The first problem is that the rules are not clearly defined. For example, things like Judge A gives -2, Judge B gives +2 against 3Lz-3F. It seems that interpretation of rules differs for each judge.
This is entirely within the rules for awarding GOEs. I think what you're seeing in an example like this is less a case of judges having different interpretations of the
rules and more of judges having different interpretations of
that element as performed.
In sports, it is an absolute requirement that rules be clear for athletes. We can not compete fairly if the measure of competition is not clear.
Skating is not a sport with discrete outcomes, such as did the ball go into the goal or not or go out of bounds or not, or even was the jump landed or not. The question isn't "Did it or didn't it?" but rather "How well?" Those kinds of decisions call for judgment of each case on its own merits,
according to guidelines/rules, but not simply mechanical application of pre-written rules.
In qualitative sports, there are always going to be value judgments made by individuals, with more room for individual variation in than in sports that rely solely on yes/no determinations. Skaters know this. Long-time skating fans know this.
Many successful elements have some positive qualities and some negative qualities. It's up to each judge to balance out the good points and bad points to come up with an overall evaluation of the element, including whether in their judgment the final score should be higher or lower (or equal to) the base value.
There are guidelines as to what judges are supposed to reward and what they're supposed to penalize, but there's no way to quantify exactly when every possible example of the element in question would qualify for a positive bullet point or when every weakness would merit a reduction, nor where to draw the line between -1 or -2 or -3 for weaknesses/errors that the rules. Every element is slightly different from the next, and different individuals will have different perceptions of how close that example from that skater on that day came to qualifying for each separate positive or negative bullet point.
And the rules explicitly instruct judges to award positive features before subtracting for errors. Now there are more explicit caps on how high they can start on the positive side for elements that don't meet all of three specified mandatory criteria or that have major errors.
With elements that have no significant errors and few positive qualities specified by the rules, we would expect GOE scores close to 0. As of last year, such an element would often have a GOE range of -1 to +1 within the -3/+3 gamut. With the new -5/+5 scoring that allows more distinctions between fine points, a range of -2 to +2 is pretty much equivalent to -1 to +1 last year. You could think of it as including -0.5 and +0.5 as additional options in last year's range and then just renaming the options.
With the additional scoring options, with 11 possible scores instead of only 7 to choose from, you are going to see more variety of GOEs from the different judges. Where before a range of 1 GOE step on either side of the median was the norm, now we're going to see more examples of 2 steps on one side or the other or both. This isn't a bug in the way the rules are written or applied -- it's an intentional feature, inherent in the nature of qualitative evaluation and in the finer increments of the revised scoring.
In the 3Lz+3F example, you're talking about an element type -- a true jump combination with a triple jump landing on the opposite foot followed by a triple flip -- that has never been performed before. There is no established standard as to what constitutes average quality on a triple something-triple flip combination. So it's small wonder that different judges had different opinions in this case.
The second problem is that figure skating judges are not professional. They are usually doctors or coaches and they are engaged in other work. I have doubts as to whether they study the rules and are familiar with it.
Coaches are skating professionals and it is their job to know the rules that their students compete under. However, skating coaches are not allowed to be judges. They may be technical specialists.
Yes, judges generally have other full-time jobs or other sources of income because judging is not a paid activity. Some have volunteered their time to judging most weeks of the year for many years and know the rules backward and forward. Others only judge a couple of competitions per year -- which may have some different specific rules between domestic vs. international, junior vs. senior, men vs. ladies, etc. -- and don't spend much time in their daily life studying skating rules when they don't have a competition that week.
Would it be possible to professionalize judges in some way so that some or all international judges are able to make it their full-time job? I don't know.
Take a look at the seasonal skating calendar, how many competitions are held each week at each time of year. Can you suggest a structure by which all of the competitions in the busy season could be covered by professional judges and what kind of commitments would be expected of those judges in the other times of year? What might be the career path to becoming a professional judge? What levels of competition would still be covered by volunteer judges, and would any types of competitions be able to use a mix of professionals and volunteers?