georgi

Here it wasn't flash or network problem. Just GIB had no explanation for what 4♦ is :). It's when there is opening 1minor - * - 4minor. Noted to be added proper meaning.

Thanks for reporting. It's fixed in next version. Problem was GIB had a flaw in Final Destination logic we started using since two versions ago. So here Final Destination ( the call we passed or we show willing to play on it limit us to no more than 5 more steps calls, or in other words if 2♦ is passed by us, we could go no higher than 3♦) should stop GIB by proceeding. So here partner ( GIB ) used his normal call to go up to our limitation. Now the "hole" allowed to GIB to re-raise additionally. In next version no matter who raises up to our limit in next round(s), we would be still under control of Final Destination. So neither East or West will get the chance to raise over partner's 3♦ even if auction was: I.e. 1♥-2♦-P-Pass X/2♥ - 3♦ - P - forced pass 1♥-2♦-P-Pass X/2♥- P - P - 3♦ P - forced pass Once human voluntarely breaks the level and go up, GIB will proceed according to current sequence, pass is no longer an obligation.

There was another solution/fix which is applied. GIB could have used direct 2♠, showing that hand and it's also forcing. 1♣/♦ - 1♦/♥ - 2♥/2♠, 6+ cards and 13+(GF) That hand had nothing else to show beside this call.

XX is addressed. 2♠ comes after GIB became dizzy about how many hearts are there at the table. After South's double they turn out to be 15 at least. So when something is unclear GIB could be confused and found 2♠ as lower escape than 3♦. So if XX shows something clear, the subsequent double by South would be more easy for GIB to swallow.

Note taken to be addressed. As could be similarly 2X - P - 4X - P P - Double

2♠ is bug for sure. 3♦ shows fit in opening at the moment. X could be one of the solutions, but i am investigating for the proper fix about it. 3♠ at the moment is 6+ cds/ 13+TP and forcing. So it's another call for GIB. So note taken and solution is on the way. Will report it when i got it.

1♣ - 1NT you could just rebid 2♣, partner has 3+ clubs for sure and unlikely 5+♦. At least unlikely GIB to have. 2♦ for GIB here would require reverse strength. 1♣ - 2NT Here 3♦ would only promise 14+ TP so this hand matches and could find this call. In all cases 2♣/3♣ are not forcing, so based on 1NT/2NT you will find the part-score you'd like. If you want to invite for game you have to risk a little, as for GIB at the moment there is no invite with 5-4 in minors. Even if you have opened 1♦ problem remains if partner bids 1NT/2NT. What would you do over NT's ? 1♦ - 1NT 2♣ - is passable 3♣ - is GF So the solution seems not safe unless there is some new path. Nevertheless, the chance partner to rebid on 1st level ( unless Soloway ) is much higher anyway than 1NT/2NT.

At the moment for GIB 2♦ is with semi-unknown strength. Max is unknown if hand is unbalanced and the longest suit is weak. Bal hand would go through 2NT which is 8+ HCP, but if you have weak by quality one suiter hand, i.e. Qx QJx Qxxxxx xx you would bid 2♦ and then could also give positive over 2♥ i.e. although having only Queen of trumps. So 3M could hold variety of positives which responder wished to show on second round. Or even singleton if hand is xxx QJx x Jxxxxx

With that hand, GIB would open 1♣. GIB would open 1♦ with 5♣4♦, only with hand under 16TP which has super diamonds and lousy clubs, where the expected reverse is impossible and rebidding 2♣ with xxxxx, while having AKJ10/AKQx in dimonds would remain concealed.

Sorry if i didn't describe the full sequence and it went unclear. The main point was when responder's calls are GF. Opener's rebids are intact as he still supposedly shows some values and stoppers. He could use Splinter to show shortness of course, etc. But 3z in my upper post was after (*) ( or any rebid by opener ), where the asterisk symbol covers these rebids for easiness. So an example described fully is : 1♣ - 2♣ 2♥ - 3♦ -> GF by responder while 1♣ - 2♣ 2♦ - 2♥ would show only hearts stopper and still INV+. opener could bid 2NT or 3♣ and then responder could proceed with max or pass with min. For GIB, opener's calls are Splinters once when there is jump, starting from 3♦ and up, but not 3♣ ( which shows just stopper ).

[hv=http://www.bridgebase.com/tools/handviewer.html?bbo=y&lin=pn|sm101,~~M19411,~~M19409,~~M19410|st||md|4STH57AD78JQKAC29K%2CS5678H69JD469C8TJ%2CS3QH248QKD2TC46QA%2C|rh||ah|Board%202|sv|n|mb|2S|an|Weak%20two%20bid%20--%206%2B%20!S%3B%2010-%20HCP%3B%209%2B%20total%20points%20|mb|3D|an|Overcall%20--%205%2B%20!D%3B%2012%2B%20HCP%3B%2019-%20total%20po|mb|3S|an|The%20Law%3A%209%20trump%20-%3E%203%20level%20--%202%2B%20!S%3B%204-15%20total%20points%20|mb|d!|an|Responsive%20double%20--%204%2B%20!C%3B%202-%20!D%3B%204%2B%20!H%3B%205-%20!S%3B%207%2B%20total%20points%20|mb|p|mb|5D|an|17%2B%20HCP%3B%20strong%20rebiddable%20!D%3B%2019%20total%20|mb|p|mb|7D|an|4%2B%20!C%3B%202%20!D%3B%204%2B%20!H%3B%205-%20!S%3B%2017%2B%20total%20points%20|mb|p|mb|p|mb|p|pc|S5|pc|S3|pc|SJ|pc|ST|pc|SK|pc|D7|pc|S6|pc|SQ|pc|D8|pc|D9|pc|DT|pc|D3|pc|D2|pc|D5|pc|DA|pc|D6|pc|DK|pc|D4|pc|H2|pc|C3|pc|DQ|pc|C8|pc|C4|pc|S4|pc|H5|pc|H9|pc|HK|pc|HT|pc|H4|pc|H3|pc|HA|pc|H6|pc|H7|pc|HJ|pc|HQ|pc|S2|pc|CA|pc|C5|pc|C2|pc|CT|pc|CQ|pc|C7|pc|C9|pc|CJ|pc|C6|pc|SA|pc|CK|pc|S7|pc|DJ|pc|S8|pc|H8|pc|S9|]400|300[/hv] Thanks for reporting. It's fixed in next version ( supposedly similar jumps to 7 in something where there is major distinction between the hand GIB promises and what it is in real ). Here North would raise to 6♦ as maximum ( and very likely behavior ).

1♣ - 2♣ (*) - 3♣ and 1♦ - 2♦ (*) - 3♦ are not forcing as they were now. 1m - 2m (*) - 3z is forcing where "z" is something else than our minor.

When GIB searches for calls and it finds an appropriate call where all constraints are ok, it doesn't need to simulate ( still based on call properties, could sim by force if there is some ambiguity ). However, based also on priorities there are some safe-belts which should protect GIB to throw a pass or select emergency call once there is a risk to make the opposite. So here what GIB found was inappropriate double based on just values ignoring possible fit. In old version West GIB gave automatically double ( simming doesn't help as the double rule was accurate ). Now in next, GIB will evaluate this lower level double additionally, and respectively having such hand like here will not use it. Instead will now consider pass / 5♥. Old version 5♥ wasn't found at all, now 5♥ is found in almost all cases for advanced-GIB.

Fixed East's double about requesting dummy's suit lead to match the expected suit by East and therefore this East will pass now as the hand doesn't correspond. Opener's XX and Pass are actually identical as such XX hasn't been planned in GIB's diary. Then from North's POV the hearts are really mismatched. East promised 5+, West 3+, South 3 and North himself holds 4. That reflected to an error evidently.

Ok, some examples 1m - 2m 2y - 2NT -> INV only, auction could stop here 1m - 2m (*)- 3z -> GF 1m - 2m 2y - 2z -> INV+, but not GF

4♣ has not really meaningful explanation for GIB now and it causes issues with further explanations. Will be addressed for more appropriate meaning than just 5+ cards and super max hand.

Not missed to be mention on purpose. Probably seeded as minor tweak. For next release could be added a little more extended note about what was changed in forum headlines. In lobby news are usually brief essentials.

All escapes from 1NT - Double - p - 2x are asking and promising currently 5+ cards and 4 or less HCP. Anyway will be addressed if it looks unsafe and 4+ is good enough.

Improved, so GIB won't use such last chance doubles once there is huge fit existing. So if it leans on double, will be in clear missing fit and points.

In last link 3♠ showed actually 5+ major cards which is impossible as South would have bid them in the lower level and won't bid 3♣. Fixed this in next version, so once South ( opener rebids his minor, major cards will be limited to 4 ( over FSF sequence ). That make the task a bit easier so to raise this North needs 4+ cards ( and also will promise them ).

Tried to replicate as with couple of dozens ( in new version ) it doesn't replicate and GIB stayed calm on 4M. Neither do old. It also stayed with small amount of hands. Then after some increase of bigger test new version has actually still 0% proceeding here after 3K hands simming. But old one had 447 x 5♦ after also 3K hands. So i believe similar tempted partners will not be encountered. Or let say in much less cases if any.

Thanks, fixed in next version (v33).

Fixed explanation in next version. -- 3+ M; 4- HCP Will check about why this hand didn't raise. 3M promises 5+ TP , so this hand qualifies in theory.

Thanks for reporting. Fixed in next version (v33).

Some versions ago it was added the automatically raise-correction by person who answered to 4NT. So if GIB answered 5♣/5♦ and partner bids 5M as sign-off, GIB used to correct to 6M showing 3/4 KC. Since v32 ( current now ), GIB does it only over 1/4 possibility. Here actually it's little more easier for South to reveal as GIB bid game over INV, showing 14+TP. Apparently it couldn't be with 0 key cards based on South's hand and having in mind the overcall and probable clubs values.