Not exactly. If guy A uses an action that costs 5, and guy B uses an action that costs 20, then A would act 4 times in the same time B would act only once - assuming they both start at the same number.
Okay, I see. You forgot to mention that you're getting rid of rounds and the limitation of only acting once per round.
Anyway, the suggestion I was making still holds, but the "instead of" in my phrasing needs to be updated.
Instead of letting tick counts go above 20, and then subtracting from them later, why not use rounds of 20 ticks each? So, instead of, for example, tick 20 being followed by tick 21, it would be round 1, tick 20, followed by round 2, tick 1.
That accomplishes the same thing, but it's a far less confusing nomenclature.
How so?
Well, you declare what you're doing, then other stuff happens, then what you declared actually takes place. What happens is that "other stuff" has made the action you're now supposed to take impossible? Which details of your action do you have to decide when you declare it, and which can you decide when it resolves? If the situation changes between when you declare an action and when it resolves, can you abort the action partway through? All those are non-trivial scenarios you'll have to account for.
More specifically though, you'll have to look at the interaction between attack range and movement. So, you move into range, then you declare your attack. What if your intended target moves out of range before the attack resolves? This is especially important if there are attacks that take more ticks than the smallest form of movement.
Imagine I'm trying to land a big attack, one that takes 20 ticks. Now imagine my target can move his speed as a 10-tick action. So, after I declare my attack, my target
will move out of range, and I
will have wasted my time. He'd be stupid to stay put and take my super-powerful attack. And he has plenty of time to get out of the way.
So, it only makes sense for me to declare a long action if my target has first locked himself into an action so long that he won't have been able to react until after I've already finished. But why will
he declare a long action? He's in the same situation as I am. I doesn't make any sense for
him to declare a long action until
I've already locked
myself into a long action. The first long attack won't be declared until after the first long attack has been declared. A paradox.
So, everyone only ever uses the fastest attack, because anything else can be easily countered. If the long attacks are never used, why are they even there?
And let's imagine if a 5-foot step only takes 3 ticks. That makes sense, that lets you move 30 feet in 21 ticks, a little slower than 30 feet per round. But let's say the fastest melee attack takes 5 ticks. That also makes sense, that's 4 attacks per round. How does a melee character hit a target who's 5 feet out of attack range?
Okay, I declare a 5-foot step. Now I'm close enough, so I declare my melee attack. % ticks later I hit. Except, no I don't. Because in the 7 ticks it's take for me to get into range and hit him, he's had plenty of time to take his own 5-foot step. That means he's
not in range to attack.
How does a melee character get any attacks in if his enemy isn't polite enough to stand still and allow himself to be hit?
Combat start - all characters roll d20 - that is your place.
Okay, but shouldn't a character with high dexterity and such tend to go before a character with low dex and such? How about, instead, you roll initiative as normal. Whoever rolled highest is assignes to tick 1. Everyone else is assigned a later tick based on how much he beat them by. You rolled 6 lower than the guy who goes first? You go 6 ticks after he does.
You're also going to need some method for determining the order in which people act when they both land on the same tick.