To make that simple, If you are taking the ball from the kick-off you will be getting 15/20 metres every time, good for the stats.
If you hit the line hard from 3m out and score a try, bad for the stats.
Some people seem to have reservations about looking at stats at all. What's the purpose of stats? They are:
1. A tool
2. Not an answer
So they are a starting place, not a destination.
Same with looking at economic growth, jobless rates, etc, etc. No one figure is an answer. All figures are distorted, simplistic and misleading. And even all the figures combined are never the answer. That's the nature of figures and models. But it would be irresponsible to run an economy without considering these figures, wouldn't it? I'd argue the same applies here.
The only stat that comes close to an answer is the footy scoreboard. Having said that, stats can open the eyes to patterns that weren't seen before. And, for me at least, these stats do that in some ways.
Re your comments, Woodsie, most variations are swings and roundabouts so they tend to even out. But agree that on the minus side, kick return metres can be easier metres. On the plus side, being selected for kick returns is a positive assessment of that player's power running. Didn't want to complicate matters with the first set of figures provided. But FYI, in 2016 Vave has 122 metres, Lussick 75, Jake 52, Starling 46, Nate 34, Burgess 14, and Lawrence 3 metres. The average between these 7 players is about 50 metres, so that's the standard for comparison. So the possible distortion in prop stats primarily applies to Vave (72 metres variation) and Lawrence (-47 metres variation). So overall, not really a whole lot in it. Unfortunately NRL only provide totals, not even weekly figures, and no indication of how many runs were kick return runs. So a bit hard to determine with more precision how much of a distortion it may be. I'll bet they have this data they just don't publish it.
If only they openly let us access all of their player statistics raw data - a simple export option is all that is needed - extremely simple to program - and essentially cost-free to the NRL - then some real good analysis might possibly be done.