AU mod: Balancing the Governments

[ducki]

We don't know what the AI's taste in goverment is determined by so we'd have to playtest that. I've seen sensible AI decisions on goverment before so it's likely to be O.K.

Surely someone's done research on what government an AI will choose, no?
I'm sure that what we see as "dumb" decisions are really just a reaction to certain flags - war weariness, number of units, NumCities/OCN, raw commerce, etc.
Additionally, a lot of the "dumb" probably comes from an inability to use the Lux slider or to chase the 8th luxury - call it general mood management I guess.

Anyway, good points. I'm surprised someone hasn't "done an alexman" on AI government decisions.


I'm getting the inkling of an idea for another AU course, or at the least, a big comparison game - a game in which you have to utilize all 6 of the governments.
Naturally, we'd want a religious civ for this.
Maybe rule out the luxury slider as well.
Might give us some insight. Might not. Who knows.

[alexman]

OK here is that Government comparison chart again, this time including PTW Democracy with an optimal FP placement.

As you can see, the PTW Democracy (yellow dashed line)and Communal democracy (green dashed line) are comparable for quite a big range of empire sizes, but for medium-sized empires, where the argument seems to focus, the old PTW double-core Democracy model was actually more powerful.

[nbarclay]

What kind of city pattern did you use to test optimal city placement in PtW?

Also, if I understand correctly, the OCN was lower in PtW. If so, how did you account for that? (If PtW was set to the same OCN value as C3C, that would result in less corruption than actually existed in PtW games.)

[alexman]

It's an ideal city placement, so I removed half of the population in the pattern (the farthest cities from the capital were removed to leave a 17 city build), and then doubled the income.

I didn't account for the C3C OCN increase, but since I didn't actually have a FP, I didn't account for the FP OCN decrease either, these two effects partially cancel each other out. I don't think this is a big deal, especially since so many other things in this analysis were arbitrarily chosen.

[alexman]

One other thing. In PTW you could actually have a more efficient empire than what is assumed above, as long as the city density around your Palace was less than that around your FP. This was due to the FP rank bug, and players took advantage of it even accidentally when they moved their Palace to a former AI core, which usually had a looser build pattern than their own core.

[Theseus]

1. BACDE
2. BCAD
3. BCA..... D
4. BAD..... C

[alexman]

4. BAD..... C? Who are you and what have you done with Theseus, who supports two government switches and the behavior shown below?



Seriously, I'm happy with the outcome , although I do think our change to Fascism will not make any difference to its balance versus Communism, and that we might have given Feudalism an unnecessary unit support cost change.

[nbarclay]

So for those of us too lazy to count the votes ourselves, what did we end up with?

[alexman]

1. Reduced corruption for Feudalism to minimal. (1gpt unit support)
2. Reduced corruption for Fascism to minimal.
3. Reduced free unit support for Republic to 0/1/1 per town/city/metro, and added a flat free support of 12 units.
4. Increased free unit support for Democracy to 0/1/1 per town/city/metro, and added a flat free support of 12 units.

For all changes in the next version, see the topped main thread. I keep all changes up-to-date there.

[nbarclay]

Thanks. Your work on the Mod is definitely appreciated!

[nbarclay]

Now that we've made our choices for the next version of the Mod, I'd like to make a proposal for the one after: eliminate the SPHQ from Communism. That would keep the production advantage that Communism can provide beginning at sizes even a little over OCN (and probaboly below the OCN for more sprawling civs) from being as overpowering, and would make it a lot harder for Communist empires to match the research capabilities of Democratic ones. The only real down side I see is that it would give players less incentive to research Espionage, but compared with how powerful the SPHQ can be for a large Communist civ, that seems like a relatively minor concern.

[pvzh]

Why not move SPHQ to Fascism? It would stengthen it to be a worthwhile alternative. Right now fascism is as good as Feudalism: higher unit support (but who need that many anyway?), but forced resetlement and xenophobic. Not worth a switch even if you are religious.

Besides SPHQ in fascism will provide only 3/8 times in OCN boost not 200% or 300% (?) like in communism. I doubt it will be game breaking.

[Krill]

Common sense should probably tell me this, but: why not wait until after AU 502 to make recommendations?
You get the extra knowledge gleaned from that game, so why are you so hasty?

[pvzh]

Well, if for every obvious change we will wait for the game to end we will get the mod right by the time people are fed up with the game. Besides, we all play other games in AU mod aside of AU courses, so we can use this experience on the balancing AU mod as well.

[Theseus]

1. alex, I am just very hesitant about the 12 free units thing. Otherwise, I agree with where the 'groupthink' is taking us. Good stuff, and kudos to you for keeping on keeping on.

2. I, too, am attracted to the idea of moving the SPHQ to Fascism... but let's take it slow.