georgi

Hi, currently the 18b ACBL tournaments' schedule at BBO is as follows: all times are EDT 11:15AM 12:15PM 1:15PM 2:15PM 3:15PM 4:15PM 5:30PM 8:30PM 10:15PM All tournaments are MP scoring.

I'm in good health :). I still work on GIBs, sorry that i missed to reply in forums. My skill level in BBO rarely deviates from intermediate or advance depend on the tourneys' restrictions i play in.

Thanks for reporting. The reason is 2NT is supposedly pass/correct. Unless invitee has something unexpected to reveal. It's not expected North ( inviter ) to have such nice fit in opening not shown up to now, so he would support second suit and if opener decides to extend this suit would find better fit. Anyway, here the bug caused inviter to "forget" to correct back. Will be fixed in v34.

GIB sitting West would pass on 5♣

Double shows -> -- 2-6 ♣; 2-6 ♦; 5-6 ♥; 2-6 ♠; 18-19 HCP; 22- total points At brief glance it need some polishing of shape, but there is explanation of the double. At least 6 cards in ♠ is no longer possible once hearts are no more then 6 as well. :)

Thanks for reporting it. The problem was this hand was too strong to bid 3♠ ( major first ) and too short to bid 4♠ ( needed 5+ cards ). If N/S weren't vulnerable that hand would have overcall 3♣ instantly. Will be addressed about better solution ( or reducing length in major to bid game ).

Thanks for reporting. There was a bug preventing GIB to rebid its minor with impossible reverse strength but highly unbalanced. So 1NT was found by mistake.

If explanation is wrong, then will be either bug or bad explanation in general. When underneath call and explanation is ok for other(s), then current user had a problem with fetching the appropriate explanation by web server. Network ISP, browser, etc reasons. If that user has problem with GIB server, robot might just stuck and not play, but it won't provide different or wrong explanation. So somehow what you received was wrong. Correct and real explanations ( for all calls ) what were supposed you to receive are: 2NT ( Two NT opener. Could have 5M. -- 2-5 !C; 2-5 !D; 2-5 !H; 2-5 !S; 20-21 HCP ) - 3♣ ( Stayman -- 5+ total points ) 3♠ ( -- 2-5 !C; 2-5 !D; 2-3 !H; 4-5 !S; 20-21 HCP ) - 4♣ ( -- 5+ !C; 3- !S; 11-20 total points ) 4NT ( -- 2 !C; 2-5 !D; 2-3 !H; 4-5 !S; 20-21 HCP ) - 5♣ ( -- 6+ !C; 3- !S; 11 total points )

Now for GIB the scheme is like that: 1NT - 2♥ - ? -> Double -> TOX, 4♠, 8-9HCP -> 3♥ -> 10+HCP 4♠, GF, no stopper ♥ -> 2NT - 3♣ - 3♥, 4♠ stopper ♥ -> 2NT - 3♣ - 3NT - 10+HCP, stopper ♥, no any major more promised, just like to play 3NT and show stopper in overcall. -> 2NT - 3♣ - 3♠ -> 5+ ♠ , INV, 8-9TP 1NT - 2♠ - ? -> Double - TOX , 4♥ 8-9HCP -> 3♠ -> 10+HCP 4♥, no stopper ♠ -> 2NT - 3♣ - 3♠ -> 4+♥ , stopper ♠ -> 2NT - 3♣ - 3♥ -> 5+ ♥, to play -> 2NT - 3♣ - 3NT -> as above, promised 10+HCP and stopper ♠, likes to play 3NT for both sequences 1N - 2M - 3NT -> 10+HCP, no stopper in overcall

Once there is opening ( for easiness level 1 ) and it goes through GIB's hand it limits the points to 16TP ( once pass , unless something latter is specified ). Once partner bids something ( Double, natural suit ) and you pass again it limits additionally the hand.

For GIB at the moment for some auctions if you cuebid their suit ( 4+ cards shown for sure ) once there is no approved suit for the pair ( yet ), you show length in partner's suit. For some auctions it shows only 13+TP and 4+ in your suit. But here it shows 2+ cards in partner's suit along your 4+ major cards.

Who is supposedly controlling the hearts so North would be certain 3NT is still better/safe once partner shows 6+ minor and 3 major cards. What if South is ♠Axx ♥x ♦KQxxxx ♣(K/Q)xx then 3NT would be -1 if hearts are only 5-5 in opponents or -2 or worse?

Thanks for reporting. Fixed in next version - v34. meaning of 2♠ would be -> -- 4- ♣; 6+ ♠; 8-12 total points

Thanks for reporting. We found a bug in this loser(s) computing function, so next version will carry fix for this ( v34 ).

1M - 2m 2OM or bidding higher suit now asks GIB to have 16+TP. The hand in the diagram has 15TP, so it's very close, but still under the range to do so. Later 1♥ - 2♦ 2♥ - 3♥ 3♠/4m Cuebid might have stronger hand, but weaker trump, so that hand was unable to bid 3♥ or 3M instantly over 2minor GF. Will be addressed for some polish as such maximum hands of the minimum range should reveal themselves once partner shows interest on next round.

We found a bug in one of the macros GIB uses to compute losers ( key cards in trump contracts ). It's in progress of fix, and would be supposedly part of next version. That will improve the fluctuations in slams, including evaluation of blackwood "news".

It looks just as bad explanation fetched by web server. There is no bug. Here 4NT is not blackwood. It's denial and suggestion to play 4NT. Over such sequence GIB plays: 2NT - 3♣ 3* - 4 minor 4X suits show -> cue and fit in minor and MAX. 4NT is to play So GIB has option to show preference in this minor as well. Over 4X, 4NT from responder is now blackwood over this minor. GIB would have bid 4♠ as West over 4♣. That denies all Aces below, So East would know their max is 6, missing Ace of diamonds.

Thanks for reporting it. 1NT call fixed in v34. It's back to lower range than 21-23. .

Thanks for reporting. Note taken to be discussed about that sequence where practically 5NT is useless ( with fit in clubs ) as none king is available to show as they are always over agreed trump 6.

GIB doesn't play forcing pass over Splinter ( at the moment ). Will be addressed, so we could add XX as showing void, or XX by opener to show the same.

Fixed in v34. 4♦ came after simulation, but anyway the problem of 4♥ X won't happen with insufficient points.

It's fixed in v33. Now GIB won't deviate from 4 cards in OM.

Thanks for reporting. Unclear how this bug appeared. Anyway, will be fixed in v34.

4♣ is not meant to be splinter. It's one of the rare places where GIB starts cuebidding with K, but not bare King if that's the case. But it could be also singleton. So that is either 1- card or KQ+ suit. So it's a cue in general. Shows a suit where declarer won't lose two cards on lead.

Thanks for reporting. In v33 ( current one ) 5) is removed , 6) fixed. Questions 1) 2) 3) are addressed for review. 4) seems caused by deal simulation flick.