Encode entire deck with one card (thumb at 0 or 2 or 4 or 6)

The Magician sends his assistant to another room.
The Magician removes the four sevens from a deck of cards claiming they are detectives.
The magician has a spectator select any other card as the target card to be the criminal and places it face up.

The Magician picks one of the sevens and gives it to the spectator telling them to hold it with a tight grip with their thumb on top and take it to his assistant.

The assistant looks at the "detective" card (the seven) and reveals the target criminal card.

How this is done:
The Magician selects the seven that is a suit one notch below the target cards suit SHoCkeD fashion.

If the value of the target card is an odd number the magician orients the one-way pointer on the detective card to be away from the spectator.

If the value of the target card is 8 or above the magician turns the detective card face up.

The Magician manipulates the spectator to hold the detective card so their thumb is in one of four positions that represent the starting value of zero or two or four or six.

When the assistant receives the detective card she notices the starting value based on the spectator's thumb position. If the card is face up she also adds the value of 8 to that starting total.

If the pointer is toward herself she adds one.
She now knows the value of the target card.

To know the suit of the target card she advances one notch SHoCkeD fashion from the detective card.

A lot of thought went in to this little system.
Background:
When using binary codes it takes six signals to identify a playing card.
Five binary signals provide 32 possibilities which is not enough to cover 52 cards whereas six binary signals provides 64 possibilities which is more than enough to cover 52 cards.

My first thought was that the magician and assistant would both have a theoretically memorized stack and that six binary signals would indicate the theoretical position of the selected card. But this was asking a bit much for my assistant who is not about to memorize a stack (I have the following stacks memorized: Aronson, Mnemonica, Tritium, currently I use the Tritium stack the most but my wife assistant, no way).

I needed to come up with a simpler more direct system for the six binary signals. With only one card (which needs to be a pointer card) there are only two natural binary signals available (card face up versus face down is one binary signal. One way pointer versus the other way is the second binary signal).

The six binary signals:
The VALUE of the target card requires four binary signals. The SUIT of the target card requires two binary signals. This is a big challenge when there are only two natural binary signals available using one (pointer) card.

I thought the SUIT of the target card could be encoded by utilizing the suit of the signal card. This required that all four of the same value pointer cards be available. The only playing card that is a pointer on all four suits is the sevens. Therefore the four sevens must be the pool of "detectives" for the magician to use.

If using standard USPCC playing cards (even from a borrowed deck) then the four Jacks or the four Queens or the four Kings could provide the pool of four detective cards because they are all one way pointers (see my discovery in the link below):

https://themagiccafe.com/forums/viewtopi......forum=37

But back to how I came up with the current zero, 2, 4, 6 method.
My first attempt was to have the thumb indicate 1,2,4,8 but I soon realized that that was no good (because several combinations are often needed but the thumb can only indicate one) so I then tried 0,1,2,3 and that worked (because the 3 combines the 1 and 2 when needed) but that ended up with the "pointer near" to indicate "4". The bad part of doing it that way is that if the pointer is misinterpreted by the assistant then they are considerably wrong with their answer (especially if using the hard to read one way pointer face cards).

So using constraint design theory I wanted the following design criteria:
1. Minimal impact if the one way pointer is interpreted incorrectly by the assistant.
2. Easy for the assistant to interpret all the signals.
3. Encode the target card value.
4. Encode the target card suit.
The above is the priority that I used to come up with the thumb at 0,2,4,6 method.

A key to this method is that signaling a 6 is combining the two binary signals for the 2 and the 4.

Note that having the pointer "near / far" indicator tell the assistant whether to add 1 or not:
This means she may say for example "The six of clubs" when the target was the seven of clubs or vice versa but either way she still looks pretty good to the audience being off by only one notch.