For example, suppose you've got a position player that is posting great hitting numbers, but has a poor zone rating. He is costing you WAR, because position player WAR is offensive WAR + defensive WAR. Moving that guy to his natural position removes the negative WAR from the calculation, and now you have a more valuable player in terms of WAR. Of course, everyone loves a CF, middle infielder, or catcher that hits well, but he's costing you if he's not fielding his position well, both in terms of his total WAR and that of your poor pitching staff surrendering more hits than they would otherwise and thus reducing your total rWAR.
By the same token, you may be better off tolerating a punch and judy hitter in an up-the-middle position if he is an elite defender. These guys can post a solid 3 WAR without even sniffing league average offense. My rule of thumb for these kind of guys is if they can muster maybe 85-90 OPS+, they're going be valuable in terms of both WAR and rWAR.
Perhaps you've got a starting pitcher that posts strong ERAs, but rarely goes more than five innings. His WAR is going to be less than a pitcher with more stamina. Good SPs are hard to find, so you'll probably want to keep him in the rotation, but consider this when you are shopping for starting pitching in the future. Good SPs that can go deep into a game are inherently more valuable than the five inning wonders like Blake Snell.
If you've read my team news, you'll know that I evaluate my rosters annually to identify positions that are not contributing their fair share to the team's total WAR. I find it to be a valuable metric to optimize my roster construction.
