If level doesn't mean power level, and thus all characters of a given level are of equal power, then what useful purpose does it serve?
If a wizard is more powerful than the fighter, then he's a higher level than the fighter.
If a fighter is more powerful than the wizard, then he's a higher level than the wizard.
If you then want to have one be consistently more powerful than another, or have one achieve power at different rates, you can do it by adjusting the rate of levelling.
Want the fighter to start out more powerful, but for the wizard to gradually eclipse them?
Then have the fighter start at level 3, but level at a steady rate and never go above some arbitrary level like 12. Meanwhile, the wizard starts at level 1, and stays there for a longer time than it takes the fighter to level, but each subsequent level is easier to reach than the last, and you aren't capped at 12.
With this approach, it's also easy to come up with power curves that haven't previously been used.
Magic can be magical, mysterious, dangerous, and able to do things that are completely outside of the capacity of nonmagical means.
It can do all of these things without being able to do the things that nonmagical means can do better than nonmagical means can do them.
It can even do all of these things if it's completely unable to do things that nonmagical means can.
Magical and powerful are not synonyms.
The issue that I have with consistent vs. swingy is that situational balance like this is only balance in some situations.
If you have a campaign where every adventuring day has at most one fight, the fact that the wizard is only at their peak for one fight per day isn't swingy at all. It's as consistent as the fighter is. So if that peak is higher than the fighter's, there's no balance, because the wizard is strictly superior.
A campaign where the adventuring day has approximately the same number of fights that the designers balanced around will have the difference between swingy and consistent average out, as designed.
Then the campaign where you have twice the amount of fights per day as the designers anticipated, the fighter is clearly superior. The swingy wizard is hosed.
This is the reason I prefer systems where everyone has approximately the same amount of swinginess or consistency, or that the degree of swinginess is a result of your choices during that adventuring day, rather than the choice of which class to write down on your sheet.
It frees you from the constraints of how much combat the designers anticipated, leaving the amount of combat up to the dictates of the individual campaign.
The issue I have with balance across pillars is pretty much the same one I have with balance across levels and balance across the adventuring day.
In any campaign that doesn't cover the range of levels the designers anticipated, in any adventuring day that doesn't have the number of fights the designers anticipated, and in any campaign that doesn't mix the percentages of play in each pillar in the proportions the designers anticipated, any "balance" the system can claim to have flies out the window.
In a system where level=power, and everyone levels at the same rate, you can have a campaign starting at any level, and running though any set of levels, and still have balance.
In a system where swinginess isn't determined by class, then any mix of classes can have adventuring days of any arbitrary length without hosing OR over-rewarding any character due to a choice made in char-gen.
In a system where each class contributes meaningfully in each pillar, in their own way, and no class is "best", then any class is viable in any campaign, adventure, or adventuring day, regardless of how much play happens in any pillar.
In a system that all three of these things are true, a player can pick any class that appeals to them, and know they're going to be able to meaningfully contribute regardless of where the adventuring day, session, or overall campaign takes them.
That's the sort of system I want to play in.
If a wizard is more powerful than the fighter, then he's a higher level than the fighter.
If a fighter is more powerful than the wizard, then he's a higher level than the wizard.
If you then want to have one be consistently more powerful than another, or have one achieve power at different rates, you can do it by adjusting the rate of levelling.
Want the fighter to start out more powerful, but for the wizard to gradually eclipse them?
Then have the fighter start at level 3, but level at a steady rate and never go above some arbitrary level like 12. Meanwhile, the wizard starts at level 1, and stays there for a longer time than it takes the fighter to level, but each subsequent level is easier to reach than the last, and you aren't capped at 12.
With this approach, it's also easy to come up with power curves that haven't previously been used.
If you're going to have magic, magic should be magical, and completely beyond anything nonmagic. Moreover, it should be mysterious and dangerous.
Magic can be magical, mysterious, dangerous, and able to do things that are completely outside of the capacity of nonmagical means.
It can do all of these things without being able to do the things that nonmagical means can do better than nonmagical means can do them.
It can even do all of these things if it's completely unable to do things that nonmagical means can.
Magical and powerful are not synonyms.
Situational balance. Fighters should perform pretty consistently across most of the adventuring day. Wizards should be more swingy. And: No one beats a prepared wizard. Conversely, the unprepared wizard dies quick.
Going back to the "combat as war/combat as sport" thread, I guess the fighter should better support the combat as sport style and the wizard should better support the combat as war style. That's not an absolute, but it's the right sort of nod for me. I'm also okay with throwing in a warlock for the players who want a magical combat as sport character and something for the martial combat as war players -- I'd probably say the ranger, but the rogue or certain incarnations of the warlord could serve that roll, too.
The issue that I have with consistent vs. swingy is that situational balance like this is only balance in some situations.
If you have a campaign where every adventuring day has at most one fight, the fact that the wizard is only at their peak for one fight per day isn't swingy at all. It's as consistent as the fighter is. So if that peak is higher than the fighter's, there's no balance, because the wizard is strictly superior.
A campaign where the adventuring day has approximately the same number of fights that the designers balanced around will have the difference between swingy and consistent average out, as designed.
Then the campaign where you have twice the amount of fights per day as the designers anticipated, the fighter is clearly superior. The swingy wizard is hosed.
This is the reason I prefer systems where everyone has approximately the same amount of swinginess or consistency, or that the degree of swinginess is a result of your choices during that adventuring day, rather than the choice of which class to write down on your sheet.
It frees you from the constraints of how much combat the designers anticipated, leaving the amount of combat up to the dictates of the individual campaign.
Could be balance across pillars. Fighters are A+ at combat. Wizards are C at best. But then when you're crossing large distances or uncovering a lost mystery (exploration), Wizards are quite a bit better than fighters, due to magic that reveals the future and bends space and time to their will.
The issue I have with balance across pillars is pretty much the same one I have with balance across levels and balance across the adventuring day.
In any campaign that doesn't cover the range of levels the designers anticipated, in any adventuring day that doesn't have the number of fights the designers anticipated, and in any campaign that doesn't mix the percentages of play in each pillar in the proportions the designers anticipated, any "balance" the system can claim to have flies out the window.
In a system where level=power, and everyone levels at the same rate, you can have a campaign starting at any level, and running though any set of levels, and still have balance.
In a system where swinginess isn't determined by class, then any mix of classes can have adventuring days of any arbitrary length without hosing OR over-rewarding any character due to a choice made in char-gen.
In a system where each class contributes meaningfully in each pillar, in their own way, and no class is "best", then any class is viable in any campaign, adventure, or adventuring day, regardless of how much play happens in any pillar.
In a system that all three of these things are true, a player can pick any class that appeals to them, and know they're going to be able to meaningfully contribute regardless of where the adventuring day, session, or overall campaign takes them.
That's the sort of system I want to play in.