While its not the answer you are looking for, the reality is that it takes experience and time to learn accurate identification and grading. Back in the day when a lot of people went to shows, identification and grading was built from handling and discussing a lot of marbles. Hands-on, in front of people who really knew. Handling large number and variation of marbles in a single room, times 25-80 rooms. It wasn't about guessing. You could rotate it in hand and see every view, get used to seeing varying degrees of wear. There were a few Gold Standards of identification and grading whose name and face-to-face reputation were on the line.
As fewer people came to shows and more came to rely on the Internet, identification became an increasing game of comparing your unknown marble of unknown condition to Web photos, many or most of which looked "kinda similar" and could be concluded to maybe the same. Learning accurate grading on Internet photos can be likened to rolling dice - with a fair helping of wishful optimism flavoring it. The concept of "as-mades" was invented. Facebook groups with contributions of questionable accuracy arose. It seems that the term "near mint" has come to now commonly cover a multitude of sins. The concept of "same run" machine mades was invented, seemingly to infer rareness and higher value, disconnected to the reality of machine made mass production over months or years.
Learning accurate identification and grading off the Internet has considerable challenges. Variances in one singular type, color combination and design of marble have a sliding scale of popularity.... that popularity changing over time. Small variations on common styles can mean one marble will sell for $50 and it's sister will sit unsold for $20. What was "cool" 5 years ago falls in popularity and price in time.
I'm always impressed that when you go to a larger show, a few hours in the rooms and you get a feel for which way the wind is blowing for what people are looking for and roughly where values may be going for those types. Values and demand/interest rise and fall. Putting values in print would mean that they would be out of date soon after publishing. And price is a function of all of these, plus what the seller has into it and how interested (or not) they are in selling/making a killing.
I suggest focusing on one or two manufacturers and develop experience with them. Learn to grade accurately and dispassionately. Try to not get sucked into wishful ID and grading. An appreciation of value in a changing market will follow. Then you can confidently expand what you are interested in collecting.
Hope this helps!