[seqfan] Re: seq needed of (0,+-1) matrices

[N. J. A. Sloane]

Robert Israel found the explanation!
I was using 

star:=A[w,x]*A[y,z]+A[w,y]*A[x,z]+A[w,z]*A[x,y]-A[w,x]*A[w,y]*A[w,z]*A[x,y]*A[x,z]*A[y,z];

and got a(4) = 125, agreeing with Max A.,

but when I corrected this to 

star:=A[w,x]*A[y,z]+A[w,y]*A[z,x]+A[w,z]*A[x,y]-A[w,x]*A[w,y]*A[w,z]*A[x,y]*A[x,z]*A[y,z];

I get a(4) = 293, agreeing with rhh.

So it does look like we get two sequences!

The second version is the one in Wesp's paper (although
he does not give any numerical values)

Robert, thanks for straightening this out!

I will add the other version - it will be A204821 (and 
A204809 is the other version)

Neil