home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: rec.games.bridge
- Path: sparky!uunet!gumby!wupost!darwin.sura.net!ukma!taddiken
- From: taddiken@ms.uky.edu (scott taddiken)
- Subject: Which inference is better, WAS - "finesse or play for the drop"
- References: <1992Nov16.143024.5460@athena.mit.edu>
- <lgfm84INNk2u@cypress.cs.utexas.edu>
- Message-ID: <1992Nov16.131237.19210@ms.uky.edu>
- Summary: How strong does point info have to be to overcome distr. info?
- Date: Mon, 16 Nov 1992 18:12:37 GMT
- Organization: University Of Kentucky, Dept. of Math Sciences
- Lines: 23
-
- In article <lgfm84INNk2u@cypress.cs.utexas.edu> varvel@cs.utexas.edu (Donald A. Varvel) writes:
- >
- >I once got a good score in an instant matchpoints game when my partner
- >failed to make an obvious weak jump overcall. The information that he
- >had 6 spades and I had 2 was enough that the booklet commentator thought
- >it obvious to play me for Qxx of hearts which in fact I held. I'm not
- >sure why, but I'm much more likely to use distributional information
- >than weak inferences on high cards.
- >
- This brings up a point which has bothered me from time to time. I find it
- relatively easy to figure out how "strong" distributional info is: i.e.
- "lefty seems to have 5 or 6 hearts, which leaves righty 3 or 4, which makes
- the odds 6-3 or 5-4 that lefty holds the queen." But what if righty is assumed
- to hold, say, 11 of the missing 15 points? What are the odds like now?
-
- The reason I did not state the problem precisely is that I'd like to pose it
- as a general question, hoping to get general answers. I suppose I could
- carry out a simulation or calculation, but are there shortcuts?
-
- ----------------------------------------------------------------------
- Scott Taddiken, University of Kentucky
-
-
-