Defending One-Trick Ponies.
Let me start by saying that there are a number of things at least potentially wrong with point-buy systems. First, I have trouble with the whole concept of building balanced characters. Characters can never truly be balanced. Different players have different meta-goals within the game. Some want to succeed in combat. Others want to be the center of attention. While you can affect some aspect of these goals through the choices you make during character creation. Other aspects will always emerge from how you play the character, or how you interact with the other players. A "strong" role-player with a "weak" character can often "dominate" the table. This is true for nearly any definition of "strong", "weak" and "dominate."
This leads me to my second, more critical complaint. Point-buy systems force players to master the system in order to build their desired characters. Lets say I want to build a master swordsman. If I'm playing a point-buy game, then I (as the player) must have a strong sense of how to tactically manipulate the system to build an effective sword combatant. If I do a poor job, then the character's in-play abilities will vary significantly from my original intent.
For me, that's a design flaw. I shouldn't have to master the system to build my character--my character should be a master swordsman because everyone at the table buys into the idea. We agree that he is a master swordsman, and as a result, he can do master swordsmanly things.
Of course, none of this really addresses Chris's post. He argued that specialization reduces the opportunities for conflict. Basically, within their realm of specialization, the character dominates. Outside that, they are dominated. Since the conflicts are a foregone conclusion, they are not interesting.
While this seems straightforward enough, something about it bothers me. Sure, overspecialization could be a problem--but I think specialization is an effective way to ensure that each player has a role within the story. Having unique roles is, to me, much more important than trying to balance power levels.
I think there's an assumed definition of conflict here, that is much too narrow for my tastes. I feel that Chris's argument applies to what I call "Mechanics Based Conflicts". Here, success or failure is determined by the roll of the dice. Character decisions are largely tactical. The conflict is considered interesting when it requires several rounds of decision making and die rolling before anyone becomes the clear victor. There is also a strong, underlying assumption that success is always good and failure is always bad. Combat is the most common (and most overused) example of mechanics based conflicts.
Personally, I feel that mechanics based conflicts are incredibly weak. They focus on game play, not story.
Instead, I would like to focus on internal conflicts within the characters. A small innocent child has been possessed by a demon. Do I kill the child, thus removing the threat. Or do I let the child live--struggling to prevent the demon from harming others while I search for a way to banish it. Here, the conflict is the decision that the character must make. Say that I choose to kill the child, the actual combat is unimportant. I should succeed easily--I probably don't need to make a single roll. But, the results of my actions are important. How will I deal with the guilt? How will the child's parents react? How will the townspeople? Once I make a decision to act, the world is profoundly changed, and I must live with the results.
Also, failure should not necessarily be a bad thing. Let's say I make a character who sucks at combat. I get into a fight with the bad guys, and draw my sword. Again, the actual combat doesn't matter. They disarm me and beat me bloody. Big deal. Then they drag me off to their camp, where I'm tied to a tree. Sure, it looks grim, but now things get interesting. I have a chance to uncover clues about their plans, and I must use my silver tongue to try and worm my way out of this mess. In many cases, a good failure can create more interesting stories than a simple success. Had I fought off the attackers, the conflict would have ended. Now, it has grown deeper and more intriguing.
Once the conflict is bigger than simple actions, then the actions themselves loose their significance. We no longer need to balance PCs and NPCs at the action level. We can often get rid of multi-round, multi-roll resolution systems. We're not focused on the action's result, we're focused on the changes those results make in the world. And, from this vantage point, Chris's problems largely vanish.
One last example, look at the Princess Bride. No singe opponent ever comes close to matching Westley with a blade. Yet, he is forced to surrender when outnumbered by Humperdinck's guards. While in the pit of despair, he is completely at the bad guy's mercy, and he is reduced to using trickery and deceit when recovering from a bad case of the mostly-deads.
Looking at it with an action-focused view, he either dominates or is dominated at every step. Yet, the resulting story is quite satisfying. The movie created numerous, interesting conflicts for Westley, without ever feeling the need to pit him against an opposing sword master of equal skill. We should strive for the same in our stories.