nand

I think what you will find is that 0= can be made to equal to logical not as long as anything which is nonzero is truthy.

Since 0= can mean logical not, that means that nand can be performed by first performing an and, and then performing a 0=. And so instead of having a nand operation, you can have an and operation and a 0= opetation.

Switching from and to nand may save you a very small number of bytes depending on how you have implemented nand.

I hope that helps.