An alternative to using true count to determine playing strategy would be to to use 2 parameters:
1. Pen = (Number of cards seen) / (Total number of cards in deck)
2. RC = Count's running count

Parameter 2 is always known to counter. A good estimate of parameter 1 could be quickly learned by counter.

(Pen,RC) is a point on a graph. Its position relative to known plotted data in graph determines strategy.

Example: Generic insurance using HiLo for 1 to 8 decks: http://www.bjstrat.net/genHiLoIns.html

Advantage is it is simple. Disadvantage is it would be impractical to consult a graph while playing. Could player sufficiently commit graphs to memory to play seamlessly? If so it would be a great method.

Side counts could be incorporated. The graph of the known data for each possible side count value would have variable starting and ending points and differing trajectories. For example, if aces were being side counted there would be 4 graphs for single deck generic HiLo insurance (1,2,3,4 aces removed). (There would be 32 graphs for 8 deck generic HiLo insurance.) Multi-parameter side counts would result in very, very many graphs.

Hope this illustrates the difficulties of side counts in addition to the generic HiLo insurance example.

k_c