georgi

4♦ is better as responder is not in the right to decide final contract after opener initiated slam try and responder is rather limited ( especially in this very case with 10-12TP ), so better to just describe himself. On the other hand, but that's only my personal call i would bid 2NT instead of 3H, showing GF HCP values. Over expected club or none singleton 4H would deny spade control so partner would be in control again to decide whether to proceed or not. Over 3♠ rebid by opener showing singleton spade there would be early Christmas / Easter :) Here GIB would do same, but it has 12TP because of ♦A in doubleton. If the Ace was in spades or clubs 2NT would follow as needed 13+TP.

2♦ looks like reverse and fixed in next version to show appropriate for reverse values: so 2♦ here would be -> -- 5+ ♣; 4+ ♦; 21- HCP; 16-22 total points So now, instead of pass, over 2♦ ( based more on 12-22 range ) GIB-W would bid 3♦ -> 4+ ♦; 5-8 total points

I am not saying 4♣ meaning of 21-22 is expected from all players as with quite less, but distributional hand slam would be cold. I am just explaining why GIB did this and what its expected. Cuebidding is as it seems not so polished in GIB's brain, but the goal is to be better for sure. Some major difference is as it seems the evaluation of hand which GIB would have more problems to see through. Human, of course, could easier "sniff" slam for which GIB needs algorithm. Nowadays GIB needs mostly Aces to cuebid, sometimes it might do it with King if holds a void ( which in theory cannot show ). To be replaced this with Kings+ or singletons/voids it needs more work to be described. There is no doubt GIB needs improvement. And it comes with every version. Unfortunately not so comprehensive as expected.

Good point, thanks. It will be investigated.

Sometimes there is lead directional double I.e. 1♦ - p - 1♥ - p 1NT - p - p - X -> Penalty double; requests ♥ lead -- 13+ HCP; rebiddable ♥; 20- total points

You're right that 4♣ denies ♠A, but 4♦ should show in theory at least K+ in spades. As otherwise if both hands don't have spades why we cuebid further. Also 4♣ shows 21-22 TP, hand almost for 2♣ opening. So a hand that North imagines is something like: ♠KQx ♥AJxxx ♦K ♣AKxx So practically missing ♠A is the only one defensive trick.

And South may shift after ♠A won, while East has no clue where is ♥A yet to ruff at first trick. If you are unsure how to proceed, maybe let opponents to help instead.

1NT would be: -- 5+ ♣; 5+ ♦; 3- ♥; 3- ♠; 11-16 total points; forcing But based on unfavorable vulnerability it needs 14-16TP With favorable vulnerability it would bid 2NT immediately. Also later if there is such a chance.

Looks like rare case as main behavior of GIB is moving in 2♣ ( even after simulating pass is not competitive call ). Would be investigated.

It used to be like that ( raising minor showed two fits ) but change was reverted in previous version. Now you show only one fit, though for minor you have two options, to bid unbid suit below 3NT or raise minor directly on level 4.

4♠ has a limit up to 18TP, as with 19TP GIB would double first time. 2♠ is also limited to 10TP, as with 11+, GIB would bid 2♥. Both cases would be adjusted in next version. Including the overcall in 4th seat.

North hand could be or 11TP or not having 3+ in rest suits. Or also could have 12-15TP, but something like both majors and OM is 4th. So cannot overcall with it.

Thanks for reporting. There used to be a bug in the sequence disallowing to proceed with 4 cards and no explanation even if you proceed if having 5. The only one info you get when you bid new suit on 4th level would be 11-16TP, but no suit length promised at all. Now would show 4+ cards and would let GIB to proceed. In diagram in question GIB would bid 4♣ ( 4+ ♣; 11-16 total points ) , 4+ cards in OM are with priority if exist. Level 3 was ok for responder to bid new suit with 4+ cards and info shows this normally.

Explanation for repeating new suit after: 1X - 1Y 2X - 2Z * - 3Z/4Z has been fixed to show 5+ cards in second suit. So in upper diagram would show: 1♦ - 1♠ 2♦ - 2♥ 2N/3♣/3♦ - 3♥ -> -- 5+ ♥; 5+ ♠; 12+ total points; forcing to 3N Similar if you jump to 4♥, but if you mouseover 3♥, maybe no reason to do it.

With same hand, but moving one of minor jacks in majors or making 9TP would make GIB using 2♣ and over 2♦ will bid 2♥ -> Invite major -- 5+ ♥; 4+ ♠; 9 HCP. So GIB would pretend it has 9HCP with 8HCP and 9TP, but somehow clear values in both suits.

GIB would rebid 1NT over 4th seat double with any hand that would rebid 2NT ( 18-19 ) over partner's bid in uncontested auction if that describes better. Would bid other suit or something else only with unbalanced hands.

When GIB shows supperaccept, 4 trumps in transferred major, it does only with 17HCP, and using 2NT with 4333 and if it's max with doubleton it bids it almost natually. The exception here would be doubleton in hearts, which GIB would show with 3♠. Idea in this sequence is to keep retransfer always possible so responder to keep 1NT opener declarer. Here North has 16HCP and therefore cannot bid supperaccept with 4 cards.

yes, GIB in v30 was improved to look at vulnerabilities.

Thanks for reporting. Fixed in next (v30).

For GIB, even opening in 3rd seat still asks for 12+TP. GIB would deal hands based on that fact. You have shown ♣KQx, ♠Axx and ♦Qx and ♥x, and hand like 4 spades + 3 clubs and 5 diamonds and 2 hearts or 3 hearts and 4 diamonds. Declarers promised 15+HCP, but showed only 7 up to this moment. Should have 8-10HCP more. If you had 5 diamonds you would have +1 TP with Axx.xx.Qxxxx.KQx But in that case declarer would be KJ in diamonds which it's not like that after play of 3♦ and Jack did not fall. So here declarer very likely already has ♥AQJ and you don't have ♦J, but missing TP might mismatch somehow GIB's plan and it went wrong and instead of finesse for -1 jumped with Ace and exits with 10.

Thanks for reporting. Fixed for such balanced and strength corresponding to 1NT rebid ( 18,19 ) to bid it instead. There was also bug preventing 1M to "see" the rebid 1NT over 4th seat balancing double. 5332 shape.

Next version will contain little help in this line. Probably for human bidders most likely. So if you bid once,twice, etc and reach slam, with partner being supposedly silent all along, he will not move. Of course if you have shown two suits or somehow last one is not your longest suit GIB would consider correction, but otherwise will calmly pass and rely on your decision to bid on your own. Example: 1♠ - p - 2♠ - 3♥ 4♠ - p - p - 6♥ p - ? ( GIB will stay ) directly 6♥ also implies, just more bidding will not change behavior of partner. 1♠ - p - 2♠ - 3♥ 4♠ - p - 2♠ - 6♦ p - ? ( Now GIB would look at cards before passing and would correct to better spot.)

Thanks for reporting. It was improved to be sensitive about vulnerability for raising after double to 3 or 4 level. It applies also for overcall. Like: 1m - 2M - Double - 3M/4M

This 4NT is not defined. So for GIB is just to pass it.

Currently GIB would bid 2♥ with priority, so explanation would be corrected to 3-♥. Even with hand 54(04) GIB won't skip it.