Faction ranking formula design

[ete]

Got the idea from this. Frankly, that formula is a horrific mess with chunky nonsensical values, and that leads to blatantly wrong results despite ridiculous fudging to try and correct it (e.g. Sharetech giving -3 points, just to try and make the Data Angles have a less insane score). But I'm confident that a fairly reliable formula can be built, which should give a reasonable estimate of each faction's power (though some parts, especially modeling the interaction between social models and innate bonuses, may require complexity to be accurate).

My current plan is to split the complexity into a number of attributes, input all the data on the original 14 to a spreadsheet, then play with the numbers a lot until it makes sense. Rough sketch:

• Builder
• • Research (Research, Energy, Bonus points per base, needs careful integration of info war) - How fast you're able to tech.
  • Infrastructure (Growth, Industry, Energy*Hurry, Terraform bonus, fungus resource bonuses, base size modifier) - How fast you're able to develop and expand. Maybe split into early and mid/late (when effic is more important)?
  • Population control (Drone, Talent, Police, WormPolice)
     
• Momentum
• • Combat (Morale, attack/def modifiers, Fanatic) - How strong your general troops are.
  • Army (Industry, Support, Energy*Hurry) - How large an army you can have.
  • Native (Planet, psy bonuses) - How strong/large your Psy army is
     
• Other/General
• • Energy (Econ, Effic, Council Votes (gov, so commerce), Interest)
  • Info war (Probe, sharetech, stealtech, Council Votes (gov, so infiltrate))
  • Subversive (Probe, Probe action cost, Energy)
  • Army (Industry, Support)
  • Native (Planet, psy bonuses)
  • Others (Mind Control immunity, Aquatic, Alien, Free Prototype, Bonus Commlink, Revolt, Council Votes)

Bonus Facilities, Techs, Units, and Abilities will take a lot of tweaking to get right, they'll probably be added mostly on-demand. Most of the bonuses should be stacked as multiplers rather than just adding them up. Anti-ideology is complex, as is Immunity/Robustness/Impunity to a lesser extent, since they're all dependent on the SE screen. I'll expect the Hive to be underrated until this gets fixed.


So.. comments? Encouragement? Offers to help with data entry would be great too, I'll be using google docs.

[Green1]

Basically a Fallen Enchantress style point buy system to balance custom factions once and for all.

(Fallen Enchantress has MUCH in common with AC for a fantasy game - barring tactics/rpg elements)

Combine it with a more modernized ACEdit that enforces this, you may be onto something.

[ete]

It'd be a useful guideline hopefully, but we'd probably be continually improving it because there will be loopholes and specific combinations which are more powerful than I'd have realised. If someone tried they'd probably be able to find plenty of exploits, but hopefully people will point them out and I'll be able to close them up until few are left. I'd not encourage adding it as enforced in any program, though perhaps it could be a displayed value.

[ete]

Okay, requesting help adding the data about each official faction to the spreadsheet here: https://docs.google.com/spreadsheet/ccc?key=0AhADYlhLMFj-dDlVZUdOZ3l4TEZJRnJWRnBLUWNDLVE#gid=0

[Green1]

Some of these guys have more than one anti ideology for AI/ diplomacy... but they do have only one they can not take.

[ete]

The anti-ideology is the one they can't take (look in the text files). They just get annoyed if you pick something OTHER than their ideology even if it's not their specified anti-ideology.

[ete]

Ok, we've got all the original factions in. Impressive how many bonuses are used by only one or even no faction, really leaves room for exploration. Will make creating the formula.. probably more hit and miss, but easier. Also.. so many variables. damn.

I think I'll include some extra variables in the formula at some point, for stage of the game where conflict occurs and native life. A lot of bonuses are cool right near the start, but long term do little. And maybe something for sea?

[Green1]

This is where Frogboy and Kael's (famous CFC modder) system they used for Fallen Enchantress comes in handy.

You have merits and flaws which have a point value.

Merit example

Plus on any SE table. Maybe point costs go up exponentially for +. ie: +1 Eff =x points. +2 Eff = 2 points. +3 eff = 6 pts.
free ability
free unit
"aquatic" package
free tech past first
and so on

Flaw example

disallowed SE choice
minus on any SE table, possibly allowing more points for each minus
minus to any research/ cannot research till certain mission year

After all the merits and flaws coded into SMAX are itemized and have value, you will have a good idea if something is truly OP.

[ete]

The issue with a simplistic addition system is that it takes absolutely no account of synergy. Some bonuses may be mild on their own, but when combined with another similarly mild bonus create a much larger advantage than the sum of the parts. And for SE settings.. well, they're complicated, and I'm posting another topic to deal with them.

[Green1]

Somewhat true.

But..  somethings balance out.

One Let's Play by GrimthR, there was a custom faction called The Lovers he had designed. Supposedly it was overpowered due to +4 Growth because the society was supposedly based on wild sex and rampant population growth. Pop booming insanity...

Did not work out that way due to extreme drone issues and nutrient limits.

Even the faction he always played - some faction based on a cult to him that had + POLICE and impunity to Fundy had severe issues at times.

But that is the reason some stats would count off less or more than others.

[ete]

I'm confident that bunching characteristics which contribute to the same area of strength and intelligently using multiplication will lead to a much more accurate measure of power than primarily using addition for a game as complex as SMAX. Perhaps one particular focus has flaws, but without a formula which at least attempts to model the synergy it's impossible to see things working anything better than terribly.

[ete]

Ok, I've got alpha formulas for: Combat, Psy, Energy, Research, Info war offense, Info war defense, Subversion offense. You can see what values it's currently producing here: https://docs.google.com/spreadsheet/ccc?key=0AhADYlhLMFj-dDlVZUdOZ3l4TEZJRnJWRnBLUWNDLVE#gid=1

There's likely to be issues, especially with Energy since that has so many possible factors and depends significantly on game stage. Still, progress. Next part after Subversion defense is to integrate Research and Info offense/defense into something which should indicate how a faction's position on the tech leaderboard should look. Then maybe I'll have a go at Native or some army modifiers.

[ete]

Got distracted trying to sort out social engineering since that's pretty important in many places. Going to take some effort to make it easy to read by my other formulas, maybe won't get back to it till after the weekend but it'll be worth it and greatly improve the accuracy of my numbers.

[Green1]

Got distracted trying to sort out social engineering since that's pretty important in many places. Going to take some effort to make it easy to read by my other formulas, maybe won't get back to it till after the weekend but it'll be worth it and greatly improve the accuracy of my numbers.

No problem. It is understandable.

Social engineering is a complex and fascinating topic that deserves scrutiny.

[ete]

I've implemented social engineering in a basic form. Current plan is to assume different SE settings for four stages of the game: Very early (no SE, or ones with starting/extremely early tech), early (most basic SE available), mid (all basic SE available), and late (all+future society), and pick one standard set of settings for each faction (in most cases this should be doable, since almost everyone runs wealth/demo (sometimes Knowl), and economics is largely dependent on stage of the game). Could possibly be two per stage, but then there will be two sets of values for the faction.

What exceptions to the following base principle are there (barring factions which can't)?
Very early: whatever's available
Early: Demo/Planned/Wealth (knowl if no wealth)
Mid: Demo/FM (which run Green?)/Wealth
Late: Demo/Green or FM/Wealth/Cyber or Eudaimonic

Early is meant to be actually early, before all the basic SE is researched probably. Is FM a good idea over growth/industry then, or wait for anti drone stuff?