Factors Behind Common Mistypings

[Krig the Viking]

I've been observing for a while which kinds of mistypings occur most frequently on this board, as well as what sorts of mistakes I tend to make myself when typing people. I've noticed some common factors that seem to occur frequently in mistypings, for example, types with shared Ego functions are commonly mistaken for one another, as are types with similar function strengths (Quasi-Identicals, etc.). I've compiled a list of the common factors that I've noticed, and given each item points, which reflect my opinion of how frequently such factors are present in mistypings.

1. Shared Ego Function. (4 pts.)

This is the most common factor that I've seen in mistypings. This would mean four points for Kindred and Business types, eight for Mirror.

2. Identical Ego/Id Function Strength. (3 pts.)

This is the second most common factor I've seen. This whole thing started when I noticed how common Benefit types are in mistypings, which I hadn't expected. Three points for Benefit types, six for Quasi-Identical.

3. Close Ego/Id Function Strength. (2 pts.)

Two points for Mirage and Semi-Duality, four for Extinguishment.

4. Shared Vertness. (2 pts.)

Two Introverts are obviously easier to mistake for one another, as are two Extraverts.

5. Shared Rationality. (1 pt.)

This seems to have a slight effect.

6. Shared Ego/Superid Functions. (1 pt.)

I've seen people mistake valued functions for strong functions. It doesn't seem logical that it would happen, but it does.

Adding up the points for each inter-type relation, we get the following theoretical list of common mistypings, from most common to least common:

Points	Name	Frequency
8	Mirror	Very Common
8	Quasi-Identical	Very Common
7	Kindred	Quite Common
7	Look-a-Like	Quite Common
6	Benefit	Common
5	Extinguishment	Kind of Common
4	Activity	Not Very Common
4	Supervision	Not Very Common
4	Semi-Duality	Not Very Common
4	Mirage	Not Very Common
3	Duality	Uncommon
3	Super-Ego	Uncommon
0	Conflict	Extremely Uncommon

Obviously, this is all still highly experimental, so I'll need some more hard data to see if it actually works. But in theory, if you have a range of possible typings that have been suggested for someone, you should be able to figure out which ones are more and less likely based on the above list. For example, if your range of suggested typings includes a Conflict pair, then it's likely that neither one of the Conflict pair is the subject's actual type, since if either one is true, then it means that someone has mistaken that type for its Conflictor, which almost never happens. The suggested types whose relations to the other suggested types are mostly the common mistypings and rarely the uncommon ones are more likely to be the correct typing.

For example, let's say people have typed a given subject as ILE, EIE, SLI, and LII. The relationships involved are:

ILE: Benefit, Semi-Duality, Mirror.
EIE: Benefit, Conflict, Semi-Duality.
SLI: Semi-Duality, Conflict, Benefit.
LII: Mirror, Semi-Duality, Benefit.

In this case, since EIE and SLI conflict with one another, the subject is unlikely to be either. The subject is likely ILE or LII.



Another example: IEI, ESI, ILE, SEI.

IEI: Benefit, Mirage, Look-a-Like.
ESI: Benefit, Conflict, Quasi-Identical.
ILE: Mirage, Conflict, Duality.
SEI: Look-a-Like, Quasi-Identical, Duality.

Since ESI and ILE conflict, neither is likely correct. SEI is Dual with ILE, which is an Uncommon mistyping, so it's more likely the subject is IEI (with Quite Common, Common, and Not Very Common mistypings).



Anyway. It's somewhat complicated, but it makes sense to me. Like I said, it's still pretty theoretical at this point, so it needs to be applied to some actual test cases, to see how it works and if it needs refining, etc. But I'm optimistic that this could be a helpful tool in narrowing down potential types in some cases.