> Thanks for the response Don!

> What if I were to count each 7 as +0.5 and each 9 as
> -0.5 instead? I am guessing that this would be a
> "watered-down" Halves count, since only the
> tag values for the 2 and the 5 would now differ from
> Halves. Do you have any idea what type of improvement
> this count would have over Hi-Lo?

I can only guess that it might be marginally better; but I wouldn't bet on it.

> Also, since Hi-Lo
> has a better insurance correlation than Halves (per
> table D18 of BJA3), would I want to just disregard the
> side count info for insurance decisions (assuming this
> new hybrid count also has a worse IC than Hi-Lo)?

Yes.

Don