Two is the new one. Five is the new two. Ten is the new five... maybe?

One of the biggest design challenges when working in an analog medium is numbers.  

MATH.  It fucks people up.  

That's a big problem when you're designing something that's meant to be fun.  People don't want to feel like they're working when they're playing (even though people are virtually always working when they're playing games)--it's the feeling that counts.  

Which is why I'm thinking about base 1.  

I'm definitely getting the terminology wrong, and I should look into what this concept is called--but I'm talking about using 1 as the lowest common denominator of the balancing system.  One is pretty average, and anything higher than one is good.  We'll say one is "normal".  

But then what's worse than 1?  Zero?  Zero means there's nothing there.  And if I want a troop to deal some damage, it has to be a fraction of 1 in order for it to be worse than 1.  

FRACTIONS.  They fuck people up worse than MATH.  

So,one is not a good starting point because you can't go down from there--but going up when your normal is one isn't great, either.  When your normal is one, you can double it, triple it, quadruple it--and those are all enormous changes, all achieved by adding only 1 at a time.  Instead, let's say your normal is two--now you can halve it, or add 50%, or double it... the same numbers are used (1-4), but their relative value is worlds apart.  See the difference?  You're able to tune your quantities more finely by changing your normal.  

However, if we make the leap to three--we have User problems.  Because my brain's a bit wired for math, I can do multiples of three pretty easily, and I can do halves or thirds of it.  Quarters are tougher, and fifths straight out--but you understand the problem already, by trying to think about those integers.  Thinking of 6 as double 3 is easy, but thinking of 11 as 3 2/3 times better than 3 is rough.  Players don't necessarily need to understand exactly how much better one option is than another, but it certainly helps when designing a deck-building game, or any game with variables the player is allowed to tweak.  When numbers are easy, understanding the systems is easier, which means players can play the game at a more optimum level with a minimum amount of confusion and frustration.  

Which is what we'll call "fun".  

Your normal needs to be something that allows for easy-to-comprehend worse numbers, but we can't get too high--100 gives you approximately one hundred times more fine-tuning capability than 1, but forcing players to add, subtract, or multiply numbers like 37 and 82 is even worse than not allowing them to understand exactly what all of the values mean.  On top of that, players understand there's a difference between 80 and 82, but understanding the difference is much tougher.  Don't forget you're going to go above your normal as well.  You'll hit values like 135 and 337, since 100 is "normal".  How different, then, is 135 from 13, or 337 from 34?  Just make your normal ten and be done with it, right?  

At the bottom of this discussion comes the question: what's the optimal normal that allows for sufficient fine-tuning of design, without compromising usability?  

Answer: Pffft, I don't know.  

Two, five, and ten seem like the best options.  Ten fails at keeping numbers small (and when tallying 8s and 7s together, the math isn't friendly), but it's a scale we all understand with ease--being able to say that a 3 is 30% of a 10 is pretty clear to most users.   

Five is nice, because there's not too much variance in the in-between integers, and working in 20% increments is pretty convenient, as well.  

Two keeps the numbers much lower and easier to deal with, but might not lack enough finesse for your game.  

Which is sort of the ultimate conclusion--you have to do what is right for your game.  Or, possibly, individual systems in your game.  

When tallying damage in a game with multiple pieces dealing damage to a variety of sources (or worse, having modified damage on top of the initial quantity), a normal of ten is probably too complex.  Five probably runs numbers too high... but two sounds pretty good.  However, if multiples of the same piece are dealing damage (so four pieces all deal 7 damage a piece), then we can depend on multiplication tables to ease the load on players (a bit).  Perhaps not enough for some, but it's certainly easier than adding 9, 6, 5, and 7 together quickly (that's not even 28... see how hard it is?).  

All of which is to say, computers are really good at math.  Utilizing a computer (as video games do) solves these User issues for you and opens up a lot of design space--but the constraints of keeping your game systems simple can lead to interesting design choices.  And if we're not making those, why are we making games?  

Although seriously, any of you feel like programming for us?  

And doing a lot of art?  

Let us know in the comments.