> Though the simulation method described earlier would
> provide the most accurate values for playing
> efficiency, I question its practicality. From Griffin,
> "Reliable simulation estimates of this...are
> extraordinarily time consuming, so the correlation
> method proves ideal...."

You forget that was was "extraordinarily time consuming" in Peter's day probably can now be done in five minutes. :-)

> The Efficiency Calculator in CVData provides PEs
> instantaneously, indicating that a calculation, rather
> than a simulation, is used to generate the numbers.

No, not necessarily true. Norm has trillions of canned simulations of hands for every conceivable card-counting system. I'd bet that he simply plugged them in to get the PEs.

> In Chapter 4 of The Theory of Blackjack, 6th ed., the
> PE and tags of an optimal single-parameter count were
> determined by calculation. PEs of other counts were
> determined similarly, though, once the tags and PE of
> the optimal system are determined, a simpler
> correlation analysis could be used to provide an
> estimate of other systems' PEs. I suspect the latter
> is what's done in CVData (Norm, please correct me if
> I'm wrong).

See above. Virtually 100% of Norm's work is done by simulation -- although the ROR and goal calculators use directly the formulas from BJA3.

Don