Open eseidel opened 10 years ago
./scripts/test-hand 2H 763.63.KQT95.AK4 1H P 2C P
Shows the bug: ForcedRebidOriginalSuitByOpener does not fit hand: And(And(hearts >= 5), And(hearts >= 6))
ForcedRebidOriginalSuitByOpener for 2H looks right in the explorer: http://z3b.saycbridge.com/explore/1H,P,2C,P
This may boil down to stoppers again. It's also possible that the 2N rebid here should be made special and intend to signify more than a minimum hand when the cheaper rebid is available. Right now the 2N is natural notrump which is lumped with all the natural bids as higher priority than ForcedRebidOriginalSuitByOpener: rule_order.order(ForcedRebidOriginalSuitByOpener, natural_bids)
If I remove that rule, then we produce "None" instead of 2N.
I don't think this problem is as difficult as you are making it. In SAYC, a 2/1 bid shows 10+ points and a 4+card suit. Opener shows a 4-card side suit, rebids the first suit with extra length, or NT when balanced (2NT minimum/3NT maximum). With 18 points and no fit, responder bids 4NT (quantitative). In modern bidding, a 32-point 6NT is a good contract. A jump to 4D would show a strong, freaky minor-suited hand (maybe 5D/6C).
http://z3b.saycbridge.com/bid/3-e3a74c1b2cf335e52ba1e1544a:1H,P,2C,P,2N,P,4D,P,P,P#
Yeah. That looks all wrong. The bot probably should be suggesting that you bid 2H as a minimum bid instead of 2N which is slightly stronger.
Even if you bid 2N, jumping to 4D seems wrong. Yes, it's a valid jump-shift, but probably doesn't apply after a NT bid like that. Jumping to 3N would have made more sense (slam is remote at 32 max points between you and no fit).
I also suspect that 2-over-1 would probably handle this better (the bot doesn't know that system yet) since part of how this auction got into that trouble is that the 2C bid N made has such a huge range.