Hi Calypso,

I mostly agree with Don, but since this is an intellectual exercise, here is my two cents worth.

I once wrote a program to calculate by brute force all the EOR's (ie betting, and all playing!) for a given game. Then I linked the resulting EORS into a BJ simulator, using Griffin's 'perfect play' scenario. That is the individual EOR for each play was used as a side count, as well as separate insurance and betting EOR's. I found that even in shoe games the SCORE can be doubled (N0 halved), compared to a strong betting count (BRH-I,Halves,etc), but almost all of the gain came from the lifting the playing efficiency up from under 60% up to nearly 100%. The difference between using the betting EOR's instead of a strong point count was negligible.

Also, strangely enough, the results were relatively insensitive to which EOR set was used, that is EOR's calculated for 8 decks worked just fine in 2 decks and vice-versa. I concluded factors such as count distribution, penetration and especially betting spread were much more important than exact EOR values.

So my advice would be to forget exactitude with the betting EOR's, and look up the playing EOR's (ignoring those affected by no-hole-card) in Griffin for the Illustrious-18 and use them as side counts for playing.

For such a flat spread as 1-4 in a game like this, you really need to avoid negative counts. Also your ramp needs to be very steep, almost straight from 1 up to 4 to make such a spread close to optimal.

Brett.