Alpha Centauri 2

Sid Meier's Alpha Centauri & Alien Crossfire => The Theory of Everything => Topic started by: ete on January 25, 2013, 01:50:33 am

Title: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 25, 2013, 01:50:33 am
As part of the aim of creating a faction rating formula, I need numeric wieghts to the benefits and disadvantages of each value for each social engineering characteristic. Simple formulas do not match up to the complex effects, for example +1 Economy is nice, but +2 Economy is many times more useful, then +3 is only somewhat more useful.

To simulate this, I need your help. Please read the instructions carefully:
Rate each value of each social parameter between 10 and -10, with ten being the best, zero the middle, and -10 the worst. For the Economy example I'm using, I would put the numbers like:
Econ Rating
0   0
+1   1.5
+2   7
+3   8
+4   9
+5   10

This indicated that a rating of +2 is several times better than that of +1.

Note that a scale should only compare with other values of the same social parameter, can scale differently scale on each side of zero! I will deal with this by scaling the positive and negative parameters separately in a different part of the equation. This means that +10 on a value is not as good as -10 is bad, same goes for other values.  It is also correct for +1 Econ to have a lower value than, say, +1 Probe, because +1 Probe is worth a larger % of the bonus of the maximum probe score than +1 Econ is worth compared to maximum Econ score.

I've made a spreadsheet with all values of all parameters which have an effect (it's actually a pretty neat resource in itself, and will be more with some tidying). Please go to it and enter the effect weighting as described above. Go check it out! (https://docs.google.com/spreadsheet/ccc?key=0AhADYlhLMFj-dGZmNGwtWmZBZnZQMGEyNjNiRUhnTXc)


And if anyone has any questions/discussion/justification post away.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 02:06:33 am
Some of these scores grant "merits" and "flaws". That is where a wieght is becomes more than just a formula.

For example,

+1 PLANET unlocks ability "capture native unit"
+3 PROBE grants immunity to standard probe team subversion.

+3 Probe would be worth much more in comparison than +1 Probe to +2 Probe.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 25, 2013, 02:14:48 am
Which is exactly why I'm not using a formula, it's too complex for that. I'm asking for subjective weightings for the importance of those merits and flaws, each scaled to only their own parameter. Your kind of examples are precisely why I made this, and are similar to my Econ one.

With +3 Probe being max (so must be +10 weight), you'd then want to rank +2 Probe much lower if you feel it is much less valuable. If you feel that the number of mindworms captured with +1 Planet is of similar value to that at higher Planet ratings, then rate +1 Planet highly, etc..
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 02:57:26 am
Easy. I come from the old school DnD pencil and paper way of looking at stuff like this. To quote us old school United States pencil and paper gamers, "we dont need no stinkin spreadsheets"... (it's a con saying lol)l Although what we did is cool.

You do have a formula for each SE stat that goes up and down mathmatically. A negative number to a positive number. Even Dungeons and Dragons, for example, is a formula and tournaments use a point buy system.

So... I propose

setting | points
-3  ------   |7
-2 -------   |3
-1 ---------|1
0-----------|0
1 --------- |-1
2 --------- |-3
3 --------- |-7
4 --------- |-13

merit
-----------
Ability: Capture Native     -5
Immunity: regular probe team -5

those are the only two abilities I see.

formula :

SCORE = (SEVALUE * 2 +1) + Merit 
If SEVALUE (PLANET) >1 then SCORE = SCORE + 5
If SEVALUE (PROBE)>2 then SCORE = SCORE +5

Of course those numbers are up for debate.

I needa beer. You guys are making me use my education. Too bad it was all in a New Orleans bar.




Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 25, 2013, 03:08:29 am
Would kind of work for some things, but does not adequately deal with the complexities in many cases. For example, making higher ratings count for much more is at odds with how efficiency works. The first few effic points are extremely important, but they become less important as you move up away from zero. Econ jumps in big and uneven steps, some giving massive bonuses, some just helping a bit. Police is similarly jumpy.

Basically, I want people to weigh up those merits and flaws, all of them, and give me their opinion. Using a formula to help with your set of results is fine, but other than Research and maybe Industry I think the opinions of players is likely to be closer to the true values than any remotely simple formula.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 03:28:06 am
Problem is, the system has to be somewhat simple or it will be hard to follow.

Now, not all SE choices are equal, but with some builds it is powerful.

In RPG terms, someone could use what we call a dump stat to mathematically get around in a point buy system. What that is is that in tournaments they would take negatives in something that did not matter to make a more important stat uber. In Dungeons and Dragons 3rd edition and greater, that stat was Charisma except for certain builds that required charisma. Or it was Intelligence.

No matter what equation, we will not be able to stop this.

In the AC world, POLICE would be a good dump stat. I could see a min max power gamer show up with his custom faction. BUT - looking at our spreadsheet, that is balanced too. Having negative 2 POLICE gives the flaw : Can not use Nerve Stapling or units as Police. Pretty bad unless you were to combine it with Free Facility: Rec Commons or something like that.

The guys at Firaxis had a decent design. All of them scale. Just add for the merit or boon that kicks in.

Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Earthmichael on January 25, 2013, 04:17:27 am
One flaw I see is using a +10 to -10 system.  Some of the negatives are far worse than the positives for some social stats, and for other social stats, the positives are far more important than the negatives.  For example, low Police ratings that starts generating drones for any aircraft or out-of-area military units is HUGE.

So I think we need to scale according to the biggest bonus/drawback.  That may mean we have a -10 to +6 rating for some area, or maybe a -4 to +10 rating for some other area.

Science is going to be hard to figure.  The percentages are only a small part of the total bonus/penalty for the research rating.  The reason I say that is that the percentage affects what gets accumulated, BUT it also affect how many points must be accumulated for a breakthrough.  Does anyone know how to figure out how to how the rating affects the required points for a breakthrough?
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 04:24:34 am
Yeah...

Had to fit my formula into where it would fall between 10 and -10.

But, I think that it is pretty fair assessment from the SP side of things.

I am sure it can be ripped apart or someone much more experienced than me will find fault with it.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 04:27:33 am
Yeah..I was ending up with - 25s and +25s according to my metrics to that scale..lol

Till an anon user put a red box saying "10". Then I had to say I knew.. but was having to adjust.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 25, 2013, 11:19:14 am
Problem is, the system has to be somewhat simple or it will be hard to follow.
That is not an issue. I will create a spreadsheet and wiki template which auto calculate everything, no one using the formula needs to understand or follow it unless they're trying to improve it. A complex formula is not a notable disadvantage, and large amounts of complexity are fine if it improves the accuracy and reliability of the formula.

Quote
Now, not all SE choices are equal, but with some builds it is powerful.

In RPG terms, someone could use what we call a dump stat to mathematically get around in a point buy system. What that is is that in tournaments they would take negatives in something that did not matter to make a more important stat uber. In Dungeons and Dragons 3rd edition and greater, that stat was Charisma except for certain builds that required charisma. Or it was Intelligence.

No matter what equation, we will not be able to stop this.
Yes, we will. Only simple additive systems are unable to prevent this, correctly scaled disadvantages with synergistic info added should be near impossible to significantly game with a dump stat. This is one of the main reasons I'm not using a primarily additive system.

Have a go flipping back and forth between your sheet and mine to see where I disagree. I've added a load of notes/justifications to my sheet.

One flaw I see is using a +10 to -10 system.  Some of the negatives are far worse than the positives for some social stats, and for other social stats, the positives are far more important than the negatives.  For example, low Police ratings that starts generating drones for any aircraft or out-of-area military units is HUGE.

So I think we need to scale according to the biggest bonus/drawback.  That may mean we have a -10 to +6 rating for some area, or maybe a -4 to +10 rating for some other area.
As stated in the opening post, I will handle that scaling in a separate part of the equation. There will be one scaling for above zero and one for below. All I'm looking for here is the importance of each negative rating relative to other negative ratings of the same parameter, and the importance of each positive rating relative to other positive ratings of the same parameter.

Quote
Science is going to be hard to figure.  The percentages are only a small part of the total bonus/penalty for the research rating.  The reason I say that is that the percentage affects what gets accumulated, BUT it also affect how many points must be accumulated for a breakthrough.  Does anyone know how to figure out how to how the rating affects the required points for a breakthrough?
Ah, that is a good point which I had not considered. If someone could find the formula for that it would be excellent.


I've added all the values to my sheet along with justifications, comments on them would be much appreciated.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 25, 2013, 01:18:20 pm
I want to get on with the formula, so I'm using my values for now. It's easy to update them with changes based on suggestions since they propagate.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 02:08:11 pm
Once we get done with this, I am interested in the relative power rating of these factions:

Buster's Uncle's SupaNoobs - Designed to be way overpowered.

The Ominiscient Antimind of Antimind- Not broken, but extremely tough. I would say tougher than Caretakers or Usurpers.

Conqueror Marr of the Manifold Usurpers- Perhaps the toughest of the canon factions.

Captian Svensgaard of The Nautalis Pirates - MP guys say it is an unfair advantage.

Chairman Yang of The Human Hive - Forum opinion from Apolyton, WPC, and CFC over a decade of posts say this is most powerful of human canon factions.

CEO Morgan of The Morganites - I sometimes disagree, but 10 years of forum lurking says Morgan is the weakest of the human canon factions.

Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 25, 2013, 02:40:57 pm
Feel free to add custom factions you're interested in to the faction info sheet, it'll be useful.

Also, the Yang/Morgan thing is only in the hands of AI. In human hands, Morgan is top notch and Yang is only decent.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Earthmichael on January 25, 2013, 05:11:12 pm
Once we get done with this, I am interested in the relative power rating of these factions:

Buster's Uncle's SupaNoobs - Designed to be way overpowered.

The Ominiscient Antimind of Antimind- Not broken, but extremely tough. I would say tougher than Caretakers or Usurpers.

Conqueror Marr of the Manifold Usurpers- Perhaps the toughest of the canon factions.

Captian Svensgaard of The Nautalis Pirates - MP guys say it is an unfair advantage.

Chairman Yang of The Human Hive - Forum opinion from Apolyton, WPC, and CFC over a decade of posts say this is most powerful of human canon factions.

CEO Morgan of The Morganites - I sometimes disagree, but 10 years of forum lurking says Morgan is the weakest of the human canon factions.

I don't know about the first two factions listed.  Usurpers is the strongest overall canon faction.

Pirates is a mixed bag.  He is pretty powerful, but less so in general than University.  UNLESS the map is mostly water.  If the land factions are all on small islands that limit expansion, Pirates rule!   I don't like having Pirates in a multiplayer game because he thrives on rising water, and therefore is unrestrained on ecodamage and attrocities.  The land factions have to spend so much time micromanaging against rising water that the game is no fun anymore.

Hive is not the most powerful faction; far from it.  Unless he gets an ally that feeds him some technology, he falls woefully behind.  He is a great faction to use in a team game, where he is teamed with Aki or Zak.  But otherwise, I would rate him middle of the road.

Morgan is far from the weakest faction.  He is one of the top 3 human factions on a large map with factions fairly widely spaced. 

Much of faction power is determined by map.  Pirates thrive on a heavily water map.  Spartans and Believers stink on a large map, but can be powerful on a map where factions start relatively close together.  Builder factions like Uni, Cycon, and Morgan are best on a large map with a fair amount of separation between factions.

You may say this is just one opinion against a decade of collective wisdom, but I am willing to back up my opinion with a game to prove my point.  Who is foolish enough to play me with Hive against my Morgan on the Vets map?  Bring it on!
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 05:43:54 pm
Depends... as you said, it is map.

If veterans map is a typical small MP map, we are looking at early conflict.

This is where Yang excels. Yang may not have energy, I give you that. But hitting Yang is like hitting a brick wall. Doctrine: Loyalty from the get-go. Impunity from Police State. Perimeter Defenses everywhere. NO ONE rexes like Yang. Little blue cockroaches scrurrying in and out of the fungus eating all the crumbs left over by those pathetic yellow Morganites. Can not run Democracy... but Yang is too much of a bad boy to run a wussy builder tech like demo. Leave that for pansies like Morgan or tech cheese guys like Zack.

Give Morgan breathing room... yeah. But in a steel cage small map with a monster like Yang? Military? Let Morgan look that up in the dictionary. Whats that? A picture of Santiago and Yang making love babies. Morgan has to buy the military from others. That is IF he beelines to Planetary Networks. But at the expense of oh no! What are those blue Impact rovers doing out there? Probes would be the only weakness.

You are right though and bring up a good point in our power rankings.

Map size/type may need to be factored.

Planet Cult is another one that comes to mind. Mediocre on most maps. Get that guy on an abundant lifeform map that has some size, I would put cult up against even the powerhouses.

 
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: BFG on January 25, 2013, 06:19:25 pm
At one point, I tried to set up a series of AI-only games using the 5 stock map sizes to answer the question of which faction (at least in the AI's hands) is most powerful.  Sadly, after about 3 games I grew tired of the effort, since AI-only still requires the player to sit there and hit "Enter" over and over again.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Earthmichael on January 25, 2013, 08:40:44 pm
The only thing you find out using AI is how the AI handles things, which is nothing like what human players do.  So you do not gain much multiplayer wisdom there.

FYI, the vets map (original or rebalanced) is a large map set up with places for 4 starting factions.  It is great for team games 2x2, or free for all, or 1 vs 1.  But being a large map, it is a builders map, not a momentum map.

Prophets just stink on any size map.  Worst faction; even with abundant native life, Gaia is better.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: BFG on January 25, 2013, 09:11:08 pm
The only thing you find out using AI is how the AI handles things, which is nothing like what human players do.  So you do not gain much multiplayer wisdom there.

FYI, the vets map (original or rebalanced) is a large map set up with places for 4 starting factions.  It is great for team games 2x2, or free for all, or 1 vs 1.  But being a large map, it is a builders map, not a momentum map.

Prophets just stink on any size map.  Worst faction; even with abundant native life, Gaia is better.
Given that I've NEVER played a SMAC multiplayer game (gasp), I'm just going to take your word for it :)
Still, such a setup might be useful to Yitzi and/or Kyrub to locate AI bugs or otherwise strengthen AI.  But I agree, it wouldn't tell you anything about the factions' strength in the hands of human players.

EDIT: Purely out of curiosity, do you ever use the Firaxis faction?
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Green1 on January 25, 2013, 09:23:09 pm
The only thing you find out using AI is how the AI handles things, which is nothing like what human players do.  So you do not gain much multiplayer wisdom there.

FYI, the vets map (original or rebalanced) is a large map set up with places for 4 starting factions.  It is great for team games 2x2, or free for all, or 1 vs 1.  But being a large map, it is a builders map, not a momentum map.

Prophets just stink on any size map.  Worst faction; even with abundant native life, Gaia is better.
Given that I've NEVER played a SMAC multiplayer game (gasp), I'm just going to take your word for it :)
Still, such a setup might be useful to Yitzi and/or Kyrub to locate AI bugs or otherwise strengthen AI.  But I agree, it wouldn't tell you anything about the factions' strength in the hands of human players.

EDIT: Purely out of curiosity, do you ever use the Firaxis faction?

Do not worry. I would say a good 60 to 70 percent of us do not do multiplayer. For all 4x games, the figure is that 10 percent do. That figure is from Kael at CFC. However, T_ras and them did want me to try to join them at some point and I may if I can get over how damn long it takes to do MP and all the headaches I need to go through to set it up.

AI is where it is at. I agree. And the AI is being worked on. At least I still find it somewhat challenging on Thinker/Transcend.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Earthmichael on January 25, 2013, 09:37:41 pm
I think everyone should play AI until transcend is a cakewalk.  It is lots of fun, and you learn lots about the game.  Even after you master one faction at transcend, you have 13 more to go.  Then you can also vary planet size and attributes, and which factions are in the game.

BUT, after you have done that until you are bored, then multiplayer is the way to go.  You never know what to expect from humans!

What I would suggest at first is a 1 on 1 multiplayer game with another player inexperienced with MP.  A 1 on 1 game typically proceed at least 5 times as fast as a 4 player game.  Check some of my 1 vs 1 game threads to see the difference.

Another interesting varient I like are two players each playing 2 factions.  It has the speed of a 1 on  1 game, but the extra factions add some variety.  To keep the speed up, make sure each player plays both factions in a row, before exchanging turns again.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Yitzi on January 27, 2013, 05:48:04 am
As part of the aim of creating a faction rating formula, I need numeric wieghts to the benefits and disadvantages of each value for each social engineering characteristic. Simple formulas do not match up to the complex effects, for example +1 Economy is nice, but +2 Economy is many times more useful, then +3 is only somewhat more useful.

To simulate this, I need your help. Please read the instructions carefully:
Rate each value of each social parameter between 10 and -10, with ten being the best, zero the middle, and -10 the worst. For the Economy example I'm using, I would put the numbers like:
Econ Rating
0   0
+1   1.5
+2   7
+3   8
+4   9
+5   10

This indicated that a rating of +2 is several times better than that of +1.

That actually might not be that accurate, as beyond +2 gives a commerce boost (which might not seem that impressive on first glance, but when you've got two or three pact brothers and Global Trade Pact passed, is extremely good.)
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: ete on January 27, 2013, 12:38:32 pm
hm, right. I was unsure how significant the commerce rate boost was, but let's see...

Quote
Commerce is computed base by base between factions with
$LINK<Treaties=4> and $LINK<Pacts=3>, as follows:
^
^(1) First, all bases for each faction are ranked from top to bottom
by Energy output.
^(2) Bases are paired off from top to bottom. If one faction has extra
bases, these are ignored.
^(3) For each pair of bases, sum the combined economic output and
divide by 8, rounding up.
^(4) Double this value if a Global Trade Pact is in effect.
^(5) Now, for each individual base, the commerce formula is as follows:
^       (ValueFromStep4) * (CommerceTech+1) / (TotalCommerceTech+1)
^(6) CommerceTech is the total # of economic technologies discovered,
plus faction & social bonuses.
^(7) TotalCommerceTech is the total # of economic technologies in the game.
^(8) Now, using the value from step 5, divide by 2 if no Pact (e.g. only a Treaty)
^(9) Add +1 if you are $LINK<Planetary Governor=28>.
^(10) Reduce to zero if sanctions are in effect against either faction.

(CommerceTech+1) / (TotalCommerceTech+1)

as a multiplier, so at low commerce ratings each step up is very significant % wise. I'll run some simulations, but you're probably right that the commerce boosts should be taken as much more significant. Though I guess in 1v1 games commerce is irrelevant. Probably not worth dealing with all the complexities raised by that till the basic formula is done.
Title: Re: Social Engineering: how much effect on gameplay does each value have
Post by: Yitzi on January 27, 2013, 01:23:59 pm
as a multiplier, so at low commerce ratings each step up is very significant % wise. I'll run some simulations, but you're probably right that the commerce boosts should be taken as much more significant. Though I guess in 1v1 games commerce is irrelevant. Probably not worth dealing with all the complexities raised by that till the basic formula is done.

Yeah, 1v1 is really a different game entirely.
Templates: 1: Printpage (default).
Sub templates: 4: init, print_above, main, print_below.
Language files: 4: index+Modifications.english (default), TopicRating/.english (default), PortaMx/PortaMx.english (default), OharaYTEmbed.english (default).
Style sheets: 0: .
Files included: 31 - 840KB. (show)
Queries used: 14.

[Show Queries]