From ENWorld, in a discussion regarding the roles in 4E D&D:

Kamikaze Midge wrote:
The fact is that in any point-based resolution system like D&D combat, there's only two things to do:

Take away enemy points

or

Stop your points from going away

I don't see that as an argument that the controller and defender are extraneous. Controllers take enemy points, and defenders protect yours, just in a different way than strikers or leaders.

But I think while that argument is technically correct, it oversimplifies things. It's like saying you can boil down all sports to getting or moving a ball. Well, yeah, but there's a bit more to it than that.

However, it does bring up a good point that if there's no movement, powers, team strategy (what makes D&D combat interesting) than your combat can devolve into a boring number depleting fest.

The use of the word 'matters' seems to skipped the question "Does anything matter in a game?" and gone straight onto "Does this and this class matter?"?

There are theoretically infinitely many things to try to do. Take away enemy points. Increase the rate at which you can take away enemy points. Increase the rate at which you can increase the rate at which you can take away enemy points. Etc. And similar for the rates of rates of rates ... of enemies taking away your points.

Leveling up, for instance, raises your take-away-enemy-points derivative and lowers enemies' take-away-PC-points derivative.

Controllers, in a fight set up to not marginalize controllers, are theoretically supposed to manipulate the derivatives of the quantities which strikers and leaders manipulate. (Obviously they can also deal damage, but I'm talking about the controllery part of the controller, not the other parts). Defenders are supposed to manipulate the derivative of the quantity leaders manipulate.

In practice I don't think second derivatives ever come up. Also in my experience a hastily designed fight often marginalizes controllers and to a lesser extent defenders, or severely overemphasizes defenders, depending.

As abstracted and reapplied to other games? Maybe a very emphasized point in the text about any situations which, if consistently adlibbed poorly, will marginalize players asymmetrically?
Or "you get what you measure", and if you measure HP vs HP, people might tend to focus just on those numbers and methods of manipulating them, even if you'd like them to do something else ("roleplay" for instance) during combat?
Or take a line from Polya and at some point during design focus on the abstract corner cases: playtesting DitV, don't just check what happens if someone has lots of d4s and nothing else besides their low d6 stats, and what happens if someone has "lots" (aka keep the average sum of dice the same) of d10s and nothing else besides their low d6 stats, but also what happens if someone has a bunch of d4s and some d10s besides their low d6 stats.

That's what I thought of while thinking about the quote. :)

Even if the game did boil down to simply "taking an opponents points and preventing someone from taking your points", I can instantly see the value in more than just two roles.

With a single pool of points for each side to play with...

(I've added in parentheses the role type that I believe 4e is trying to make synonymous with this function, as far as I understand things...)

You get:

• The character who's primary objective is to take away someone's points. (Striker)
• The character who restores your points when they get lost. (Leader)
• The character who increases someone's potential to steal points. (Controller)
• The character who reduces someone's potential to steal points. (Defender)

Without the second two options you get a very bland game, but with them a range of basic strategic options manifest.

And that's only if each side shares a single communal pool.

If each character on each side had a pool to worry about you could also include:

• The character who transfers points to where they are most needed. (Probably a sub-role of Leader or Defender)
• The character who identifies the weakest/strongest characters on a given side. (Probably a subrole of Striker or Controller)
• The character who hides the weakest/strongest characters on a given side. (Probably a sub-role of Defender or Controller)
• (You could come up with 6 different combinations of the base roles and work out a subrole that both could cover adequately)

Once you throw in two basic pools of points for each character (Hit Points/Willpower, Hit Points/Magic Points, Force/Chi, Glory/Honour, etc...) you instantly get a range of character archetypes that can shift points between the pools in addition to the shifting points between characters.

But this is all in reference to 4e D&D, and it seems that the current incarnation has only one pool to worry about. Hit Points.

I generally agree with Whiteknife's final line, and that's one of the reasons why I try to ensure that there is plenty of colour to mirror within the narrative the effects that are playing out within the dice mechanisms. I also try to ensure that when I design games there are tactical and narrative benefits from working strategically with other players (or against them), rather than just going head to head. 

Just some thoughts...

V

It also just occured to me that one might also look at it in terms of arithmetic operations.
Addition
Subtraction
Multiplication
Division
Not quite sure where to go with this though.

There do need to be other things going on beyond numbers though.
There is something rather attractive however about big numbers.
A lot of people on other boards have been commenting that strikers are the "sexiest" of roles simply because of their ability to deal lots of damage.

I think that if I were designing a game to use roles kind of like 4E does, I think I would give every role the ability to deal big damage and reduce damage, but have different qualitative factors riding on the numbers.

Another point of view: I used to have an old AD&D 2nd Ed manual, Wizards' Manual or something. (I loved playing wizards)

It classified the spells in terms of their relation to COMBAT specifically, as being in one of six categories:

• Offensive One: Deal damage directly to your opponent.
• Offensive Two: Effect a change such that your team is better able to damage your opponent (e.g. buff spells)
• Defensive One: Make it harder for your opponent to cause damage (e.g. Touch of Fatigue)
• Defensive Two: Make it harder for you to be damaged
• Reconnaissance: Learn more about the combat situation to make better decisions
• Miscellaneous: non-combat spells

The difference between category one and category two is that category one targets the opponent, and category two targets the good guys. These are somewhat analogous to the categories of Striker, Leader, Controller, Defender though it's not a perfect match. I would argue that you need all four categories to survive when you don't know what you'll be facing.

For example, if you've got nothing but strikers in the form of fighters and fireball-casting wizards, they're in a tough spot if they can't hit an enemy's AC and Saving Throws. If, instead, a few of those wizards had "Bull's Strength" spells, together they are both suddenly useful. The wizards don't need to beat any Saving Throw, and the Fighters now climb to a reasonable attack score to hit the enemy.

At least, that's how it is intended to work in principle I believe. D&D itself (in my experience) has enough loopholes that tougher-than-one's-level-would-indicate characters are easy to generate.

Dan