If there was one "best" setting, they wouldn't need to put knobs on it, would they? :?

Set it up, listen, fiddle with the knobs, and stop when it sounds good. Then you push the big red button and go at it.

Mehrad, welcome to this forum!!
First off, that Rode is a pretty nice mic, eh? Very versatile, to be sure. I don't own that particular microphone, but what I can contribute can be applied to any variable-pattern mic.
The settings for the polar pattern selection are ultimately going to be dependent upon the (a) source, and (b) the acoustical environment the source is in. If you are tracking, say, a vocalist in a nice large room with controlled, fairly neutral accoustics, you can play with the pattern selector abit more without getting into trouble. You should notice that as you vary the control, there will be subtle tonal changes between "wide" and "narrow" cardioid patterns. This will change with male vs. female singers. It will also help to go for the "wide" if the talent moves their head around a lot. Narrowing the pattern usually tends to emphasize the "proximity effect" you'll get from the talent.

If the room is not so flattering, you will definitely need to "tighten up" the pattern to minimize the amount of "room" being picked up.
Putting the mic in "omni" will "flatten" the freq response and eliminate the "proximity" issue. This is generally used if the mic is being used in a good room for acoustic instruments (i.e.,piano, strings,etc)...
The Figure 8 pattern will many times work for vocals and/or horn sections, with the talent split between the 2 sides. This tends to have a slight midrange emphasis, and a "voice" all its own.
Like Zemlin said, you gotta experiment, baby!

Good advice, moon baby! the only thing I'd like to add is that I've found that 11 am omni to cardioid works well for some of the female singers I've been tracking, not so much with the males. - by the way, this is on k2's, same capsule but the tube version. My k2's are also super on percussion