The questions answered are:

1. How many apparel qualities and conditions are there and how are they ranked?
   
   
2. How many quality and condition combinations are possible for a single apparel item?
   
   
3. How are apparel descriptions written?
   
   
4. What apparel descriptors are affected by quality and condition?
   
   
5. How are apparel item quality and condition variants ranked by defense score?
   
   
6. Is apparel quality, condition, or neither a better determinant of apparel defense score?
   
   
7. How are apparel item quality and condition variants ranked by purchasing and selling price?
   
   
8. Is apparel quality, condition, or neither a better determinant of apparel purchasing and selling price?

The short answers to all of these questions can be found in the Conclusion section.

The long answers to all of these questions can be found under the Results section.

I gathered information on apparel item qualities and conditions primarily through excel and a cheat engine folder. My investigation followed the chronology of the questions:

1. For the first three questions:
   
   
   • I altered the qualities and conditions of different apparel items, all within the game-code value range (0-255). I found that there were distinct number intervals which denoted each quality and condition level for every apparel item. After recording the in-game names of those levels, I organized them into a chart. Some more tinkering later, I noticed how qualities and conditions in apparel item descriptions had no repeating adjectives.
   
   
2. For the remaining five questions:
   
   
   • I used 363 quality and condition variants of one apparel item as a sample I generalized for all 65,536 variants of every apparel item. It is only meant to be an accurate estimate, not a perfectly precise predicting tool.
     
     Hence why, all of the excel charts and regression models below this section use 121 lower end, 121 middle, and 121 upper end value interval variants of one apparel item: internal name PltArmour SetB. Here are visual representations of the model, ordered from least, to medium, to greatest defense score (i.e. rank followed by quality and condition):

⠀
⠀
⠀
⠀

Summary
In this chart, 15 quality and 19 condition names are arranged into 11 value intervals within a 256 total byte range. Each row represents a level of quality and condition in terms of game-code value intervals in bytes. The rows are organized top to bottom in order of decreasing apparel defense score — sum of coverage, impact, slash, crush, and pierce.


Further Explanation
As an example of equivalency, almost perfect and extremely good do appear in different items but still correspond to the same value interval of 15 - 31.

As an example of difference, "an exceptionally well made [Apparel Name] in flawless condition" is a variant of that apparel item with possibly the highest defense score. Likewise, "an extremely crude [Apparel Name] in extremely worn/terrible condition" is a variant of that apparel item with possibly the lowest defense score.

Further on differences, "an exceptionally well made [Apparel Name] in flawless condition" with a 255 byte quality value and 0 byte wear value is greater in defense score than "an exceptionally well made [Apparel Name] in flawless condition" with a 248 byte quality value and 7 byte wear value.



In terms of game code values, there are 256 byte values for apparel quality and condition which means 256*256 or 65,536 quality and condition game code value combinations are possible.

In terms of value intervals, there are 11 levels for quality and condition which means 11*11 or 121 quality and condition value interval combinations are possible.

In terms of apparel descriptions, there are 15 name slots for quality and 11 name slots for condition which means 15*11 or 165 apparel descriptions are possible.



After observing the chart, notice how condition "terrible" could either be equal to condition "extremely worn" or "very worn." One possible way to know which level "terrible" corresponds to is through the quality. In a single apparel item, when the quality has the adjective "extremely" in it, the condition "terrible" is equivalent to "extremely worn" and for all qualities with the adjective "very" in it, the condition "terrible" is equivalent to "very worn."

Moreover, it seems as though adjectives in quality and condition names do not repeat in the same apparel description. The adjectives which could repeat but are usually never repeated are, "extremely," "very," "fairly," and "somewhat." For example, a level 5 quality "fairly well made" paired with a level 8 condition "fairly worn/relatively poor" will most likely have the apparel description "A fairly well made [Apparel Name] in relatively poor condition."

Regarding qualities, the qualities coarse, coarsely made, and somewhat crude denote the same level of quality but switch places on each viewing of an apparel item description text. The same is true of qualities very coarse, crude, and crudely made.



Only "coverage," "impact," "slash," "crush," "pierce," and "encumbrance" are affected by quality and condition. The descriptor "points" remains unaffected by quality and condition.

I suspect an apparel item's tier has the greatest affect on points.

⠀
⠀
⠀

Key
Quality: 11 units in total. Each unit corresponds to a quality level:

• 1 (extremely crude)
  
• 2 (very crude)
  
• 3 (very coarse/crude/crudely made)
  
• 4 (coarse/coarsely made/somewhat crude)
  
• 5 (fairly coarse)
  
• 6 (mediocre)
  
• 7 (fairly well made)
  
• 8 (well made)
  
• 9 (very well made)
  
• 10 (extremely well made)
  
• 11 (exceptionally well made)

Condition: 11 units in total. Each unit corresponds to a quality level:

• 1 (extremely worn/terrible)
  
• 2 (very worn/terrible)
  
• 3 (worn/poor)
  
• 4 (fairly worn/relatively poor)
  
• 5 (somewhat worn/below average)
  
• 6 (fair)
  
• 7 (fairly good/above average)
  
• 8 (good)
  
• 9 (great/very good)
  
• 10 (almost perfect/extremely good)
  
• 11 (flawless)

Defense score: Sum of coverage, impact, slash, crush, and pierce of each apparel item varying by quality and condition.



Short Answer
Quality, as compared to condition, is a better determinant of apparel defense score.

Long Answer
As shown in the regression models:

The measurement of strength of the linear relationship between independent and dependent variables (i.e. Multiple R) is greater in the case of quality and defense score than condition and defense score: 0.6953 > 0.6783. Quality is more linearly related to defense score than condition.

The proportion of the variance in the dependent variable explained by the independent variable (i.e. R^2) is greater in the case of quality and defense score than condition and defense score: 0.4835 > 0.4601. Quality explains a larger proportion of defense score than condition.

The value indicating the overall significance of the model as compared to a model without the independent variables (i.e. Significance of F) is further from 0.01 in the case of quality and defense score than condition and defense score: 9.9097E-54 < 2.9860E-50. Quality more accurately fits defense score data than condition.

The value indicating the increase in the dependent variable for every unit increase in the independent variable (i.e. Coefficient) is greater in the case of quality and defense score than condition and defense score: 0.6981 > 0.6810. Every unit increase in quality increase the defense score more than every unit increase in condition.

The value indicating the significance of a coefficient (Coefficient P-Value) is further from 0.01 in the case of quality and defense score than condition and defense score: 9.9097E-54 < 2.9860E-50. Every unit increase in quality more accurately fits defense score data than every unit increase in condition.

⠀
⠀
⠀

Key
Quality: 11 units in total. Each unit corresponds to a quality level:

• 1 (extremely crude)
  
• 2 (very crude)
  
• 3 (very coarse/crude/crudely made)
  
• 4 (coarse/coarsely made/somewhat crude)
  
• 5 (fairly coarse)
  
• 6 (mediocre)
  
• 7 (fairly well made)
  
• 8 (well made)
  
• 9 (very well made)
  
• 10 (extremely well made)
  
• 11 (exceptionally well made)

Condition: 11 units in total. Each unit corresponds to a quality level:

• 1 (extremely worn/terrible)
  
• 2 (very worn/terrible)
  
• 3 (worn/poor)
  
• 4 (fairly worn/relatively poor)
  
• 5 (somewhat worn/below average)
  
• 6 (fair)
  
• 7 (fairly good/above average)
  
• 8 (good)
  
• 9 (great/very good)
  
• 10 (almost perfect/extremely good)
  
• 11 (flawless)

Purchasing and Selling Price: sum of (12^2)Gold + 12*Silver + Copper varying by quality and condition. (1 gold coin = 12 silver coins & 1 silver coin = 12 copper coins)



Short Answer
Quality, as compared to condition, is a better determinant of both purchasing price and selling price.

Long Answer
As shown in the regression models:

The measurement of strength of the linear relationship between independent and dependent variables (i.e. Multiple R) is greater in the case of quality and purchasing price than condition and purchasing price: 0.6471 > 0.6229. The same is true of selling price: 0.6470 > 0.6229. Quality is more linearly related to both purchasing price and selling price than condition.

The proportion of the variance in the dependent variable explained by the independent variable (i.e. R^2) is greater in the case of quality and purchasing price than condition and purchasing price: 0.4187 > 0.3880. The same is true of selling price: 0.4186 > 0.3880. Quality explains a larger proportion of both purchasing price and selling price than condition.

The value indicating the overall significance of the model as compared to a model without the independent variables (i.e. Significance of F) is further from 0.01 in the case of quality and purchasing price than condition and purchasing price: 1.9162E-44 < 2.1956E-40. The same is true of selling price: 1.9635E-44 < 2.1475E-40. Quality more accurately fits purchasing and selling price data than condition.

The value indicating the increase in the dependent variable for every unit increase in the independent variable (i.e. Coefficient) is greater in the case of quality and purchasing price than condition and purchasing price: 284.1083 > 273.4719. The same is true of selling price: 71.0317 > 68.3854. Every unit increase in quality increases both the purchasing price and selling price more than every unit increase in condition.

The value indicating the significance of a coefficient (Coefficient P-Value) is further from 0.01 in the case of quality and purchasing price than condition and purchasing price: 1.9162E-44 < 2.1956E-40. The same is true of selling price: 1.9635E-44 < 2.1475E-40. Every unit increase in quality more accurately fits purchasing and selling price data than every unit increase in condition.

The questions and answers summarized:

1. How many apparel qualities and conditions are there and how are they ranked?
   
   
   • There are 15 qualities and 19 conditions and they are sorted in 11 game code value intervals within a total 256 byte range.
   
   
2. How many quality and condition combinations are possible for a single apparel item?
   
   
   • 65,536 game code value combinations, 165 apparel descriptions, and 121 value interval combinations.
   
   
3. How are apparel descriptions written?
   
   
   • All 165 apparel descriptions are written without repeating adjectives. The adjectives which can repeat but usually never do are "extremely," "very," "fairly," and "somewhat."
   
   
   
4. What apparel descriptors are affected by quality and condition?
   
   
   • Only descriptors "coverage," "impact," "slash," "crush," "pierce," and "encumbrance" are affected by quality and condition. The descriptor "points" remains unaffected by quality and condition.
   
   
   
5. How are apparel item quality and condition variants ranked by defense score?
   
   
   • See section "Defense Score" for answer.
   
   
   
6. Is quality, condition, or neither a better determinant of apparel defense score?
   
   
   • Quality, as compared to condition, is a better determinant of apparel defense score.
   
   
   
7. How are apparel item quality and condition variants ranked by purchasing and selling price?
   
   
   • See section "Arena Shop Price" for answer.
   
   
   
8. Is quality, condition, or neither a better determinant of apparel purchasing and selling price?
   
   
   • Quality, as compared to condition, is a better determinant of both purchasing price and selling price.

Closing Commentary
Since I am only comparing defense scores and prices of one apparel item's quality and condition variants, my findings can tell you what quality and condition variant of an apparel item is highest in defense score and price.

For example, if you are choosing between two different items, like different plate armour models, you could know if their defense score rank is equal within their own quality and condition variants (e.g. 1/363 & 1/363). If it is and one has a higher defense score than the other, know that the former is likely higher due to an apparel descriptor other than quality and condition.

https://gtm.steamproxy.vip/sharedfiles/filedetails/?id=2711813328

I will be updating this guide. So do check in often!

I will post apparel quality and condition guides for subsequent versions of Exanima.