I know cat divisions is said to be next stop on the ZRS development roadmap. Just my observations below, adding an aspect that is rarely mentioned - finish times.
The nice thing about results-based categorization is that you don’t really have to look exclusively at signup numbers anymore. You could also start glancing at finish times.
The cat divisions are not perfomance-based anymore and, besides, winner finish times can depend on race tactics rather than all-out efforts. But still, the spread of finish times is kind of interesting. It does say something, however fuzzy that something is.
So let’s say the cat A winner finishes at 36:00, the B winner at 38:00, C winner at 40:00 and E at 42:00, then in a sense that would be a nice and intuitive spread. Of course every cat should have a fair number of participants but it isn’t really crucial for them to be exactly evenly balanced because there seems to be another type of balance in there too.
But when looking at e.g. the current and rather well-attended Fresh Outta 24 races, another pattern emerges. The top finish times of A, B and C are often quite close to each other. So much so, in fact, that one of these categories can seem redundant, not only because of finish times but also because there are usually few participants in one of the three categories, usually either A or B.
To make matters worse, the cat D winner usually doesn’t finish all that far from the cat C winner, although the gap can be a little wider than between B and C. But then the time gap from top D down to the E winner is often huge. In addition, cat E is often small in numbers.
I like v3. Do keep tweaking, but it’s better than ever before. As for cat divisions, however, the above just seems wrong. Cat D is clearly too big and too wide, in numbers and in time spread. Cat C is quite chunky too. And at times the A-B division seems completely redundant, all while cat A’s often complain about low participation and B’s about tanking.
See where this is getting?
The problem is even though the ZRS ranges for each category seem comparable score-wise, points in different ranges seem to have different weights. Inadvertently, I assume, but it could be a problem.
First, look at the score ranges of the current cats. It’s:
A: 310 or whatever, there is no ceiling(?)
B: 170
C: 170
D: 170
E: 180
The problem is that, as an example, the value of each point in the cat D range seems to be higher than in other ranges since the time gap between winner and the cat E winner is so wide. It takes a lot for someone to go from top E to a competitive top D. Thus each point from 181-350 carries a heavy weight. Sure, it takes more to winning than just shaving off minutes on your finish time, but at the same time you can’t ignore that aspect either.
Of course, there is diminishing returns as you creep closer to the top, A and B. You put in a huge investment in training for rather small gains in race times. But still, I’m not so sure it’s a desirable property of ZRS scores to be “weighted” in this way. It’s thought to be a scalar, no? What kind of statistical distribution govern the score values? I bet it’s a weird looking one.