TA Derivative Balancing Theory, Part Deux
Moderator: Moderators
History seems to be repeating itself.
The analogy between argumentation and balancing is flawed. Truth is an ideal, it does not recognizably appear in reality. It is merely one end of an artificial binary between "Truth" and "Falsehood" on which we attempt to measure the abstract "Validity" of something.
Balance is still a matter of approximation, and it is dependent upon individual interpretation. By reducing a model to minimal complexity you will get closer and closer to "Balance" as long as everybody is speaking in the same terminology with the same assumptions and the same individual affects. However, the problem of perspective complicates this.
To provide a practical demonstration of individual interpretation, Tired, you posit that I closed the prior thread due primarily to a perceived abuse of a moderator's power and position. I, however, did so for a complex of reasons of which that is similar to only one.
That said, I must close this post on two notes which together cannot but produce discord. The first is that your theories on balance, while flawed, are still useful and should be further explored. If we were to discard every statement for the slightest error, we would not speak or write. Second, I see no reason to close this thread as of yet, and thus bid you a fair day.
The analogy between argumentation and balancing is flawed. Truth is an ideal, it does not recognizably appear in reality. It is merely one end of an artificial binary between "Truth" and "Falsehood" on which we attempt to measure the abstract "Validity" of something.
Balance is still a matter of approximation, and it is dependent upon individual interpretation. By reducing a model to minimal complexity you will get closer and closer to "Balance" as long as everybody is speaking in the same terminology with the same assumptions and the same individual affects. However, the problem of perspective complicates this.
To provide a practical demonstration of individual interpretation, Tired, you posit that I closed the prior thread due primarily to a perceived abuse of a moderator's power and position. I, however, did so for a complex of reasons of which that is similar to only one.
That said, I must close this post on two notes which together cannot but produce discord. The first is that your theories on balance, while flawed, are still useful and should be further explored. If we were to discard every statement for the slightest error, we would not speak or write. Second, I see no reason to close this thread as of yet, and thus bid you a fair day.
Your meta-posting and contemptuous attitude does not make you more right, Tired. Argh point still stands.
Just like a chaotic system being deterministic yet impredictable, many people feels TA has so many factors interacting in so many ways it is impossible to write an ever-valid balancing formula.
If you disagree with this statment, there is a simple way to settle the issue. Tell us what the formula is. Let us check it works. After all, you did state it's "easy easy easy", right?
Many people think it's not true.Tired wrote:Anyway lurker, balance is easy easy easy to achieve, as I've pointed out here and in other threads. All it does is force all units through their common characteristics into a common formula where you can achieve a zero-sum.
Just like a chaotic system being deterministic yet impredictable, many people feels TA has so many factors interacting in so many ways it is impossible to write an ever-valid balancing formula.
If you disagree with this statment, there is a simple way to settle the issue. Tell us what the formula is. Let us check it works. After all, you did state it's "easy easy easy", right?
I'll give you a simple general formula:
1. Gather data on how often players build different units, normalized with respect to cost of course.
2. If a unit is built more often then others, increase it's cost, and vice versa.
3. Repeat
Of course, it's just one example among many, that would appear simple and effective in most situations.
1. Gather data on how often players build different units, normalized with respect to cost of course.
2. If a unit is built more often then others, increase it's cost, and vice versa.
3. Repeat
Of course, it's just one example among many, that would appear simple and effective in most situations.
-
[Krogoth86]
- Posts: 1176
- Joined: 23 Aug 2007, 19:46
Imo at a certain point you have a situation where maths won't bring you any further and it just gets "personal"...
To give an example:
You now have a situation where to unit types are balanced the way you wanted them via your formulas - e.g. let's say 3 type A units kick one type B unit's ass. So everything seems to be fine. But now when you go ahead and put up the numbers suddenly the type B unit wins the "equal" fights because of things like let's say AoE which has more importance here. What to do now? When assembling in greater numbers a specific amount of A units don't beat the type B units anymore. When you go and try to "correct" this you'll probably find out that there's no real solution as when "fixing" the situation with lots of units you'll probably mess up the situation with few units and the other way round...
So in the end it's "a matter of taste" which way you'll put it - maths won't help you here anymore. That's of course no degradation of using some sorts of calculations for balancing - generally it's all about numbers and so you have to use some kind of system if you don't want to have a more or less random balance. But when it comes to such details you have a situation without a right or wrong and so maths aren't of any help here anymore...
@Zpock:
Your "Trial&Error" method might be a method you can use though it has two disadvantages imo:
1.)
It take helluva time as you might ping pong around the values you'll have in the end...
2.)
The method might lead to "wrong" results. It only works on an ideal, plain map. When playing on maps with lots of relief there might be a high demand for artillery or high-trajectory firing units which aren't that much prefered on other maps. So the Trial&Error Method fails to consider those facts too and you for example end up with a crippled down artillery which is of only little use on maps with few relief (and so you'll improve your units to find them OP on the hilly map again)...
To give an example:
You now have a situation where to unit types are balanced the way you wanted them via your formulas - e.g. let's say 3 type A units kick one type B unit's ass. So everything seems to be fine. But now when you go ahead and put up the numbers suddenly the type B unit wins the "equal" fights because of things like let's say AoE which has more importance here. What to do now? When assembling in greater numbers a specific amount of A units don't beat the type B units anymore. When you go and try to "correct" this you'll probably find out that there's no real solution as when "fixing" the situation with lots of units you'll probably mess up the situation with few units and the other way round...
So in the end it's "a matter of taste" which way you'll put it - maths won't help you here anymore. That's of course no degradation of using some sorts of calculations for balancing - generally it's all about numbers and so you have to use some kind of system if you don't want to have a more or less random balance. But when it comes to such details you have a situation without a right or wrong and so maths aren't of any help here anymore...
@Zpock:
Your "Trial&Error" method might be a method you can use though it has two disadvantages imo:
1.)
It take helluva time as you might ping pong around the values you'll have in the end...
2.)
The method might lead to "wrong" results. It only works on an ideal, plain map. When playing on maps with lots of relief there might be a high demand for artillery or high-trajectory firing units which aren't that much prefered on other maps. So the Trial&Error Method fails to consider those facts too and you for example end up with a crippled down artillery which is of only little use on maps with few relief (and so you'll improve your units to find them OP on the hilly map again)...
You're all idiots.
How much "right" and "wrong" is there in 2+2? Does popular opinion matter? Does wikiality apply? If 2/3 players say that 2+2=5, does 2+2 really =5?
Saying that I'm wrong without understanding my argument is a very clever approach towards frustrating me, but at the moment I'm all frustrated out. Try me again tomorrow - if I check the forums you might get better results.
For those who aren't deliberately baiting me, which I'm sadly left to presume is all of you, I can't understand how you can't follow the simple line of logic, which no one here's debated.
For a balanced equation, I could say that speed was worth 10x as much as Health. As such, it would take many times - perhaps 5x - as much metal invested in Flash Tanks as invested in a Sumo to kill a Sumo. In BA, that would equate to something like 10,000 metal in Flash Tanks. This would be balanced.
I could say that range was worth only 2% as much as dps, and redesign a Flea that could fire from one corner of a 16x16 map to an opposing corner and cost 200 metal without changing any other basic unit stats. This would be balanced.
I could say that airplanes should cost 57x as much as comparable ground units, put wings on a Stumpy, and send 11,400 metal PteroStumps to their doom against normally priced Slashers. Would this PteroStump be effective? No. Would it be balanced? Yes.
If you haven't followed this line of reasoning so far, and if you can't draw your own accurate example based on those provided, if you can't answer, "Then what's the point of balance?" then you're not qualified for intelligent discourse, because you weren't reading when we defined the damned terms.
You're all idiots.
How much "right" and "wrong" is there in 2+2? Does popular opinion matter? Does wikiality apply? If 2/3 players say that 2+2=5, does 2+2 really =5?
Saying that I'm wrong without understanding my argument is a very clever approach towards frustrating me, but at the moment I'm all frustrated out. Try me again tomorrow - if I check the forums you might get better results.
For those who aren't deliberately baiting me, which I'm sadly left to presume is all of you, I can't understand how you can't follow the simple line of logic, which no one here's debated.
For a balanced equation, I could say that speed was worth 10x as much as Health. As such, it would take many times - perhaps 5x - as much metal invested in Flash Tanks as invested in a Sumo to kill a Sumo. In BA, that would equate to something like 10,000 metal in Flash Tanks. This would be balanced.
I could say that range was worth only 2% as much as dps, and redesign a Flea that could fire from one corner of a 16x16 map to an opposing corner and cost 200 metal without changing any other basic unit stats. This would be balanced.
I could say that airplanes should cost 57x as much as comparable ground units, put wings on a Stumpy, and send 11,400 metal PteroStumps to their doom against normally priced Slashers. Would this PteroStump be effective? No. Would it be balanced? Yes.
If you haven't followed this line of reasoning so far, and if you can't draw your own accurate example based on those provided, if you can't answer, "Then what's the point of balance?" then you're not qualified for intelligent discourse, because you weren't reading when we defined the damned terms.
You're all idiots.
Zpock: That is statitics and opinion polls, in short it's an empirical process, not the mathematicaly true formula that can be applied to kown how to perfectly balance a mod once and for all.
Beside, your method has some flaw. You assume that humans will naturally use the best units. This is taking humans for machine, and ignoring the whole human aspect. Player won't use the best units. They'll use the units which they think are the best. (And not even, a proportion of players significant enough to screw your stats are influenced by the looks, the fun using the unit provide, the ease of use, ...)
To know what units to use, player won't re-analyse the mod with a fresh eye each time. They'll go with what they used to use. So for instance if you buffed a previously underpowered unit, people will still avoid it, and so not notice until very long it has actually become a good unit. They'll do with what other player do, so you can have not-so-good units overused because it's trendy, while better units see no use because if no one use them, everybody think they must suck since no one use them.
If you look for instance at the TA scene, you'll see that despite having loads of player, and not any change ever in the unit properties, it took years and years for nowadays way of playing to settle.
Your process would work better if you used emotionless AI instead of humans, save that AI don't enough play like humans, they are more stupid, and can't invent trick like humans, they have such ubermicro not even a Korean can rival, etc...
So, in short, while I agree that you process is about the only way to balance a mod:
- It takes very long, if not an infinite time, to reach the balance.
- It's of no help to know how to balance beforehand new unit
- It's not a formula, but a procedure based on the polling of humans people opinion.
- It's not simple nor easy if it requires the use of a hundreds of player over years and years!
- Like pointed in the post above, even if that process actually reach a stable equilibrium (and does end up toggling between two if not more state, or has some units properties tending for infinity), you would have only reached balance for a particular set of condition. Change the nationality of your players, release new maps, have computer hardware improves, and the conditions change enough that your mod is imbalanced again. Tis not a view of spirit, in TA the brawler owned when their EMG had lag-attack, but stopped being used when peple started having faster CPU. Or in Core Contingency, urban maps makes mexx irrelevant but builder micro and metal storage all important.
Edit: Hmm, I see now you actually posted half-assed attempt at actual formulas. Each one is dead wrong:
I guess your exemple weren't supposed to be taken litteraly, yet I still showed you that no matter what your formula is, it can always be proved wrong.
Beside, your method has some flaw. You assume that humans will naturally use the best units. This is taking humans for machine, and ignoring the whole human aspect. Player won't use the best units. They'll use the units which they think are the best. (And not even, a proportion of players significant enough to screw your stats are influenced by the looks, the fun using the unit provide, the ease of use, ...)
To know what units to use, player won't re-analyse the mod with a fresh eye each time. They'll go with what they used to use. So for instance if you buffed a previously underpowered unit, people will still avoid it, and so not notice until very long it has actually become a good unit. They'll do with what other player do, so you can have not-so-good units overused because it's trendy, while better units see no use because if no one use them, everybody think they must suck since no one use them.
If you look for instance at the TA scene, you'll see that despite having loads of player, and not any change ever in the unit properties, it took years and years for nowadays way of playing to settle.
Your process would work better if you used emotionless AI instead of humans, save that AI don't enough play like humans, they are more stupid, and can't invent trick like humans, they have such ubermicro not even a Korean can rival, etc...
So, in short, while I agree that you process is about the only way to balance a mod:
- It takes very long, if not an infinite time, to reach the balance.
- It's of no help to know how to balance beforehand new unit
- It's not a formula, but a procedure based on the polling of humans people opinion.
- It's not simple nor easy if it requires the use of a hundreds of player over years and years!
- Like pointed in the post above, even if that process actually reach a stable equilibrium (and does end up toggling between two if not more state, or has some units properties tending for infinity), you would have only reached balance for a particular set of condition. Change the nationality of your players, release new maps, have computer hardware improves, and the conditions change enough that your mod is imbalanced again. Tis not a view of spirit, in TA the brawler owned when their EMG had lag-attack, but stopped being used when peple started having faster CPU. Or in Core Contingency, urban maps makes mexx irrelevant but builder micro and metal storage all important.
Tired here refer to a common formula where you output unit stat sheet on one head and it gives you perfect balance on the other end! (or rather, I guess his formula would take all the characteristic about a unit, and outputs a single value out of it, and that mod is balanced when that value is the same for all unit.)Anyway lurker, balance is easy easy easy to achieve, as I've pointed out here and in other threads. All it does is force all units through their common characteristics into a common formula where you can achieve a zero-sum. You could pull this formula out of a hat, and have a very dull game with perfect balance.
Yet Argh said many times already he agrees Nanoblobs was a balancing & gameplay failure, yet Argh's following post addresses none of these two points.Tired wrote:Argh's a case in point of this, but he's a flat out moron. Watch him defend his contention that NanoBlobs was anything other than complete ass, or that he's not a moron, in a post soon to follow this one if you have doubt.
WTF???? Theorically, at this point, I should stop talking to you, as are you clearly impervious to simple obvious facts.Tired wrote:Argh: *Responds, as predicted.*
No, you never made it clear that your line of defense against Argh was to let it him write overlong posts and ignore him (which wouldn't work anwyay, since when ignoring valid points make you appears more clever?). You pulled that after his second reply. However, you are an arrogant idiot for pretending to have flow-charted the human people you talk with. And I hate that breed of idiots, that writes correctly and pretends to know better than anyone while they're full of shit, because they're the most dangerous ones.Tired wrote:Flow Chart #2:
Tired: "Argh's a moron, so he'll respond to what I'm typing here even though I've made it pretty clear that I won't give him the time of day. In responding to this statement, he falls for a troll feint and proves himself a moron!"
Because calling people flat out moron, and their work complete ass, is every bit polite? Or maybe "that one response" refers to post out of this thread, but my point still stands, one post you all out insults, the other you pretend to be a polite personTired wrote:I should note that in this one response I was every bit as polite as Argh has even been on these forums, and probably a fair bit moreso.
Edit: Hmm, I see now you actually posted half-assed attempt at actual formulas. Each one is dead wrong:
Then buildings must have infinite health.I could say that speed was worth 10x as much as Health
I'll suppose you mean "range" not rage. This is not true, because if a turret unit has a range R, and another a range R+epsilon (where epsilon is very small compared to R), then they would be about equal, while in reality the turret with range R+epsilon is incomparably better than the turret with range R, as it outrange it and so win against it everytime, without the enemy able to retaliate, when properly placed.I could say that rage was worth only 2% as much as dps, and redesign a Flea that could fire from one corner of a 16x16 map to an opposing corner and cost 200 metal without changing any other basic unit stats.
Assuming a map with lots of flat free room (so collision issues are negligible), and a plane that move and turn as slowly as a ground unit, the plane worth would be equal to a ground unit worth. You fail again.I could say that airplanes should cost 57x as much as comparable ground units.
I guess your exemple weren't supposed to be taken litteraly, yet I still showed you that no matter what your formula is, it can always be proved wrong.
You're the only idiot.Tired wrote:You're all idiots.
In that case you just expand the data gathering to include hilly maps? Maps need to be balanced after the game,more the other way around also, for obvious reasons. Data gathering can be BTW anything, watching replays, some big database system, a few playtests with your buddy. I wanted to keep it simple but it should be obvious that in step 2 you might want to do some kind of deductive analysis to select appropriate balance changes (IE ask yourself why is this unit not used, etc).[Krogoth86] wrote: Your "Trial&Error" method might be a method you can use though it has two disadvantages imo:
1.)
It take helluva time as you might ping pong around the values you'll have in the end...
2.)
The method might lead to "wrong" results. It only works on an ideal, plain map. When playing on maps with lots of relief there might be a high demand for artillery or high-trajectory firing units which aren't that much prefered on other maps. So the Trial&Error Method fails to consider those facts too and you for example end up with a crippled down artillery which is of only little use on maps with few relief (and so you'll improve your units to find them OP on the hilly map again)...
Math is just a tool, and it's not the tools fault that your using it wrong. You still have to do the thinking. It's foolish to neglect it just because it doesn't give you the perfect answer right away.
Last edited by Zpock on 20 Nov 2007, 01:53, edited 1 time in total.
Tired Flow Chart:
Insult previous poster blandly.
Add racial slur.
Make pitiful attempt at angering by attacking whatever work of previous poster he can think about, no matter how irrelevant it is.
Adressing the damn point? No way, he's way above that!
Edit: bah, I was too long to reply, and the moderator was quicker.
Insult previous poster blandly.
Add racial slur.
Make pitiful attempt at angering by attacking whatever work of previous poster he can think about, no matter how irrelevant it is.
Adressing the damn point? No way, he's way above that!
Edit: bah, I was too long to reply, and the moderator was quicker.
Canadians are not a race. Learn to speak the language. 0oAdd racial slur.
Do you know anything at all about Accounting, Zpock?
Anyway, I'll quote myself, since asking people to read the thread that they're commenting on seems too much to ask.
"11. (mathematics chemistry) equality of elements in equation: a state in which the elements of a mathematical or chemical equation are equal on both sides."
http://encarta.msn.com/dictionary_18616 ... lance.html
Keep it simple, and repeat.Balance only impacts gameplay by introducing consistency.
I'm not canadian, and you attacked me on the ground I was a stranger. Anyway it's pointless to keep talking about a nuked post.Tired wrote:Canadians are not a race. Learn to speak the language. 0oAdd racial slur.
Do you know anything at all about Accounting, Zpock?
Strictly speaking there are no human races, so I guess racial slur never exist, duh?
-
[Krogoth86]
- Posts: 1176
- Joined: 23 Aug 2007, 19:46
Now let's stay with Flash, Sumo (the two you mentioned in a previous post), a third let's say Goliath-like unit and my AoE example. A specific amount of Flashes beat a Sumo. A specific amount of Goliath-style units kill a Sumo. Does the flash horde kill the Goliath-like units then (or at least nearly)? No. The reason: Because of its AoE it has a huge advantage against numerous units which pretty much kills your equations...Tired wrote:"11. (mathematics chemistry) equality of elements in equation: a state in which the elements of a mathematical or chemical equation are equal on both sides."
Or in other words:
x Flash = 1 dead Sumo (plus just one damaged flash alive)
y Goliath-style-units = 1 dead Sumo (plus just one damaged Goliath-unit alive)
As you are focused on maths it now also should be...
y Goliath-style-units = x killed Flash (with one Gollie-unit alive)
(or maybe the other way round)
... which isn't true...
This is an example that you can't apply simple equations on complex situations as some things are ambivalent and can't be measured in a single way what would be needed for a single balance system concerning of equations...
Krog, what is the point of light cavalry? You've played Medieval:TW or something. What purpose does light cavalry serve that heavy cavalry doesn't?
I'd go on to explain that a direct Health / dps relationship really does balance out the way you seem to want every single time all other things being equal, but just a simple question to start you off easy.
I'd go on to explain that a direct Health / dps relationship really does balance out the way you seem to want every single time all other things being equal, but just a simple question to start you off easy.
Krog I think you can still use the equations system, assigning a points system to each attribute that seems important like, area of effect. You would then balance the total score of each unit just as tired says. However choosing the weighs is the tricky part. And the system would not take into account synergies like between speed and range unless you complicate things a lot. This could probably come up with a crude balance, and then refined by changing the weighs instead of individual units values. This does have some merit and I could very well see why it would be a pretty efficient method for a TA mod with many units.
Thanks for reitiring in a bluntier wording your pointless attacks.Tired wrote:zwzsg, be less French.
To reformulate Krogoth point: Tired's balancing theories suffer from the same fundamental flaw that many modder around, including Caydr, fell into.
You are all assuming that the "worth" of unit can be summed up in a mono-dimensionnal value. That's there's an isomorphism between the "worth" state and the |R space. That there is a "<" comparaison operator valid in the "unit's worth" space.
Now, let's imagine a simple mod, with three units, rock, paper scissor. Now imagine that, as Tired implies (but without wording it properly because he cannot) that there is a function (the fabled "formula"), ValueOf, taking a unit as an argument, and outputing a real.
ValueOf("Rock")=r
ValueOf("Paper")=p
ValueOf("Scissor")=s
Units being true to their name, we have:
p>r
r>s
s>p
so, p>r and r>s => p>s
yet s>p
Oops!
Therefore our hypothesis are wrong. There is no such formula.
Formal proof you are wrong.
(I spared you the details on building a RPS mod.)
Also, proof that your mind is to small to even envision that there can multidimensionnal spaces, where no comparaison operator exists.
Last edited by zwzsg on 20 Nov 2007, 02:23, edited 1 time in total.