Systems Archive 4

[ Read Responses | Return to the Index ]
[ Previous | Previous in Thread | Next in Thread | Next ]


Re: Need Clarification on Variations

Posted by John May on 11 August 2001, at 7:53 a.m., in response to Re: Need Clarification on Variations, posted by New Counter on 17 July 2001, at 8:19 p.m.

Thanks very much. My publication "Get the Edge at Black Jack" by John May has misprints or gives incorrect information.

You admit you are a rank newcomer yet you apparently feel you can criticize someone who has been doing this for rather a long time without any substantiation at all. Honestly I throw up my hands in despair some times at the level of pedantry I have to deal with.

For the record, most of these indices have been derived independently and can be found in texts by other authors-as you could found out with a two-minute search on the web.
The reason for the odd discrepancies are a) Most older texts calculate indices on the basis of a flat bet. This is erroneous. Bet spread has to be factored in because certain composition-dependent totals are associated with larger wagers (because of the count) and have greater importance in determining the total-dependent strike number. b) Older texts don't use risk-aversion, ie the largest amount of risk will be traded for EV. This leads to sub-optimal bankroll growth. c) Some indices are time-dependent, eg the extra time involved in splitting a given hand may counter any small amount of extra EV to be gained. d) Indices shift with overall penetration which has been receding steadily through the decades. e) The "floating indice" effect whereby a strike number shifts at a specific level of penetration causes a further bias whereby larger bets should weight the "global" indice differently. f) The cut-card effect alters some indices by, as a practical matter, shorting the deck of a card or two.

I could go on here. The older texts don't deal with any of these problems so don't be surprised at the odd discrepancy, or that their strike numbers may be sub-optimal for your game.


Responses


Password:


The Systems Archive 4 is maintained with WebBBS 2.24.060398.