Quote Originally Posted by sindri View Post
You are right, the ENTP notation is arbitrary, but it is the standard established by Reinin himself in his mathematics and is standard practice for all socionics. This is already complicated enough, so let's go with it for now, but, you can redefine the system to have any type as the base type, meaning each corresponding dichotomy trait for that type is now the positive one.

What can it represent? Pretty much any system in socionics because everything was made with the same logic in mind. Specifically, the pyramid represent the abelian group Z2^4 (Z2 x Z2 x Z2 x Z2). But I don't like talking about the math because it is unnecessary and it tends to turn people off to the actual idea.

The image in the original post is a space showing all possible types, but a single type can be represented by defining the charge for each point (check out the post I just did for QuickTwist to see an example). For each type, all the small groups will have either 1 or 3 positive traits. They will look like the fano planes for the eight information elements and functions in the paper, except colored and 3d.

Thanks for the compliment, I put a lot of work into the visuals so it would make it easier for people to get into such an intense subject ^u^
Ok, cool, I wasn't criticizing with the ENTP being arbitrary btw, just wanted to understand the representation without any confusion on my part.

I have some more question and comments. I've always seen information preference typing as a matrix of traits(whatever) for which the solution is the type. This can apply to a empirical model as well not just socionics. Most big data marketing categorize people by some kind of single value decomposition which solves matrixs for any number of traits.
From what I understand a ableian group is just a special kind of matrix. This is essentially the same sort of technique you're proposing some sort of matrix solving technique.

There's a few directions this can be taken which can,

A. Guess at a type based on observed traits
B. Check type for accuracy
C. Derive additional traits

There's a few ways to do this.

Solve the Matrix to guess the type, then derive reinin from type and check this against traits

So Traits -> Type -> Reinin -> Parity check, I think this is close to your methodology?

This is pretty close to the way I see this occurring.

However, with enough data you can do something else too, which is find clusters of traits that arise in the data that we would need for the above method because you wouldn't really know what traits are applied to reinin.

So Traits -> Type -> Find traits clusters(should match Reinin)

The way I look at a socionics is that it presents a plausible solution to information precedence so this would be solution of some big data analysis where you would have type trait clusters and reinin trait clusters.

FWIW, you would not need to know what type these individual are or what even the data represents initially, you could do that by manually looking at the traits and associating with the sociotype.

Any big data analysis should produce some sort of type and sub clusters of dichotomies that matches Model A and Reinin Dichotomies. I think if a system can be created to analyze traits and create something that's close to socionics model(maybe not possible to be perfect because measurable traits are going to be likely behavior or self reporting) and then empirical out of this system can be analyzed for problems(which it will have) and inaccuracies.

I don't have a deep knowledge of the maths involved but you might so you might help me correct any understanding/misunderstanding I might have.