Huh.
I'd like to say that I appreciate that people on both sides ran tests to try to figure out what was really going on instead of just taking sides and getting angry at each other. I'll admit, after philthepat's confirming test I jumped and was already thinking, "well, that's obviously screwed up and not how it should be!", but after Garlon's and jpeters's latest tests I thought, "hmm, okay, maybe it's not that bad for it to work that way".
Quote:
Originally Posted by OutS|der
There should be no scenario where a 1 rating should be competing in any level over Rookie if even that, I'd put them more in HS/College.
...
|
I don't really agree, only because do we ever see a 1 current rating? I mean, sure, I'd prefer if a 1/600 didn't result in this, but if it does simply because it makes the ratings that we do see work better for whatever reason, then okay, so be it.
Regardless, I find it all very interesting and I appreciate that we understand it better now. I'm glad that I don't really pay that much attention to MiL stats (I do a bit, like for deciding when to promote someone), but I'm even more glad I no longer feel like this is something we need to freak out about, just look into more.
Also, for the people who might not understand what some people are talking about with the word "floor", it means that between some range of ratings OOTP might have one formula work, but between the next range of ratings some other algorithm works. The way I learned about OOTP floors was when I looked at a spreadsheet someone (sorry, I forget who) made that "reverse engineered" how position training worked in OOTP and it clearly showed that between some range of ratings one thing happened and between another something else happened. Like, "if range < X, then do Y, otherwise do Z". You could also think of it as, instead of a straight diagonal line or curve on a graph, the line or curve suddenly changes its angle after a certain point. It might seem like a strange thing to program in, but I think it just sometimes accomplishes a certain purpose better.