs (e.g. chess programs that routinely beat grandmasters), we substitute brute computational power for thinking. Many of GIB's rules allow it to perform simulations.

GIB starts by finding the matching bid in its bidding rules (we call this the "book bid"). If simulations are allowed, it then makes some adjustments to its hand (adding a card to each suit, adding/subtracting a few total points) and finds the book bids for those similar hands. Then it deals out cards to the other hands at the table consistent with the rest of the auction. For each hand and possible bid, it does 2 things:
1.Determine how the auction will probably continue if it makes that bid (to avoid exponential complexity, it only considers book bids for this continuation, not what would happen with subsequent simulations), and
2.Calculate the double-dummy result of the final contract determined in step 1.
It then selects the bid whose expected value is highest across all the hands. This takes into account the form of scoring; this is how we emulate rules of thumb like "bid games more aggressively when vulnerable at IMPs". The description that's displayed comes from the specifications of the bid that was chosen; this corresponds to the standard rule about disclosure in bridge: you must describe your agreements, not your actual hand

[Bbradley62]

One of the primary reasons that GIB's bid don't always match the descriptions is because there are (at least) two separate parts of GIB programming: one that tells GIB what to bid and another that displays explanations at the table. Sometimes, there are mistakes in the coding and the two are not consistent; correcting this is a major function of this forum.

I still think you're overstating the frequency with which GIB simulates during auctions, and will await clarification from staff/programmers. I don't think GIB will ever, for example, bid 3N over 3♣ as several humans did in the OP. Another example would be when GIB decides whether to super-accept a transfer; I think this is also decided by rules, not by simulation. It appears to me that these simulations are reserved for final placement of contracts.

[mike777]

Bbradley62, on 2014-December-03, 21:53, said:

One of the primary reasons that GIB's bid don't always match the descriptions is because there are (at least) two separate parts of GIB programming: one that tells GIB what to bid and another that displays explanations at the table. Sometimes, there are mistakes in the coding and the two are not consistent; correcting this is a major function of this forum.

I still think you're overstating the frequency with which GIB simulates during auctions, and will await clarification from staff/programmers. I don't think GIB will ever, for example, bid 3N over 3♣ as several humans did in the OP. Another example would be when GIB decides whether to super-accept a transfer; I think this is also decided by rules, not by simulation. It appears to me that these simulations are reserved for final placement of contracts.

3nt would be a "final placement of contracts"

[gordontd]

mike777, on 2014-December-03, 23:18, said:

3nt would be a "final placement of contracts"

But a premature one, by the hand that is not in charge.

[mike777]

gordontd, on 2014-December-04, 03:08, said:

But a premature one, by the hand that is not in charge.

OK, you tell that to the machines.

[johnu]

mike777, on 2014-December-03, 21:14, said:

When humans play bridge, they don't just follow rote rules for bidding; they often use their judgement to find better bids, or fill in holes in their system.

Hmmm, when my partner bids Stayman and I count the cards in my major suits and if I don't have 4+ cards in either major, I bid the appropriate level of diamonds. I don't use judgement to find a better bid because in response to a standard Stayman bid, there are only 3 expected responses. You can't go around making up responses to a conventional bid and expect partner to figure out your brilliancy.

[mike777]

excellent points and I agree.

Again feel free to tell that to the machines.

My Bridge Encyclopedia indicates there seems to be more than one standard reply but your points are excellent.