AU: A MOD for the curiculum

[nbarclay]

Originally posted by player1

Just don't forget that Amp. ability. It actualy makes low defense unimportant.
Personnaly I think that 5/2/1/cost50 is not less powerfull unit at all. Although my main resoning was that Bezerk offense should not be the same as of Cavalry. Kills the flavor for me (but maybe not for you).

Unimportant only when attacking coastal targets, and the fact that the ships the Bezerks travel on can be sunk partially offsets the invulnerability to land attack while being transported. You're right that something feels off about Bezerks being as powerful on offense as cavalry; your values probably do make more sense, but I still prefer to stick with the stock version.

In regard to battleships, in the stock game (at least according to the PtW Civilopedia), the ROF is 2. But bombers have a ROF of 3 and only cost 100 shields. Also note that the ROF is irrelevant when attacking directly, so if you increase the ROF and then charge more as a result, you charge players for an ability they may or may not use depending on what role their battleships play.

Nathan

[alexman]

I need some feedback about an idea.

The AI places a huge importance on technologies that give units that don't require resources. This means that all civs research Bronze Working first (to get Spearmen), and then Warrior Code (to get Archers).

The problem with this approach is that all civs research the same things, and don't take advantage of their unique starting techs to research something more valuable.

For example, if Commercial civs (Alphabet) researched Writing instead, they would be in a very good position to either a) get lots of goodies from trading a technology that nobody else has or 2) get a head-start towards Literature and Republic.

Similarly, Religious civs would benefit by researching Mysticism first, and then b-lining towards Monarchy.

I have a solution that works well but is a bit drastic, so I need some feedback. The idea is to give Spearmen and Archers no technology requirement. Bronze working would still allow the Collossus, and we could have Warrior Code allow Barracks. I think it would be best to leave UUs like Hoplites, Impi, and Bab. Bowmen still to require their current technologies. That way UUs, in addition to starting techs, would influence research path.

The downside of this change is obviously the fact that it is a big change. But will it really change gameplay that much? For sure not much for the AI, because those are the two techs it researches first anyway. The human would be able to use Archers immediately, and that could be fun. Note that militaristic civs (except Japan) would still have the advantage for Archer rushes, because they would be able to build barracks immediately.

What do you all think?

[Dominae]

Re: Spearmen and Archers requiring no tech advance: too drastic. This is precisely the kind of change that is used solely to make the AI more optimal, that is, closer to what humans believe to be the best strategy. The net effect is arguably slightly better AI at the expense of variability in gameplay.

Other reasons why I don't think it's a good idea:

1. Warrior Code is now an empty tech.

2. The Colossus is never built right away, so saying that Bronze Working is still a useful tech isn't entirely true. By the time the Colossus is begun, at least half of the Ancient tech tree has been researched already.

3. The difference between Militaristic civs and others is that the Militaristic civs don't have to "waste time" researching Warrior Code in the first place (in order to get Archers). Thus, Militaristic civs can go "all out" aggression right away, while others must do so with a time investment.

4. Spearmen really are essential units, so I don't completely disagree with AI's tendency to research it first all the time. I'm sure there's some sort of randomization involved, and I believe that's enough (in this case). Archers are a different story. The "reason" why all AI research Warrior Code so early is that to be able to capitalize on early conquest, if the opportunity presents itself. Obviously the AI doesn't know when the opportunity presents itself, so it is coded for all eventualities, as it were. I don't think this is a flaw/mistake.


alexman, I honestly like where you're going to improve AI research, just not this particular solution. Keep the ideas coming!

By the way, are we going to include the relocation of special abilities (ROP, etc.) in the next version of the mod?


Dominae

[alexman]

Originally posted by Dominae
1. Warrior Code is now an empty tech.

The ability to build barracks is quite important. Don't you think?

2. The Colossus is never built right away, so saying that Bronze Working is still a useful tech isn't entirely true.

Don't forget that you need Bronze Working to research the very important Iron Working tech. That makes Bronze Working still extremely useful.

3. The difference between Militaristic civs and others is that the Militaristic civs don't have to "waste time" researching Warrior Code in the first place (in order to get Archers). Thus, Militaristic civs can go "all out" aggression right away, while others must do so with a time investment.

You can't really go "all out" without barracks against spearmen! You'll get slaughtered!

I'm sure there's some sort of randomization involved.

I am now sure there is no random factor. It's the difference in expected turns to completion that makes one civ choose one tech and another choose a different one. But in the case of Bronze Working and Warrior Code, no difference in expected completion time can ever make the AI not research them #1 and #2 respectively.

In summary, I really believe that the change is not going to affect gameplay, or human strategy that much. The options are still there. It's just the reasons for pursuing them that are different (Barracks instead of Archers for WC, and Iron Working + Collossus is already enough for BW).

By the way, are we going to include the relocation of special abilities (ROP, etc.) in the next version of the mod?

I would like to move Alliances to Polytheism, at least in addition to having them in Writing. (We believe in the same Gods, so together let's kill all those who don't...). This would make Polytheism more attractive and thus Monarchy more common, especially for religious civs.

I would also like to move the Pyramids to Mathematics, so the latter is researched more often by the AI. Industrious civs would still be the only ones one tech away from that Wonder. Also, the idea is to make starting technologies less valuable to the AI in order to encourage researching deeper in the tech tree and get the first tier from trade.

Also, we should make taxmen require currency. This will make this technology more valuable to the AI (marketplaces!), but not by much.

Finally, we should increase the AI value of Literature (by a lot), Code of Laws, and Philosophy, so that the AI ends up getting the Republic, Libraries, and the GL sooner. We would do that by duplicating some of the special abilities of Writing in those techs (that have Writing as a prereq. anyway). Similarly, I think we should increase the AI value of Printing Press and Radio a little, so the human can't as often take advantage of researching them with one scientist and selling it for a fortune. In the Industrial age, we need to increase the AI value of Electricity and Atomic Theory, so the AI doesn't get beat to the classic turning points of ToE and Hoover as often.

These changes, combined with the original two I proposed, make for very reasonable AI research choices, believe me!

[ducki]

I like the Pyramids to Math move, if only conceptually/logically for now, but...

Isn't Math a 2-prereq tech? I thought it required... uhh... hell.... Alphabet! And uhh... grr... what's the Industrious one? Masonry?

Anyway, if I'm misremembering, ignore me, but I was positive Math had 2 prereqs, making it 2 techs away from any industrious Civ.

I don't understand Alliances to Polytheism, though. I take it we decided the Gardens were not to be moved from Monarchy?

[alexman]

Yes, you're right. Math requires Alphabet too. Oh well...


We could definitely move the hanging gardens instead of the Alliances, although the effect isn't as great and we would also have to do something else like reducing the cost of Polytheism. And I still say it would also make Monarchy very close to useless for human non-religious players.

[Dominae]

Originally posted by alexman
The ability to build barracks is quite important. Don't you think?

Heh, you got me on that one! Then again, I don't normally consider building Barracks early on unless I'm Militaristic.

Originally posted by alexman
Don't forget that you need Bronze Working to research the very important Iron Working tech. That makes Bronze Working still extremely useful.

Agreed. I was actually wondering why you didn't mention this in your original post!

Originally posted by alexman
You can't really go "all out" without barracks against spearmen! You'll get slaughtered!

Yes, "all out" is too strong a term. I'm referring to simple Archer rush, which isn't all out, but comes close to qualifying considering the resources you have in the early-game. Perhaps I'm doing it wrong, but I don't build Barracks before Archers in an Archer rush. In any case, your example solidifies my point that giving everyone Spearmen right away will reduce some strategies, namely really early Archer rushes. Giving Archers to everyone doesn't solve this problem.

Originally posted by alexman
I am now sure there is no random factor. It's the difference in expected turns to completion that makes one civ choose one tech and another choose a different one. But in the case of Bronze Working and Warrior Code, no difference in expected completion time can ever make the AI not research them #1 and #2 respectively.

Wow. I'd really like to get confirmation from Soren on this one (although I do trust your judgement). This means that we can play with the AI research strategies quite drastically.

Originally posted by alexman
In summary, I really believe that the change is not going to affect gameplay, or human strategy that much. The options are still there. It's just the reasons for pursuing them that are different (Barracks instead of Archers for WC, and Iron Working + Collossus is already enough for BW).

You're also probably never going to see an AI Warrior (if I understand the build algorithms correctly). Am I the only one that thinks Barracks aren't that important in the very very early game when you're conducting your first little military excursion? I mean 5-10 units max, a combination of Warriors and Archers.

Another point that I have against your proposed change is that it is likely to spill over to all the other techs that we think the AI doesn't research correctly. Nationalism is preferred, make Riflemen available for free in Industrial age (not possible, but you know what I mean), etc. I've also showed (maybe not convincingly) that the AI's preference for Bronze Working and Warrior Code makes sense: those that have it already gain a very early advantage, those that don't need to catch up (militarily), but have other, non-military advantages to offset this with (like Granaries, etc.).

Originally posted by alexman
I would like to move Alliances to Polytheism, at least in addition to having them in Writing. (We believe in the same Gods, so together let's kill all those who don't...). This would make Polytheism more attractive and thus Monarchy more common, especially for religious civs.

What about Philosophy? I still think Hanging Gardens at Polytheism makes sense, and doesn't render Monarchy useless for non-Religious civs. I do like your "realism" rationale, though.

Originally posted by alexman
I would also like to move the Pyramids to Mathematics, so the latter is researched more often by the AI. Industrious civs would still be the only ones one tech away from that Wonder. Also, the idea is to make starting technologies less valuable to the AI in order to encourage researching deeper in the tech tree and get the first tier from trade.

Very nice. As you said, it would make Mathematics a more useful tech, and it would also mean that the option to build the Pyramids wouldn't be available without some work (Masonry is just too easy to get). Very nice.

Originally posted by alexman
Also, we should make taxmen require currency. This will make this technology more valuable to the AI (marketplaces!), but not by much.

Sure.

Originally posted by alexman
Finally, we should increase the AI value of Literature (by a lot), Code of Laws, and Philosophy, so that the AI ends up getting the Republic, Libraries, and the GL sooner. We would do that by duplicating some of the special abilities of Writing in those techs (that have Writing as a prereq. anyway).

These changes, combined with the original two I proposed, make for very reasonable AI research choices, given their starting techs, believe me!

If it has the effect that you say it does, then I'm all for it. I'll just have to deal with "gaining" abilities I know already with each new advance!


Dominae

[alexman]

Dominae, thanks for your input.

This has to be quick because my boss will kill me!

In case you've missed it, here is Sir Ralph's definitive guide to the Archer Rush.

Also, there is no way for the human player to know about the extra abilities that we give techs that come after Writing. They don't even show in the civilopedia. This is purely an AI-value increasing trick.

(By the way, Pyramids require Masonry in the stock version, not Pottery)

[Dominae]

Originally posted by alexman
Dominae, thanks for your input.

No problem. I just want to make it clear that I'm not putting down your efforts toward a better mod in any way. Excellent work.

Originally posted by alexman
In case you've missed it, here is Sir Ralph's definitive guide to the Archer Rush

Yup, I read it. A great guide, but I pick and choose the bits that I incorporate into my games. If I build Barracks, I'll probably be building up a large force of Horsemen or Swordsmen (perhaps via upgrades), not Archers.

Originally posted by alexman
Also, there is no way for the human player to know about the extra abilities that we give techs that come after Writing. They don't even show in the civilopedia. This is purely an AI-value increasing trick.

Oh, ok, I thought it would appear in the Civilopedia description.

Originally posted by alexman
(By the way, Pyramids require Masonry in the stock version, not Pottery)

Yeah, I know. I have a funny habit of posting then proof-reading, rather than using the Preview Reply button. You got me before I caught that mistake.

I would like to get other people's input concerning Monarchy (because I think we've both made our points and still haven't reached an agreement). Is Monarchy (the advance) not a viable option if it doesn't come with Hanging Gardens? More specifically, do non-Religious civs have any reason to switch to Monarchy before settling into The Republic (or Democracy)?


Dominae

[ducki]

On the barracks issue, once I've got my 3 or 4 warriors cranked out at the beginning, I will forgo any and all military buildup until I have barracks.

I'll even build a barracks in my main settler city in between workers/settlers, so I can use it to build up military in between later settlers.

I think the importance of barracks is highly subjective, so for some, giving it a required tech would be a big change. For other, it wouldn't.

I do know that the AI tends to send out lots and lots of Regulars, so I don't know that it would affect them on the military front all that much.

Me, I don't like fighting with Regulars if I can avoid it, captured border towns notwithstanding - sometimes you just have to rush that Regular Defender.

[ducki]

Originally posted by Dominae
I would like to get other people's input concerning Monarchy (because I think we've both made our points and still haven't reached an agreement). Is Monarchy (the advance) not a viable option if it doesn't come with Hanging Gardens? More specifically, do non-Religious civs have any reason to switch to Monarchy before settling into The Republic (or Democracy)?

If Monarchy didn't have the Gardens, I would almost never research it. Almost never. Even with a religious Civ. Then again, I'm not that experienced of a warmonger, so take that with a grain of salt. I prefer Luxuries+Republic->Democracy to FreeUnits+MilitaryPolice+Monarchy for warring.

[Dominae]

In my experience, a protracted war isn't always possible in Republic, even with 2-3 luxuries. If you're playing a very aggressive game (Zulus, Vikings, but any other civ will do), you just don't ever want to be outside of war. In those cases, researching Monarchy and staying there until much later has worked for me. Depending on how powerful the AI is relative to you, jumping into The Republic won't get you back into the game if you're behind. You have to go to war to catch up. Enter Monarchy.

Strangely enough, Monarchy and The Republic are the two techs that I find myself researching the most in the Ancient era (although never both). I suspect this is because the AI prioritizes required techs (any info on that alexman?). Thus, I seem to be researching Monarchy for its government and sale value, not the Hanging Gardens.


Dominae

[alexman]

Originally posted by Dominae
I suspect this is because the AI prioritizes required techs (any info on that alexman?).

Optional techs have their AI value reduced by 2/3.

From what I have seen, it's true that the AI doesn't place a high enough value on the Republic (because it doesn't also have a Wonder, like Monarchy does, hint, hint ). That's why I suggest adding one of the Writing attributes to the Republic.

As for Monarchy, the AI gets delayed because Polytheism has a very low priority, and is often the last required tech of the ancient era to get researched. Once in the Middle ages, Feudalism has a much greater importance to the AI than Monarchy.

[ducki]

Originally posted by Dominae
In my experience, a protracted war isn't always possible in Republic, even with 2-3 luxuries.

I was thinking more like 4-5 with marketplaces, but I get your point.

If you're playing a very aggressive game (Zulus, Vikings, but any other civ will do), you just don't ever want to be outside of war.

Like I said, almost never. I'm still new to warmongering, so "playing a very aggressive game" would fall under almost never. Almost.

In those cases, researching Monarchy and staying there until much later has worked for me.

I agree. Monarchy->Democracy is very powerful if you plan on lots and lots of war and don't have the luxuries and production for Republic to be workable.

Depending on how powerful the AI is relative to you, jumping into The Republic won't get you back into the game if you're behind. You have to go to war to catch up. Enter Monarchy.

Still, if I've got 4-5 luxuries, or can get them in a relatively short war, I still prefer Republic, but have tried out the Monarchy route.

Strangely enough, Monarchy and The Republic are the two techs that I find myself researching the most in the Ancient era (although never both). I suspect this is because the AI prioritizes required techs (any info on that alexman?). Thus, I seem to be researching Monarchy for its government and sale value, not the Hanging Gardens.

Interesting, I may have to try that.
I usually start out on The Wheel then head for Literature unless I'm Commercial, in which case I skip The Wheel.

Once I hit Literature - actually, once I hit Writing - I've got a pretty good idea whether I'm going to need the Library. If not, I'll head for whichever techs the AI is not going for, determined mostly by seeing what they have for trading.

I'll have to try beelining for Republic some time. I've beelined for Monarchy a lot, but most often so I can build the Gardens, not to switch governments.

I predict that moving the Gardens to Polytheism will have the AI consistently beating me to the Gardens, which is a "good thing" for AI strength, though not so good for my own citizen happiness.



This sort of discussion is what, I think, makes Civ such a great game. There are so many different ways you can play that changing stuff like Gardens->Polytheism might or might not affect all of us. And the fact that the decisions I make are somewhat influenced by things I read here, there is, by no means, a single "formula" for play, which would be boring.

[Dominae]

alexman, if I understand your work on AI research priorities, Monarchy and The Republic should always be researched before Feudalism, because they allow new governments. Feudalism grants a Wonder and a resource-required defender, which I don't believe add up to surpasss the value for new governments (I don't remember the precise values).

Edit:

Feudalism: Pikemen 70, Med. Infantry 6, Sun Tzu's 6 = 82
Monarchy: Monarchy 256 / 1.5 optional (excluding Hanging Gardens) = 170 (approximately)


Dominae