site algorithms

[aestrivex]

actually, i didn't realize quite the way that this site was going to work. i think the design is ultimately very interesting -- it could potentially, given a varied body of respondents, be very interesting to see what this site concludes about people's types and the relations therein based on typings of others.

at the same time, obviously, as ryan knows, i think the questions for the test and the intertype relations component are horrible.

[RSV3]

actually, i didn't realize quite the way that this site was going to work. i think the design is ultimately very interesting -- it could potentially, given a varied body of respondents, be very interesting to see what this site concludes about people's types and the relations therein based on typings of others.

at the same time, obviously, as ryan knows, i think the questions for the test and the intertype relations component are horrible.

Welcome aestrivex! I am trying to make some minor changes to the test. With respect to the intertype relationship component, that is still in the early stages and will probably evolve considerably before any results appear. As always, any suggestions for improvements are appreciated.

[aestrivex]

one suggestion and complaint that i have is this site's reliance on this number system of subtypes (ie LSE-3 or whatever). basically in my opinion accepting/producing subtypes being bad enough, this system is essentially an arbitrary one that you made up. while i recognize that you can and do crunch the numbers using these subtypes in strange ways (which, obviously, in my opinion, is inherently a silly approach), the resulting % compatibility numbers based on them are just silly; if you must use them, i would advise getting rid of this % compatibility thing and use it as behind-the-scenes aggregation.

though naturally why would you listen to me anyway; i think most of the point of this website (solely excepting the algorithm to determine one's likely type based on having people type others) is a mistake to begin with.

[RSV3]

one suggestion and complaint that i have is this site's reliance on this number system of subtypes (ie LSE-3 or whatever). basically in my opinion accepting/producing subtypes being bad enough, this system is essentially an arbitrary one that you made up. while i recognize that you can and do crunch the numbers using these subtypes in strange ways (which, obviously, in my opinion, is inherently a silly approach), the resulting % compatibility numbers based on them are just silly; if you must use them, i would advise getting rid of this % compatibility thing and use it as behind-the-scenes aggregation.

though naturally why would you listen to me anyway; i think most of the point of this website (solely excepting the algorithm to determine one's likely type based on having people type others) is a mistake to begin with.

The subtype system is essentially the accepting/producing system; I've just added the preceding number to represent the magnitude or strength of that subtype (0 being no subtype, 1 being minimal, 2 being moderate, and 3 being strong). To visualize this graphically, a person's subtype represents their position along his/her particular temperament ring. E.g., an LII-3Ti would be on the far side of the alpha quadrant in close proximity to beta (and farther away from delta); an LII-0 would be of centered in the alpha quadrant.

Although I'm not getting rid of the subtype system for a number of reasons, I do agree with your suggestion that it is unnecessary to display the percentage of compatibility, so I will make that change when I get some time.

[aestrivex]

The subtype system is essentially the accepting/producing system

you carry the assumption that the accepting/producing system is a continuous spectrum (or at least divisible collection) of emphasis, rather than a discrete grouping. there is nothing to support this assumption in the classical literature on accepting/producing subtypes (that i have ever seen, anyway) -- rather, every single description or bit of classical material that i've seen on the matter of accepting/producing subtypes has treated them as simply descriptions of accepting and producing categories, and nothing more.

indeed, there are people who have done work with accepting/producing subtypes conceptualized very much as a discrete grouping -- anything the ashtonian people do, for instance, holds a very different and very discrete concept of subtypes.


all of this is not to suggest that your system is inherently wrong for assuming this type of spectrum, or that i agree with the ideas dealing with discrete subtype groupings (obviously, with regard to the ashtonian system i think it's completely full of shit and am not so keen on any other system of subtypes either).

but, the point is, there is nothing to suggest that the assumptions you're working with are "essentially" classical accepting/producing subtypes; that is merely nonsense.

[RSV3]

The subtype system is essentially the accepting/producing system

you carry the assumption that the accepting/producing system is a continuous spectrum (or at least divisible collection) of emphasis, rather than a discrete grouping. there is nothing to support this assumption in the classical literature on accepting/producing subtypes (that i have ever seen, anyway) -- rather, every single description or bit of classical material that i've seen on the matter of accepting/producing subtypes has treated them as simply descriptions of accepting and producing categories, and nothing more.

indeed, there are people who have done work with accepting/producing subtypes conceptualized very much as a discrete grouping -- anything the ashtonian people do, for instance, holds a very different and very discrete concept of subtypes.


all of this is not to suggest that your system is inherently wrong for assuming this type of spectrum, or that i agree with the ideas dealing with discrete subtype groupings (obviously, with regard to the ashtonian system i think it's completely full of shit and am not so keen on any other system of subtypes either).

but, the point is, there is nothing to suggest that the assumptions you're working with are "essentially" classical accepting/producing subtypes; that is merely nonsense.

I basically agree with everything you've said here, and my previous wording was a little misleading. My system is now substantially different from the classical subtype system, although I did start with it. Imo, a continuous subtype system fits better with Model A for a number of reasons including using it to derive relative functional strength and more specific intertype compatibility.

[aestrivex]

suggestion: i would like to see a category of something like "conjugate type ignoring test results." personally, i would be interested in seeing how the resultant data compare to my understanding with the addition of this category of responses.

[RSV3]

suggestion: i would like to see a category of something like "conjugate type ignoring test results." personally, i would be interested in seeing how the resultant data compare to my understanding with the addition of this category of responses.

I'm probably not going to implement this at this time, although I will say that most of the individual component data you need to compute this manually is already provided, assuming you come up with your own weight for each component.

[aestrivex]

I'm probably not going to implement this at this time, although I will say that most of the individual component data you need to compute this manually is already provided, assuming you come up with your own weight for each component.

well, duh. it's also the case that all we need to do is make our own website with the same numbers and function and that would also provide the desired result. it's also a possibility that we could meet and type all the individuals independently and ignore your information as worthless, which i do already. finally, it's a possibility that we could all retire and marry sanitation workers and then murder them by hitting them repeatedly on the head with violins.

[RSV3]

I'm probably not going to implement this at this time, although I will say that most of the individual component data you need to compute this manually is already provided, assuming you come up with your own weight for each component.

well, duh. it's also the case that all we need to do is make our own website with the same numbers and function and that would also provide the desired result. it's also a possibility that we could meet and type all the individuals independently and ignore your information as worthless, which i do already. finally, it's a possibility that we could all retire and marry sanitation workers and then murder them by hitting them repeatedly on the head with violins.

Your caustic remarks add very little to the purpose of this forum, so if you're going to continue posting, I'd prefer you stick with more constructive commentary.