Mathmatical Balence Model?
-
holtonster
- Posts: 55
- Joined: Mon Apr 23, 2007 7:42 am
- Location: San Leandro, California
II am just ranting here, but Alleg might be a bit too much to model according to signal/system theory. Maybe try the soft computing approach instead?
"We are all atheists about most of the gods that humanity has ever believed in. Some of us just go one god further." - Richard Dawkins
"Patriotism is the virtue of the vicious." - Oscar Wilde
"Patriotism is the virtue of the vicious." - Oscar Wilde
Snack wrote:QUOTE (Snack @ Jul 8 2007, 03:50 PM) II am just ranting here, but Alleg might be a bit too much to model according to signal/system theory. Maybe try the soft computing approach instead?
Hi Snack,
Can you tell me more about this? I'm no math genious, just know enough to be a danger to my check book.
Actually this iz a way wrong way to think of alleg BALANCE.... U should think of it the way I do when making a new faction..... U should balance out all strengthz with other weaknessez... AKA Giga'z MAD money and LUX shipz.. Balanced BY weak Hull etc....Dreg MADD speed..balanced BY large hitboxxez....etc..etc...BIOS superior tech..balanced BY huge research time... ITZ not math that will balance out and keep to the fun of the game.....ITZ logical and Imaginative thinking that will work in the end....To bad no haz even attempted to enlist my help on this core.. Az I think I really have a great grasp on how to balance and Maintain the underlying fun of alleg at the same time
[img]Try Dungeonz and Dragonz online..itz Free
Well the theory of signals and systems can probably be found in MIT "open source" curriculum as it can be found on most comp/electrical science universities. Basically, you have bunch of these differential equations that represent the system you are modeling and it's a pain (well for me at least /mrgreen.gif" style="vertical-align:middle" emoid=":D" border="0" alt="mrgreen.gif" /> ), to do it properly. On the other hand, you can model the system using some soft computing method like fuzzy logic which will eventually alow you to describe the system in more human terms and most important, use simpler calculations while giving you a good result.
Anyway, if you want to model the Allegiance as some kind of math system, you really need to be familiar with that math stuff and invest a lot of time and effort in doing it. I just was intrigued to see how someone would actually do that with Alleg, since IMHO it would require a lot of theoretical and practical knowledge and a lot of math, so I joined the thread. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
I would really like to see what you can come up with in a sort of a very simplified Alleg model for start. So maybe go from there and see what happens. I welcome the creative effort. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
Anyway, if you want to model the Allegiance as some kind of math system, you really need to be familiar with that math stuff and invest a lot of time and effort in doing it. I just was intrigued to see how someone would actually do that with Alleg, since IMHO it would require a lot of theoretical and practical knowledge and a lot of math, so I joined the thread. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
I would really like to see what you can come up with in a sort of a very simplified Alleg model for start. So maybe go from there and see what happens. I welcome the creative effort. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
Last edited by Snack on Mon Jul 09, 2007 10:38 am, edited 1 time in total.
"We are all atheists about most of the gods that humanity has ever believed in. Some of us just go one god further." - Richard Dawkins
"Patriotism is the virtue of the vicious." - Oscar Wilde
"Patriotism is the virtue of the vicious." - Oscar Wilde
-
AaronMoore
- Posts: 471
- Joined: Sun May 06, 2007 3:09 pm
- Location: Australia
Hey guys,
On the concept of "soft" computing, I thought I would jump in. I also have been thinking about balancing the core/creating a new faction etc.
My thought was to simplify the core so that all factions have equal and equivilant ships as a starting point, then re-introduce the faction specific ships, and improved modifiers one-at-a-time, with a discussion/poll etc about the balance/tradeoffs.
Anyway back to the point:
I have done Genetic Algorithms as a project at Uni, and I was thinking I would try my hand at writing something that would read a core, play many hundreds of simulated games (read: problem=Commander AI) and use the process of natural selection to adjust the core to the desired state of 50% wins for each faction.
I know this may/may not just bring the factions totally inline, but hopefully, it will take into account the additional ship types etc. and would allow us to focus on creating new factions and ship types and let the process of natural selection balance it for us...
Just throwing it out there!
/blink.gif" style="vertical-align:middle" emoid=":o" border="0" alt="blink.gif" />
On the concept of "soft" computing, I thought I would jump in. I also have been thinking about balancing the core/creating a new faction etc.
My thought was to simplify the core so that all factions have equal and equivilant ships as a starting point, then re-introduce the faction specific ships, and improved modifiers one-at-a-time, with a discussion/poll etc about the balance/tradeoffs.
Anyway back to the point:
I have done Genetic Algorithms as a project at Uni, and I was thinking I would try my hand at writing something that would read a core, play many hundreds of simulated games (read: problem=Commander AI) and use the process of natural selection to adjust the core to the desired state of 50% wins for each faction.
I know this may/may not just bring the factions totally inline, but hopefully, it will take into account the additional ship types etc. and would allow us to focus on creating new factions and ship types and let the process of natural selection balance it for us...
Just throwing it out there!
/blink.gif" style="vertical-align:middle" emoid=":o" border="0" alt="blink.gif" />
What would your chromosomes and population look like? How would you represent all these Allegiance variables with GA? Seems like a really tough job to me. Then again, I am no expert. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
On ACE forums, I was toying with the idea of using GA to find (optimize) the best (well in stohastic terms) Finite State Machine algorithm for Miner AI. That seems doable to me with my limited knowledge. However, modeling entire Allegiance seems like incredibly complicated task to me. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
On ACE forums, I was toying with the idea of using GA to find (optimize) the best (well in stohastic terms) Finite State Machine algorithm for Miner AI. That seems doable to me with my limited knowledge. However, modeling entire Allegiance seems like incredibly complicated task to me. /smile.gif" style="vertical-align:middle" emoid=":)" border="0" alt="smile.gif" />
"We are all atheists about most of the gods that humanity has ever believed in. Some of us just go one god further." - Richard Dawkins
"Patriotism is the virtue of the vicious." - Oscar Wilde
"Patriotism is the virtue of the vicious." - Oscar Wilde
-
AaronMoore
- Posts: 471
- Joined: Sun May 06, 2007 3:09 pm
- Location: Australia
Thanks Snack,
The chromosomes would be made up of the balancing factors of the core, so all the hundreds/thousands of numbers and modifiers for the factions/ships/bases/cost for the core. I know this is massive and there is a huge chance of producing something that may be "balanced" according to whatever the performance funtion is (50% wins, and game length), but this will probably produce something that is drastically different to what every1 is used to (will Bios reasearch time still be so long etc.)
This is not a good thing, and we need to keep the fundimentals of what we know and love about allegiance constant, so possible only a select number of these modifiers will be avaliable for mutation.
You're right, for this to work, I not only need to create a commander AI, but also simulate the entire allegiance game (reasearch, mining, combat) so will be very complicated, nearing impossible.
However if we simplify everything (teams of constant size, same simple combat ai/formula, etc.) we will just be left with simulating the mechanics of allegiance and the commander ai, and the balance of the core will have a direct impact on the game result and length.
I'm going to need to think this through a lot more, i know .....
/unsure.gif" style="vertical-align:middle" emoid=":unsure:" border="0" alt="unsure.gif" />
The chromosomes would be made up of the balancing factors of the core, so all the hundreds/thousands of numbers and modifiers for the factions/ships/bases/cost for the core. I know this is massive and there is a huge chance of producing something that may be "balanced" according to whatever the performance funtion is (50% wins, and game length), but this will probably produce something that is drastically different to what every1 is used to (will Bios reasearch time still be so long etc.)
This is not a good thing, and we need to keep the fundimentals of what we know and love about allegiance constant, so possible only a select number of these modifiers will be avaliable for mutation.
You're right, for this to work, I not only need to create a commander AI, but also simulate the entire allegiance game (reasearch, mining, combat) so will be very complicated, nearing impossible.
However if we simplify everything (teams of constant size, same simple combat ai/formula, etc.) we will just be left with simulating the mechanics of allegiance and the commander ai, and the balance of the core will have a direct impact on the game result and length.
I'm going to need to think this through a lot more, i know .....
/unsure.gif" style="vertical-align:middle" emoid=":unsure:" border="0" alt="unsure.gif" />
Last edited by AaronMoore on Mon Jul 09, 2007 11:30 am, edited 1 time in total.
that approach seems quite cool, though i'm just wondering how long it would take?
i mean, the number of variables, doing it for each faction vs each faction, it may end up you set it off running, and its estimated completion time is in roughly 563 years time...
perhaps distributed computing ... boinc?
i mean, the number of variables, doing it for each faction vs each faction, it may end up you set it off running, and its estimated completion time is in roughly 563 years time...
perhaps distributed computing ... boinc?
