I don't honestly know how anyone can believe that Dice Pools are a "lighter" system than straightforward checks.
Now, I understand that certain people have certain types of brains, and others have other kinds of brains, and that certain brains may be more predisposed toward the more linear mechanics of a straight die + rank vs. difficutly or even simpler percentage or roll under check; whereas other brains are more connected to the concrete visual characteristic of rolling lots of dice.
But there are ways to address that which do not involve the complexities of the typical dice pool system. Whatever they might gain in terms of mathematical simplicity (which usually isn't that much anyways) they lose in terms of the complications and time involved in a typical die pool system. Its certainly not condusive to the system "just fading into the background".
The best dice pool system, and the only one that isn't really a hassle to play, is one where you roll a number of dice (either set or variably modified by the action/skills etc) against a flexible target number (difficulty) and then only consider the highest die rolled. Optionally, you can have a high roll "explode" (re-roll and add to the original value). This is easy enough, because you are only considering the highest die, the others can be totally ignored.
But very few die pool systems work that way. Octane is the only one I can think of offhand.
There are some, like Over the Edge or the D6 system, that work with a pool of dice, where you add the whole pool together and compare it to a difficulty. This kind of system is alright, as long as you don't end up having pools of too many dice, where adding becomes a serious hassle (this and other reasons cause the D6 system to suck ass).
But the default systems most people think of when they think "dice pool" are the WoD or Shadowrun style of dice pools; where you have a fixed or variable difficulty, roll the pool and then have to count up the number of "successes" (dice that overshot the difficulty) to determine a "degree of success". This essentially creates two different difficulty thresholds, the original difficulty number, and the number of successes. This essentially means you're figuring out two mechanics, which takes twice the time.
What really baffles me in all this is that recently both WW and Fanpro have thought to "simplify" their mechanics. But how have they done this? By replacing a variable difficulty
number with a fixed difficulty number.
This is the least useful "fix" imagineable. Its the kind of thing I would expect from WW, but it disappoints me that Fanpro is going that way with it, and there appears to be some truth to the claims some have made on Dumpshock that Fanpro is "imitating" the WW concept. I certainly can't believe the Fanpro boys just thought of the very same "fix" (which is hardly a fix at all) in a total vacuum.
This fix won't really be doing much to streamline the complexity of the system. You still have to check the difficulty, you still have to count successes. While having a fixed difficulty number is something of an improvement, it would have been a vastly greater improvement to get rid of the idea of "counting successes".
Have a mechanic where you have a variable difficulty, but ONE success is enough to succeed, always.
You want to make it a harder test? Raise the Difficulty.
Want to be able to judge whether something succeeds just a little or by a lot? Compare the single result to the difficulty; the higher it was, the better you succeed.
This is what seems to be the smoothest way to handle a die pool mechanic.
So why doesn't anyone have the guts to do it?
RPGPundit June 16th 2005