This journal is written because I need to explain something that is palpably unjustifiable: you would submit your picture Y, which is "clearly" of higher quality than a certain picture X in this group's gallery, yet I declined X. I wish to convince you that this is not due to my deliberate subversion.
There are two reasons for this unfortunate phenomenon. The first reason is that judgment in art quality is different for everyone; the second is that my judgment is different everyday. Both of these variations are not very big, but these variations can make a difference when the picture under question is near the boundary of acceptance/declination.
I'll explain the effects of this variation in the case of my own judgment. The effects are similar in the other case.
The most appropriate way to model my judgment is to think of me as a program that takes in a picture and outputs a number based on its sadness, artistic quality, and a bunch of other factors. If the output number is greater than 0, then that art gets included, else it doesn't.
Suppose now we have three pictures A, B, C, D.
A is of amazing quality, D is of very bad quality, and B, C are of boundary quality. Now, suppose I look at these four pictures once a day, record down my scores for these pictures, then wipe my memory about these pictures (so that each day's judgment is independent). After doing this for many days, I list all the data I wrote down, and what I would see is this:
A: 100, 103, 97, 80, 102, 110, ... 100.
B: -1, 3, 0, -3, 5, 3, ... 1.
C: 2, -2, 0, -3, -1, ... , 1.
D: -10, -11, -8, -12, ... -5.
I computed the average scores. A, 100; B, 1; C, -1; D, -10. Thus, it would be reasonable to say that I should always include A and B, and always exclude C and D.
However, there are daily variations in these scores. On certain days, I may be more touched by sad art and give them a higher score, or may be too tired to judge with sufficient attention and give them a lower score, and the variations in my judgment can be large: I have judged A to be 80 on one day, and C to be -5 on another day.
However my judgments may vary, since A has such a high score, the variations in my judgment can't make a dent, and A would get accepted any day. Similarly, C would not be accepted any day. BUT, for B and C, things are much less predictable. It's quite possible that I would accept C and exclude B, had they come at the certain days when I would score B lower than 0, and score C higher than 0.
Thus, as a general rule, whether I accept art that is of boundary quality is highly random, and the randomness increases as the quality nears the boundary.
Similarly, there are some art that are so good that everypony would agree that they deserve to be included, and some are so bad that none would. And there are the boundary cases where personal variations matter a lot. It's conceivable that, given two boundary cases X and Y, you would consider X to score 2 and Y to score -2, while I consider X to score -1 and Y to score 4.
What does all of these mean?
If you submit a piece of art that is of quality that is near the boundary, then whether I accept it is quite random. Thus, this unfortunate phenomenon appears: sometimes I would decline a piece of art that has a higher average score than another piece of art that I have accepted, simply because of random effects (in less amicable words, “my whims and fancies”).
To those who are upset that I declined your art when some lesser art exist in this gallery... I owe you an apology, even though I'm not deliberately guilty. The most likely thing that happened is that your art falls near the boundary of quality, thus whether it gets included becomes highly unpredictable, subject to random chance. Maybe if you had used a different color scheme, or drew some other pony, or submitted it on a rainy day, or… I would have included it.
Above all, please don't be upset... I try my best to make this group useful and an interesting place to visit... I try my best, but I can't always do my best, and I can't prevent the inherent randomness from causing me to accept some art but decline some other equally good art.