I made 4 very mild posts on this subject on rge21.com, and the screams of the clones were such that even the highly fair Parker has caved in rather to them. Similar alleged cracks by the usual folks were of course ignored.
A sausage hides its ingredients, where one is left at the mercy of the FDA, and USDA, but they are tasty and we don't ask. Similar occurs IMHO with claims of a floating advantage. Many effects hide and interact under the skin.
The only part of the FA, that Peter Griffin was able to verify, he put in later editions of his Theory of Blackjack. He explained it rather intuitively: as the true count goes extremely high, your edge tends to top off, with ties and who has the blackjack(s). As the count goes extremely low basic strategy is very inferior to the dealer being able to hit to at least 17, with little chance of busting, while you stand.
This makes the overall curve of edge versus the true count curl downward for very high and low counts. Yet the total sum of all true counts must balance, to the overall full deck, off the top edge, such that this curve will bend into a bow in the middle, to balance, making for slightly higher edges for counts in the middle ranges.
Two things have to be added to Griffin's explanation(s):
The bow effect does indeed float to a very limited degree. This is only because deep penetration means that more extreme counts have a higher relative frequency, resulting in a deeper bow effect.
The same effects that cause the bowing also limit your edge in the higher ranges of your spread. Your edge is lower there, and they are more frequent. You simply don't have the edge for every plus count increasing.
Mixed in with these effects is the added spice of a fallacy in the way true counts are examined in similatations. True counts are never continous. They always occur in discrete steps, where each step is 1 divided by the number of decks left. Your true count steps 0,2,4... if you have 1/2 of a deck left; there is no such thing as a true count between 0 and 2, 2 and 4, etc.
What this means is that if you simulate the edge for a true count range of 1 to 2, with just under 1 deck left, you get a true count only just above 1. Now the same sample tabulation for having just under 4/5 of a deck left, deals with a true count of just over 1.25, and no others. Clearly this inflates the alleged floating edge. It is ludicrous and highly lemming like, for any skilled investigator to have made this mistake in accounting (humorous crack: do all the clones have ties to Enron?), but it is the only effect that would have ALL counts show increased edges. The bow effect only exists because they DON'T. And the bow effect has been verified in several other games besides blackjack.
To further examine the sausage ingredients you have to examine just what is involved when your top end of your spread is less advantageous and is more frequent: you are overbetting if you try to exploit this FA.
I think the renewal of FA claims is just the desperate attempt to "data mine" new returns from conventional counting, without wanting to admit that increased edges do indeed require moving from conventional counting to shuffle tracking, sequencing, location plays, and banger type plays. Heavens, that would require giving, John May, alienated, myself, and Grifter some credit.
Too bad that view and pointing out of missed obvious points is so disturbing to some.
