It's one of my pet peeves when I'm playing a game, is when I see something that I would really like to use (be it a stat or a weapon), but it was implemented into the game (or something else was implemented into the game that renders it useless) in such a way that you have to bend-over backwards to find a use for it, and even then it's very weak at serving it's purpose. So why did the developers take the time adding that item when they could have spent a little more time on other aspects of the game? I'm all for things being included for the sake of novelty, but sometimes it's less obvious that the thing you're aiming for is a novelty, so before you know it you're at a really high level in an RPG, and you find-out you've been pouring your points into a stat that is completely useless or has a much lesser effect later in the game than you'd think it would.
A lot of times I chalk this up to a lack of play testing. Many times when playing a game that lets you allot stats to your (mostly RPGs of course), there will be that one stat that looks very tempting. Upon putting points into the stat, you find that you like the effects given by this stat, say Magic as an example. So you spend your time playing this game putting all of your points into Magic (most of them anyway), but then you run into a problem. Closer to the end of the game you're having a much harder time playing the game, possibly even finding it impossible. So you decide to research it, you find some forums that discuss the game, and low and behold, everyone giving advice says to stay away from Magic because it becomes highly useless at the end of the game due to a large number of enemies that have a high resistance to Magic.
This is actually a true story, to some extent. Diablo II. In this game there are 8 classes to choose from, and each class has 3 skill trees you can work down. In theory, this leads to 24 possible character builds to choose from. In reality, you have roughly 20 potential builds that are great for the first stretch of the game, and a handful of builds that can continue to be as useful when playing in the higher-level content. So why bother with those 20-some other character builds? Why put them into the game? For flavor? Well, that's great, but now I have to start a new character because I've hit a sudden road block where my current character just struggles too much to get even one enemy down, and he spawned with 10 friends all right next to him that I still have to deal with.
Don't get me wrong, I love Diablo 2, and I even play a couple of those 20-some builds that are near-useless in the later game. The problem is though, I have to really dance around to make kills on those characters, meanwhile someone else will come along with one of those 5 builds and rips the enemies apart in mere seconds. If it's even a plural number of seconds. It's good to know that I'm skilled enough to play one of the lesser builds at a higher-level dungeon, but it's disheartening to be spending a good 10 minutes working on clearing my way through one hallway of a dungeon, then someone joins the game and they basically just walk through the place, barely breaking stride as the enemies seemingly drop to their knees in awe of his damage potential.
Which is basically what Rule 1 of my Design Philosophy is about. It asks the question when something is included into the game: "Who would want to use this, for what purpose, and is it a viable answer to that purpose to the same degree as other available answers?" If any of those questions cannot be answered properly, then it's back to the drawing board. Novelty is nice, but it should be saved for situations where it's more obviously a novelty.