[virgosrock]

Rule should be very simple:
If simulation suggests insane double/triple jump to 6 or 7 use RKC anyway. Can't be difficult to implement.

---

[steve2005]

At least this could have a play if east has Q♥. Often on these jumps there is zero play.

Costs nothing to ask for KC and find out missing Q♥

Most ridiculous jump I have seen.
2♣-2♦-2♠-6NT
Missing 7♥ and 7NT

Here jump was ridiculous and premature checking for KC would not have helped find the heart fit.

[ahydra]

Well, it might be slightly harder to implement than most GIB bidding fixes, as it probably requires code changes rather than bidding database changes.

ahydra

[virgosrock]

ahydra, on 2018-June-25, 02:35, said:

Well, it might be slightly harder to implement than most GIB bidding fixes, as it probably requires code changes rather than bidding database changes.

ahydra

Well, if BBO got the binary probably impossible. IF database changes the algorithm should be something like look for all double and triple jumps to 6 or 7. Insert rkc and responses where possible.
I am probably oversimplifying hugely.

[Cascade]

4♥ is an overbid. No fix is going to correct all the problems from an overbid.

[johnu]

Cascade, on 2018-June-25, 17:19, said:

4♥ is an overbid. No fix is going to correct all the problems from an overbid.

So what if 4♥ is an overbid. The point is that a percentage of hands are going to be short a keycard if GIB jumps to 6 or 7 when RKC would allow GIB to discover the partnership is short on keycards. e.g. An overbid hand would have a 90% chance of being short on keycards, a down the middle hand might have a 25% chance of being short on keycards. Would you feel better if you had the maximum for your bidding and GIB put you in 0% slam?

[smerriman]

Quote from another thread:

jdonn, on 2018-April-26, 07:55, said:

It has been a long time goal of mine to have a rule created that GIB always bids blackwood on the way instead of jumping to 6 or 7 of any suit.

Maybe in the next couple of years they'll get around to implementing this.

[virgosrock]

johnu, on 2018-June-26, 00:21, said:

So what if 4♥ is an overbid. The point is that a percentage of hands are going to be short a keycard if GIB jumps to 6 or 7 when RKC would allow GIB to discover the partnership is short on keycards. e.g. An overbid hand would have a 90% chance of being short on keycards, a down the middle hand might have a 25% chance of being short on keycards. Would you feel better if you had the maximum for your bidding and GIB put you in 0% slam?

Oh! Cascade. Thou knowest not what thee speaketh or something like that.
It is an approximate world we live in but we hafta use the tools to make it more predictable.

and johnu thanks once again for standing up for the downtrodden!

[Cascade]

virgosrock, on 2018-June-26, 04:14, said:

Oh! Cascade. Thou knowest not what thee speaketh or something like that.
It is an approximate world we live in but we hafta use the tools to make it more predictable.

and johnu thanks once again for standing up for the downtrodden!

I know more than you give me credit for.

Ironically, it is you by mixing it up with 4♥ that is not using the tools you were given to make your result more predictable.

GIB is a tool. It bids and plays a certain way. Yes, some things could be improved. However, it is the old adage "garbage in garbage out" when you are dealing with a computer. If you bid 4♥ and GIB thinks you have something approximating 20 hcp and you don't then you cannot expect GIB to make a good decision whatever its rules. GIB does not play bridge the way that people do. Perhaps on this auction, it could have more and or better rules. If you choose to play with GIB then you have to accept its rules - it is absolutely inflexible in that regard.

[smerriman]

Cascade, on 2018-June-26, 18:14, said:

If you bid 4♥ and GIB thinks you have something approximating 20 hcp and you don't then you cannot expect GIB to make a good decision whatever its rules.

That's true, but not relevant. The point was even if you *do* have 20 HCP, 7♥ is still ridiculous, because you can easily be missing the ♥Q.

Simply tweak the diagram so South has 20 points, and the bug is still the same.

This has been mentioned many times in the past, but bug reports in this forum need to be approached from the perspective of "if the human hadn't lied, does GIB's bid make sense"? In this case, the answer is no, so it is a bug regardless.

[virgosrock]

Cascade, on 2018-June-26, 18:14, said:

I know more than you give me credit for.

Ironically, it is you by mixing it up with 4♥ that is not using the tools you were given to make your result more predictable.

GIB is a tool. It bids and plays a certain way. Yes, some things could be improved. However, it is the old adage "garbage in garbage out" when you are dealing with a computer. If you bid 4♥ and GIB thinks you have something approximating 20 hcp and you don't then you cannot expect GIB to make a good decision whatever its rules. GIB does not play bridge the way that people do. Perhaps on this auction, it could have more and or better rules. If you choose to play with GIB then you have to accept its rules - it is absolutely inflexible in that regard.

Firstly Cascade I apologize if I offended you. Was trying to be funny.

The point was to illustrate the importance of RKC vs. Insane jumps. smerriman has also clarified in a later post.