I'm afraid the numbers that have been published are way off. The strategy is very, very weak. About 1/3rd of KO. The risk is so high that it is worse than Basic Strategy for many people.
The first option for counting large cards with OPP amounts to an extrapolation based on an average distribution of large/small cards (one each for every 2.6 cards). The potential for a high risk situation can develop. For example: suppose you start the count at 6 for a six deck game with 5 players (4+dealer). Several rounds into the shoe you are at +10 running(4 extra large cards are available). The next round all players have 20 and the dealer has BJ. You count zero small cards and then subtract 5 from 10 for a total of 5(one extra small card available..-1 running). What is wrong with that? The count should be 0 with 6 extra small cards available. Now suppose at 3 decks remaining the same type of round occurs. By then, you may be betting big bucks into what is a very negative count and be worse off than a basic strategy player.
What is wrong with the concept in general? Not enough cards/rounds before the shuffle which creates a new day, not enough cards for the distribution to average out TRUE each and every time.
jimpenn- In both methods you count the small cards that are seen. The second method applies a number value based on what cards are contained in each hand rather than the, so called, average distribution. I would say that the second method is an attempt to get around the type of problem that I've described. Does it? I don't know because I lost interest after the first attempt. I had seen enough................