> If playing two hands is always, unambiguously the
> right thing to do, why isn't it consistently,
> unambiguously recommended in BJA3?

It isn't always the best thing to do. Also, different considerations are taken into account for positive and negative counts. In favorable decks multiple hands increases the action we are allowed within the same risk of ruin, but it has a negative card eating effect. It is generally the best thing to do in positive counts and the case for it becomes stronger as we add more players to the table. For negative decks it is the opposite. It has a positive card eating effect but becomes diluted, and less effective, as we add additional players to the table.

We will look at just the case of multiple hands in negative counts here since that is the matter under discussion. I believe in most heads up cases or where there is perhaps one other player hand at the table that multiple hands in negative counts will usually enhance win rate and SCORE. There are two ways we can measure these things, gain per round dealt or gain per hour (time unit). Using either of these measures, multiple hands in negative usually comes out on top. Those that have argued the other side, and you are not the only one, have often introduced a different method of quoting win rate, win per hand dealt, rather than per round. Using that method the multiple hand in negative count model does not finish on top. Those of that school of thought reject the idea that there is time savings in playing multiple hands in negative counts.

While I am of the group that playing multiple hands in negative counts helps considerably, I do not think the gain is nearly as large as it appears in simulations. Sims generally quote in terms of per round dealt without regard to time usage. This method will, of course, give a much bigger edge to the model which uses up more negative cards per round and thus greatly inhances the frequency of positive rounds dealt.