Social Engineering Mod

From Alpha Centauri Wiki
Revision as of 19:20, 29 January 2017 by Tnevolin (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Template:Header Nav

Contents

Preface

This is an attempt to enhance and balance social engineering choices. Thanks to all contributors on many other forums for expressing their opinions and sharing ideas.

Goals

Adjust efects range usage

In vanila version some combinations of models bring effect values beyond allowed range thus wasting it and diminishing model value in general. The opposite is also true. Some effect values are unreachable by any means. Adjusting this alone will bring more variety to the game.

Adjust models appeal

I agree with multiple posters that some models are inferior by all means and never used. This too cripples game variety. I'd like to adjust their value to let it be chosen not necessarily often but at least sometimes when time is ripe. Thus bringing additional twist into the game by using means already encorporated into game mechanics.

Adjust extra sharp effects

Some effects of social engineering could be so good that it brakes the game. I am specifically talking about population boom. With democracy + planned + CC it comes too early in the game ruining the rest of the game experience.

Resulting configuration

Here is my resulting configuration. The rest of the article is just a rationalization for those interested in academics.

<work in progress>

Effect weights

Effect weight is a relative effect value comparing to others. Understanding effect weights allows easy social models evaluation and comparison to each other. Still such estimate is a very rough approximation other people may argue with. I used multiple assumptions in below calculations. Feel free to correct them. I will appreciate any input. Thank you.

Comparison method

Obviously different effects have different impact on your empire. For the sake of comparison I try to emulate the same outcome for two effect and see how much of each I need to achieve it. In other words, how strong one effect should be to compensate another effect weakness. Then I average my comparisons by historical period, average city, the effect frequency and applicability. That gives me some insight about relative effect values.

Historical periods and city yelds

I consider three historical periods for purpose of averaging. Latest part of the game is excluded as at that point game is either won or lost and no social engineering can help it.

Early game

  • Technology: restrictions are not lifted
  • Number of cities: 6
  • Distance to HQ: 5
  • City size: 3
  • Terraforming: minimal
  • City yield: 8-6-6
  • Production bonus: none
  • Eco/Phy/Res bonus: none
  • Small army, occasional conflicts

Middle game

  • Technology: restriction are lifted
  • Number of cities: 12
  • Distance to HQ: 10
  • City size: 5
  • Terraforming: partial heavy terraforming
  • City yield 14-12-16
  • Production bonus: none
  • Eco/Phy/Res bonus: +50%
  • Medium army, occasional conflicts

Late game

  • Technology: Advanced Ecological Engineering
  • Number of cities: 18
  • Distance to HQ: 15
  • City size: 7
  • Terraforming: heavy terraforming
  • City yield: 20-20-30
  • Production bonus: +50%
  • Eco/Phy/Res bonus: +100%
  • Medium army, permanent conflicts

Effects weight calculation

Industry

I take this effect as a base because it is linear and affects whole empire all the time.

Weight = 1

Research

Industry to Research:
+1 Industry => 10% production increase => stockpile excess production => reallocate excess economy to research
By historical periods: 10%, 5%, 5% = 7% average

Research to Industry:
+1 Research => 10% research increase => reallocate excess research to economy => convert excess economy to production for 1/2 cost => production increase
By historical periods: 2.5%, 5%, 5% = 4% average

Weight = 0.5

Economy

Economy effect is drastically non linear. Let's try to estimate how energy flow is affected by effect change on average.

Averaging by effect value change and by periods:

  • -3 to -2: add 1 energy/base = by periods (15%, 6%, 3%) = 8% average
  • -2 to -1: add 1 energy/base = by periods (15%, 6%, 3%) = 8% average
  • -1 to 0: add 1 energy at HQ = 0%
  • 0 to +1: same as first one: 8% average
  • +1 to +2: jump from 1 energy/base to 1 enerty/square = by periods: (67%-15%, 38%-6%, 27%-3%) = average 36%
  • +2 to +3: add 2 energy/base and 1 commerce = by periods (30%+5%, 12%+10%, 6%+15%) = average 26%
  • +2 to +3: add 2 energy/base and 1 commerce = by periods (30%+5%, 12%+10%, 6%+15%) = average 26%
  • +2 to +3: add 1 commerce = by periods (5%, 10%, 15%) = average 10%

Therefore, +1 Economy is equivalent to 15% city energy yield increase on average. That is if whole range of available values would be used which is not the case in vanila version. The most popular and most used combinations before future society is FM + Wealth. Which makes transitions 0 to +2 and +2 to +3 the most popular choices during the important part of the game. Look at the finer tuned weights table below. Both these transitions has Economy weight 2.5. You can use that if you stick to vanila models. With more dispersed distribution of Economy modificators across SESocial Engineering models the weight will be close to average 1.5 as calculated below.

Economy to Industry:
+1 Economy => 15% energy increase => allocate excess to economy => 30% economy allocation increase => multiply by bonus => rush production for 1/2 cost => production increase
By historical periods: 7.5%, 15%, 15% = 13% average

Economy to Research:
+1 Economy => 15% energy increase => allocate excess to research => 30% research allocation increase => 30% research increase

Weight = 1.5
Very nice match by both convertion calculations!!

For balance freaks we can set finer tuned weights for different steps on Economy scale since this is the only effect that is unlinear in the middle and not on the extremities.


Economy scale step weights
From To Weight
-3 -2 1.0
-2 -1 1.0
-1 0 0.0
0 +1 1.0
+1 +2 4.0
+2 +3 2.5
+3 +4 2.5
+4 +5 1.0

Energy

Based on above Economy calculations we can introduce simplified Energy effect as a linear substitute for non linear Economy effect. This effect doesn't exist in a game but we can use it for ease of calculations. One step on Energy effect scale changes raw energy yield by 10%. Miraculously this effect has a weight of exactly 1 as Economy has weight of 1.5 and one step on Economy scale changes energy yeld by 15% on average.

Weight = 1

Efficiency

Efficiency effect is threefold. It descreases inefficiency, decreases number of b-drones, and decreases loss of uneven Eco/Res allocation. Let's consider them one by one.

Inefficiency formula tells us that increasing efficiency by 1 decreases inefficiency percentage on base by approximately D/128, where D is a distance to HQ.
By historical periods that gives us 4%, 8%, 12% = 8% average energy yield increase which is equvalent of 0.8 Energy effect rate.

B-drones formula tells us that increasing efficiency by 1 decreases number of b-drones by N/24, where N is number of cities. Let's calculate the phych allocation required to quell these additional drones. This way we can compare how increase in Efficiency compensates decrease in Economy and vice versa.
By historical periods that gives us 8%, 4%, 3% = 5% average energy allocation which is equivalent of 0.5 Energy effect rate.

Finally, the penalty for unequal economy/research allocation. Of course, the absolute effect depends on unequality percentage. Let's take a smallest unequal distribution of 60%-40%. For that distribution you loose 2% per Efficiency which is equivalent of 0.2 Energy effect rate.

All together it equals to 0.8 + 0.5 + 0.2 = 1.5 of Energy rate.

Weight = 1.5

Growth

Converting Growth to Industry or Economy gives pretty measly values because 10% growth speed increase is equivalent of 10% nutrient surplus increase only. Comparing to total nutrient production it is only about 2%. Reassigning workes to get 2% minerals instead is economically unnoticeable. Growth power lies not in immediate economical effect but in economical development acceleration instead. Roughly you'll have 10% bigger sities with 10% growth rate bonus. That gives you 10% more workers => 10% more of both minerals and energy. That is +1 on Industry scale and +1 on Energy scale totalling at weighed value of 2.
Skipping early population boom topic here as this breaks this effect completely and makes it invaluable.

Weight = 2

Support

Support is directly translated to minerals. Roughly each step on Support scale frees one mineral/base except maybe highest values when not every base can exaust free unit limit.
By historical periods: 15%, 8%, 3% = 9% average mineral increase.
Clean units bypass support limit thus diminishing this effect weight a little. Yet it is still clearly comparable to Industry effect.

Weight = 1

Police

Police effect is simialr to Efficiency b-drones effect and to Phych energy allocation effect. Both reduce number of drones. Let's convert Police rating to Economy rating using Efficiency b-drones calculation approach.

Police to Economy:
Police scale is more or less linear one. With one wandering unit per base it is approximately one induced/quelled drone per step on Police rating scale. Effect increases with more wandering units and with bigger bases. +1 Police => roughly -1 drone => equivalent of phych allocation or pacifying facility cost + maintenance (whichever is cheaper). At early game stages and in smaller bases it is cheaper to build pacifying facility. Recreation commons, for example, quells 2 drones and costs 40 minerals + 1 energy/turn maintenance. Roughly it is about 1 energy/drone ratio.
By historical periods: 16%, 8%, 3% = 9% average energy allocation which is equivalent of 0.9 Energy effect rate.

Weight = 1

Morale

This one is difficult to compare with others as it manifests itself only at war time and only when fighting tough opponents. That is when your fighting odds are around 50% so slightest morale boost really matters. That 1-2 moral boost significantly cuts your casualties and saves you production. Now the proportion of minerals spent on army and the moral effect is highly dependent on game situation and playing style. In my experience the Morale effect is invaluable at early game stages when armies are small and winning or losing single tactical combat turns the tide of whole war. It dissolves to the end of the game when you have plenty of morale enhacement facilities and winning is based on empire productive strength rather then tactical one.

Weight = 0.5 (subject for discussion)

Planet

One of the most usefult Planet effects is an ability to capture worms and slight battle odds improvement. For starter, +1 Planet lets you amassing huge worm army at the early game. Again, same as Morale it dissolves with Centaury Empathy which lets you producing worms at higher rate.

Weight = 0.5 (subject for discussion)

Probe

This is IMHO the least usefult effect of them all. Mostly because it works so rarely in the game that doesn't have major impact on outcome. The only exclusion is +3 Probe that makes your cities and units immune. And even then it makes no difference.

Weight = 0 (subject for discussion)

Effects weight table

Sorted by weight

Effect Weight
Energy (artificial effect not existing in game) 1.0
Economy (perfect average) 1.5
Economy (vanila version) 2.5
Growth 2.0
Efficiency 1.5
Support 1.0
Police 1.0
Industry 1.0
Morale 0.5
Planet 0.5
Research 0.5
Probe 0.0

Effect range usage analyzis

Analyzing how effect is used and whether its usage need to be adjusted. I.e. add more negative/positive modificators or use more rare/often.

Economy

Values below 0 are not reachable by common means. Values -1, -2 are reachable by Hive only.
Proposition: add some negative values somewhere.

Efficiency

Effect continues to work above value +4 even though manual describes value range -4 to +4 only. This is because all effects are calculated by formula that can accept any value.
Strongly unlinear toward -4 value. Stepping from -3 to -4 takes away 50% of your energy yield and makes all citizens drones! Whereas stepping from +3 to +4 changes same values by about 1%. Other than that works fine.
Proposition: disperse -2 modificators from PS and Planned to other models so they still can be used in large empires.

Support

Can go below and above value range which has no effect. No issue.

Morale

Can go above value range which has no effect. No biggie.
Worse thing is that besides SE models morale is also boosted by facilities and by promotion. As a result war oriented factions soon run swarms of elite units.
Proposition: decrease Morale modifier and give more negative Morale value across models so many combinations would push Morale down.

Police

Can go waaay below and above value range which has no effect. No issue.

Growth

Generally skewed toward higher values which allows early pop boom.
Proposition: decrease positive modificators, add more negative modificators.

Planet

Can go up to +4 which still has effect on combat odds. No issue.

Probe

Can go up to +6 which has no effect. No issue.
Proposition: use more rare and use more negative values.

Industry

A little skewed toward higher values. No issue.
Proposition: add more negative values.

Research

A little skewed toward higher values. No issue.
Proposition: add more negative values.

models analyzis

Analyze models popularity and discuss fix.

Police State

Slightly unpopular due to big Efficiency penalty.
Proposition: decrease Efficiency penalty, increase Police bonus

Democratic

Overpowered due to big Efficiency and Growth bonuses combination.
Proposition: decrease Efficiency and/or Growth bonuses, maybe add some lightweight effect like Research

Fundamentalist

Unpopular due to lack of good bonuses.
Proposition: add hefty Production bonus and/or morale

Free Market

No major issues since Economy is balanced by Police making it usable during peace time only. Plus you need extended worm protection for cities.

Planned

No major issues.

Green

Apparently slightly weakeash model judging by other people feedback.
Proposition: add Police bonus

Power

Another slightly weakeash model judging by other people feedback. Lack of serious values.
Proposition: add Probe and/or Research

Knowledge

No issues.

Wealth

No issues in general except that -2 Morale is not always a major flaw.
Proposition: add another -2 Morale and/or hinder Research

Cybernetic

No issues.

Eudaimonic

Insanely overpowered. Each bonus is a huge boost by itself and Morale in later game means nothing.
Proposition: decrease or remove bonuses, cut on Research

Thought Control

Too weak comparing to Eudamonic. Police is the only valuable bonus.
Proposition: actually maybe swap some bonuses with Eudamonic? That would fix them both. Need to think.