[ArtK78]

Playing in the BBO Forums daylong tournament today, I picked up a hand like this (the tournament is still running so I don't have the hand record):

Kx Kxx A AKQJxxx

I opened 2♣ (whether this is right or wrong is not the issue - I am playing with a bot) and saw my bot partner respond 2NT. Leaving aside the merits of the 2NT response to a 2C opener (I play that it does not exist with many partners), I thought about just bidding 6NT as the most likely correct contract. However, I checked the meaning of 6NT before I bid it and found that it showed 25-27 HCP. Since that might provoke the bot to bid 7NT when we were missing an ace, I decided to take it more slowly and bid 3C. The bot bid 4NT - RKCB. Again, leaving aside the merits of the weaker hand bidding RKCB, I chose to respond correctly - 5♣ showing 0 or 3 key cards. The bot, holding an ace and the values promised for its 2NT bid, PASSED!


+420 was not a good score.

When will bots learn that the purpose of RKCB is to avoid bidding a slam off 2 key cards, not as some sort of slam exploration method?

[johnu]

ArtK78, on 2019-August-04, 10:27, said:

The bot bid 4NT - RKCB. Again, leaving aside the merits of the weaker hand bidding RKCB, I chose to respond correctly - 5♣ showing 0 or 3 key cards. The bot, holding an ace and the values promised for its 2NT bid, PASSED!

+420 was not a good score.

When will bots learn that the purpose of RKCB is to avoid bidding a slam off 2 key cards, not as some sort of slam exploration method?

Ahh, the ambiguous 0 or 3 key card response. I'm (purely) guessing GIB couldn't figure out whether you could have 0 keycards, so passed the 5♣ response. GIB bids one bid at a time without planning for the next round. I don't see BBO fixing this problem in the next 10 years, if ever.