The balance formula thread

[Warlord Zsinj]

Deleuze is not pleased

[Peet]

This is my balance formula

[TheFatController]

balanced annihilation 2.0 (beta) will harness the power of crowdsourcing to achieve perfect balance no maths involved!

[1v0ry_k1ng]

the perfect balance of con vehicles to flash tanks

[KDR_11k]

Dem red maffs go fasta.

[smokingwreckage]

OK: here's a mod. Every single unit has the same cost and buildtime, but every time someone in a game builds it the cost and buildtime go up. In fact they increase at an increasing rate for every unit built.

Ideally this mod needs a very large number of somewhat similar, partly interchangeable units.... in other words a TA derivative of some sort, like UTASP or whatever it was.

EDIT: it wouldn't be balanced for any given game, but on average over time it might be

[Evil4Zerggin]

It's already been all but done (although perhaps not in Spring), and it's actually fairly common...














...it's called fantasy draft >_>

[Argh]

Dear Bana: "fungal math" is an apropos description of a type of chaos that occurs within a given timeframe.

The reason I've made up this term is not to be cute.

Like a fungus growing in a petri dish, the chaos of a given set of starting conditions inevitably leads to certain chaotic growth curves, assuming that an event has not interrupted the cycles that permit the chaotic event to continue.

However, the exact size of the results is not predictable at any point in time, other than theoretically if all events happen perfectly. Which, since they never do, means that you never can really tell people anything, other than what's theoretically possible, vs. what actually happens.

On the "full of shit" argument:

People like you tend to get pissy when I actually show you the math to perform serious analysis with, so I've quit bothering, and usually just ignore these threads, since it's the same stupid crap every time by people who don't get it.

But since you want to play with numbers, here's some math for you to look at:

In a two-resource game design model (let alone how many resources are actually present in *A games- they use six if we're being minimal, 10 if we're being really accurate)...

If one resource is Metal and the other is Energy, and Energy may be spent to create more Metal on a 1:5 tradeoff on initial cost (i.e., no continuing cost), then how quickly do you gain Metal, if you gain 5 new Energy a turn (a linear progression, no tradeoff)?

Turn 1: 1 Metal. 5 Energy.
Turn 2: 3 Metal. 10 Energy.
Turn 3: 6 Metal. 15 Energy.
Turn 4: 10 Metal. 20 Energy.
Turn 5: 15 Metal. 25 Energy.
Turn 6: 21 Metal. 30 Energy.
Turn 7: 28 Metal. 35 Energy.
Turn 8: 36 Metal. 40 Energy.

Each turn, we're seeing 1 more at a tradeoff of 1:5. It's a steady growth curve, and the actual metal per turn is growing, as well as our total.

Now, let's look at trades, or the non-free growth curves followed by most RTS games (in one way or another- that's a complex topic).

In our second model, you can trade 3 Metal for 10 more Energy on the next turn and from then on (i.e., the MM / Fusion econ with a lot of noise removed).

So, if the player does that every turn, using our previously-defined tradeoff of 1:5, starting with that same starting Energy output of 5:

Turn 1: 1 Metal, 5 Energy.
Turn 2: 2 Metal, 5 Energy.
Turn 3: 3 Metal, traded for 10 Energy, 15 Energy for next turn.
Turn 4: 15 Energy = 3 Metal. Traded 3 Metal for 10 more Energy.
Turn 5: 25 Energy = 5 Metal. Traded 3 Metal for 10 more Energy, 2 Metal remains.
Turn 6: 35 Energy = 7 Metal + 2 from previous turn, 9 Metal available. Traded for 30 more Energy next turn.
Turn 7: 65 Energy = 13 Metal. Traded 12 for 40 more Energy. 1 Metal remains.
Turn 8: 105 Energy = 21 Metal. Traded 21 for 70 more Energy... sorry, just had to add that turn, so that people would really get it...

Thus far, the second player has zero Metal, because they keep trading it for Energy. The other player has 36. Looks great for the first player thus far.

But I can tell you, with utter certainty, which player's going to win this game, if the Metal is no longer being spent on Energy, and that 36 doesn't show up in player two's base by turn 10...

I assume most of you won't bother doing the math. Let's just say it doesn't take long before the second player's curve skyrockets, and then there's a staggering difference in available Metal per turn at that point. And even if they trade some Metal, they're still so far ahead, it's not even funny.

At any rate... if you want to talk seriously about this stuff, that's cool and all. I just showed you why DPS is pretty much the wrong end of the math to look at, though. Economies determine what's actually available when. It's why most serious Starcraft games in Korea rarely involve many tech upgrades, and why they don't play on huge maps.

[1v0ry_k1ng]

its called MM economy

[Regret]

It's why most serious Starcraft games in Korea rarely involve many tech upgrades, and why they don't play on huge maps.

Oh hahahahah, we have an expert in our midsts!
Seriously, try to sound less like a pompous self-proclaimed genius.

Blizzard rolled a double six when developing warcraft, then used similar principles in starcraft+a decade of balancing and you say it all has divine math behind it.

Goodness Argh, I wish I could channel your ego-inflating bullshit powers for they are truly legend.

[Regret]

As a separate point:

I made this thread have a good laugh about all you people arguing about balance.

Now to spoil my own fun, here's the answer on how balance is done:

You balance a complex RTS game by giving units some arguably logical stats (read random) at the beginning of game creation, after which you proceed to tweak the stats constantly over a undefined period of time until you get a game that 'feels' right (starcraft) as opposed to game that feels wrong (perimeter).

By stats I mean unit stats/features/techtree/whatever.

[Licho]

There is now simple way to check how accurate your balancing really is - just by checking what ACTUALLY happens in REAL games.

http://planet-wars.eu/ModStats

I wonder how are other than CA mods doing though.

[Gota]

Pointless for CA though...since u keep making changes...If you stopped making changes than collected statistics for a year...
Like..the radar vehicle...it might have decent hp and excellent radar coverage for its cost but people don't bother using it atm cause they do not realize how good it is...will you still buff it?no...you need a stable release to test over time...You have the statistical tools but you still balance intuitively...

[Licho]

Waht are you talking about? Radar vehicle is used in 13% of games atm.. ships or hovers are used in less % :)

Median of unit type usage is 10%

And I would not be surprised if BA actually changed more over a year period. Units becoming OP and UP. We tweak often thats true, but tweaks are usually small scale few % changes.

[Gota]

Eh..not according to your own stats website..radar units are hardly used...also....what do those statistics give you?should an artillery piece be used as much as a mex?should it be used as much as a fusion?should it be used as much as a gator?

[Licho]

I think you are confused, as I said, its being used in 13% of games. Median is 10%. Its unit used more than average unit :)

You probably fail at understanding data.

Anyway most important for practical balancing are cost damaged/investment and cost damaged/cost lost battle effectivity.

Unit usage % was added just to identify underused unit for deletion, since we have too many models to replace.

[Gota]

Under % of games used the arm t2(or however u want to call it) radar vehicle gets 3.33%

And yeah keep up the good work...Ill check ca out when its in the 5k+ to see how it worked out for you.

[Licho]

There is no arm t2 radar vehicle.. nor kbot... in CA..

Please if you are totally ignorant about something, at least stfu and dont make fool of yourself..

[Argh]

@Bana and Regret:

I love how you guys can't actually talk about the issues. It's a lot easier to insult me than to provide a serious counter-analysis, isn't it?

Blizzard rolled a double six when developing warcraft, then used similar principles in starcraft+a decade of balancing and you say it all has divine math behind it.

Who thinks I've said differently? They made a good simple game, and have kept honing it, because their sales figures allowed for it. There's nothing magical about it.

High-end StarCraft play is short, and is not played on large maps, because that's where the initial economic peaks are, and the games are invariably played for rush.

Large maps would totally change the value of time, basically. Which is largely what I've been trying to drive home here, twice now.

[imbaczek]

http://planet-wars.eu/ModStats

this is almost what I had in mind when I was talking about unit efficiency matrices and RPS cycles in the other thread, but not quite - it comes close, though.