If we assume a 2D algebra, one possible algebra that can be represented is complex numbers, but are there any other types of algebra that are valid groups?
let the 2 dimensions be e1 & e2.
In order for the tables to obey the rules of a group (identity element, one element in each row & column, etc)
e1  e2  
e1  e1  e2 
e2  e2  e1 
The table only represents the type of each entry  not yet whether there is any sign reversal.
There are only two valid possibilities, for a valid group, e2^{2} = +e1 equivalent to dual and e2^{2}= e1 equivalent to complex numbers.
There are a number of questions that occur to me about this, for instance:
 Can we calculate the sign of the other terms from the sign of the terms on the leading diagonal (from the squares of each dimension).
 How independent are the signs of the other terms?
So if we start from the squares of the dimensions, we either have the dual
 e1^{2} = +e1
 e2^{2} = +e1
or the complex numbers.
 e1^{2} = +e1
 e2^{2}= e1
I cant think of a way to derive the other terms from these terms?
If we include e1 * e2 = e2 as a given fact, can we derive the last entry: e2 * e1 = e2 ?
We can try this for both cases, as follows:
since: e2 = e1*e2
square both sides:
e2^{2} = e1*e2*e1*e2
but also we can put one or more e1* in front
e2^{2} = e1*e1*e2^{2}
therefore this gives e1*e2 = e2*e1
This is true regardless of whether e2^{2} = +e1 or e1 so it applies to all cases:
dual  complex  


So we can derive the right identity operator from the left identity operator, but this seems to be the only thing that is connected. The other results seem to be independent of each other?