[thorvald]

---

4♠= 11- HCP, 5-12 total points

How bad a hand evaluation is that?

[thorvald]

In the 2012 version 4♠ promised 3+♥, and then it might be more reasonable, and the bot using the old version will pass 3♠

In those days the bot would open this hand with 3♦

[bobmcc]

Do the programmers ever explain this type of behavior?
Do they try to fix the problems?

[smerriman]

bobmcc, on 2022-September-08, 22:20, said:

Do the programmers ever explain this type of behavior?
Do they try to fix the problems?

The programmers stopped working on GIB 3.5 years ago.

But the explanation is simple, and comes up regularly in the forum - GIB adds points for shortness, regardless of whether it's in its partners suit or not. It shouldn't, but it does, and that's about all there is to say about it.

Edit - I lie, there's a little more to say about it, and the shortness in hearts probably isn't the explanation. In this case GIB believes the 3♠ bidder is showing a very good hand (something like AK 7th in spades with at least a KQ outside). In order to make up your 2♣ / 3♥ bids, it believes you're basically guaranteed to have a spade void and all the other points, making 6♦ a virtual certainty.

Not sure what Thorvald means about the older version GIB passing 3♠; it doesn't, it considers various options then jumps to 6♦ and expects to make (no other call is even remotely close). 4♠ is the slower route taken by the later robots.

What did West actually have to bid 3♠? And was 6♦ making, or was 2♣/3♥ a stretch and not what the robot would do?

[thorvald]

smerriman, on 2022-September-08, 23:01, said:

The programmers stopped working on GIB 3.5 years ago.

But the explanation is simple, and comes up regularly in the forum - GIB adds points for shortness, regardless of whether it's in its partners suit or not. It shouldn't, but it does, and that's about all there is to say about it.

Edit - I lie, there's a little more to say about it, and the shortness in hearts probably isn't the explanation. In this case GIB believes the 3♠ bidder is showing a very good hand (something like AK 7th in spades with at least a KQ outside). In order to make up your 2♣ / 3♥ bids, it believes you're basically guaranteed to have a spade void and all the other points, making 6♦ a virtual certainty.

Not sure what Thorvald means about the older version GIB passing 3♠; it doesn't, it considers various options then jumps to 6♦ and expects to make (no other call is even remotely close). 4♠ is the slower route taken by the later robots.

What did West actually have to bid 3♠? And was 6♦ making, or was 2♣/3♥ a stretch and not what the robot would do?

I think 6♦ is a much better bid, I wonder If I would pass that,

The board is not visible as my opponent hasn't played it, but I think my hand was

---

I might have made an error in the simulation, as the bot also bids 6♦ (by far) now

I was wondering if it was better to bid 4♥, than 3♥ showing longer ♥, but now West will not bid 3♠, and the bot will still jump to 6♦

I think learning to subtract points when void in partners suit is important