@Sephiroth: (I apologize in advance for the word count/length, but I'd really, really appreciate it if you'd skim and clarify) I'm not entirely sure you're saying about the graph- there's multiple ways it could go. Are we talking about it creating intuitive assumptions (like "If WG play at a 75% win rate from now on, they can hit 50%?) that are wrong? Success rate = points won/points contested, so going by v = d/t, this is analogous to a graph of velocity over time. Those graphs are widely used and do convey useful data- like acceleration (the analogous thing here, a change in success rate, is pretty useful for picking up changes in our knowledge and understanding of clan performance). I mean, to simplify things, here's how I arrived at this point:
- Points vs. Time Graph is standard, it gives us an idea of where clans are at a given update period in terms of points won. But some clans play significantly faster/slower than others and consequently inflate/deflate their apparent performance on this graph.
On this chart (from Division D a week ago), you see that with Dutch. They look like they're at a better place than they really are, 'cause they've contested more points.
- So I tried Points vs. Points Contested, because now it puts everyone on the same clock.
This one helps you see that the Juggernauts have actually been outperforming the Dutch for a while. There's also that dotted "ceiling" line indicating the max points possible at that many points contested. If we just take the slice below that ceiling and rescale it so that the ceiling touches the top (100%), we end up with this graph-
- Success Rate vs. Points Contested
This mainly exaggerates/makes clearer the differences between clans, since being ahead by 3 points 30 points in is different from being ahead by 3 points 90 points in. Like you pointed out, the amount of uncontested points places constraints on how much improvement's actually possible for a competing team. So that's the advantage I saw to that particular visualization.
Or even more simply, it's just rescaling a win rates vs. time graph to account for differences in play speed.
Sorry if I'm being daft here; I'd just appreciate a little bit of clarification since I actually don't understand. It'd be genuinely helpful.
I agree on brevity. I hadn't really looked at everything beforehand so part of it was just me figuring things out. In Division D, I tried to go a bit more in depth but here, well, I might've been out of my depth and unable to make up for it. :/
@MotD: I can see where I'm in the wrong here. Apologies for repeatedly railing on you in my frustration. It seems to me that there was some avoidable miscommunication on both our parts, then, and I accept full responsibility for the large chunk of it that happened on my end. At least on my end, I was projecting some malice and disdain onto your behavior, so I'll also freely admit that I'm probably in the wrong there too. I believe you were projecting some of that onto me as well... but then again I don't know the extent of your views beyond the very limited information I have. I'd be more than happy to avoid this issue, though; it seems like I've crossed a few lines already.
Edited 7/22/2016 03:29:03