GIB Leads and Followups Consistency

[carmelbobc]

I'm certainly not an expert BBO player. But often I'm confused in reviewing a board when GIB's opening lead is the same for all players, yet its follow-up leads vary for reasons I cannot surmise. Example: In the ACBL Daylong (MP) A -2019-10-10 Board#2. With the bidding the same for all players, S ends up declaring 4H:

 

- GIB's opening lead in all cases was the D4. E plays its singleton D, and NS wins the trick.

 

- If NS takes the CA and then lead hearts, W wins the HA and plays a diamond for E to ruff.

 

- But, when NS next leads hearts (trumps), W wins the HA but returns a club, which lets S make as many as 6H.

 

Very much will appreciate someone enlightening me as to how GIB decides it second lead after the opening lead.

 

CarmelBobC

[smerriman]

The basic logic:

 

- GIB deals a certain number of hands at random, "consistent" with the bidding and play so far

- for each one, it calculates the best card to play, double dummy

- it then plays the card that has the best average result

 

If the bidding/play is identical up to that point, the random seed is equivalent and it will deal the same hands, and end with the same result.

 

If things aren't identical, a different set of random hands would have been dealt, which can give a different result.

[steve2005]

- If NS takes the CA and then lead hearts, W wins the HA and plays a diamond for E to ruff.

 

- But, when NS next leads hearts (trumps), W wins the HA but returns a club, which lets S make as many as 6H.

NS played differently. So EW can play differently

Gib figures out what to do at the moment, what it played earlier doesn't affect its new decision.