Join Bridge Winners

Bridge Winners Profile for Hans Kuijf

Hans Kuijf
Hans Kuijf
  • 0
    Following
  • 1
    Followers
  • 0
    Posts
  • 0
    Favorites

Basic Information

Member Since
Oct. 7, 2013
Last Seen
Oct. 17, 2013
Member Type
bridge player

Bridge Information

ACBL Ranking
None
Sorry, this user has no cards yet.
the state of computer bridge
I have not checked the example in Jack. It was just to make the point clear. As for probability tables (like suit play) I can be very short. Jack itself does it a lot better. Probability tables very rarely work because the auction and the lead and the signals are ...
the state of computer bridge
As for the Qxxx problem. I constructed an example that might help you understand. N: AJxx/AQ/xx/xxxxx S: KT9xx/x/AKQJx/Ax You are in 6S on a C (or D)lead. Humans would play S and if the wrong view is taken they would finesse in H ...
the state of computer bridge
Putting some hate is artificially lowering the expectation of certain positions just because the double dummy evaluation will unlikely produce a good estimate. Positions which have the Qxxx missing for example (and many more) are lowered a bit. This causes the engine to hate ending up there.
the state of computer bridge
English obviously is not my native language but rereading it I see no obvious flaw. Sorry. Can you ask a specific question?
the state of computer bridge
The latter. The lead itself is based on a simulation (strongly guided by experts rules and convention card settings). The analysis of the lead made by someone else is also done with expert rules. The possible layouts are not simulated themselves to check whether the lead with that hand is ...
the state of computer bridge
I am a bit in the dark what you are asking. Perhaps some general example will do. Suppose you miss Qxxx (drop or a (two way?) finesse). The simulation will try to push the decision over the horizon, i.e. finding a way to postpone it for 4 tricks or ...
the state of computer bridge
My program, Jack, uses a vector alpha beta algorithm to effectively produce single dummy results, usually up to 4+ tricks. There still are some horizon effects and for declarer it works much better than for the defender but double dummy side effects are not a big issue. Variance-reduction techniques are ...
the state of computer bridge
WBridge5 is really an excellent program and this play is rather unfortunate. Only the author can explain it I think but after losing the final it might not be the right time to ask Yves about it.
Not following anyone yet
.

Bottom Home Top