Updating the AU mod for C3C.

[Dominae]

Until we get decided on what types of threads we want, I figured I would get the ball rolling here.

Here are my thoughts on changes to units (including UUs). Note that I've not really looked at planes and ships, as I still have little experience with them.

Curragh

Increase cost to 20 Shields.

I've only seen the AI use these on Sid. This means that in a vast majority of games they're a human-only unit. This gives the human player an important advantage in terms of early Contacts in almost any given game. The "suicide Curragh" strategy (also human-only) reinforces this problem. Although not a super-solution, making Curraghs more expensive addresses this to some extent.

I'm also toying with the idea of restricting Curraghs to 1-tile movement only in Sea and Ocean, and making them only available to Seafaring civs. This is quite a drastic change, however.

Cavalry

Reduce Attack to 5.
Create new "Improved Cavalry" available at Nationalism with old Cavalry stats.

As discussed previously, most games "end" with a Cavalry offensive by the human player. The AI needs Riflemen to effectively defend against Cavalry, but Nationalism is often a long way off when the human player beelines to Military Tradition.

The way this modification works is that with the advent of Nationalism a civ can "upgrade" their Cavalry to Improved Cavalry at no cost (nationalistic sentiment makes them more effective in battle, or whatever). A nice side effect of this is that it should give some incentive for the human player to research or trade for Nationalism.

I'm eager to see if this change works out in actual play.

Infantry

Increase Attack to 8.

The AI uses Infantry as its primary attack unit until Tanks, so making them more effective on the offensive helps it out. This worked out great in previous versions of the mod and I think we should use it here too.

Jaguar Warrior

The Shield cost of the Jag was increased solely for the MP community. Should we revert back to the 10-Shield version, since AU is a SP mod?

Chasqui Scout

0/0/1, ATAR, 20 Shields.
Scout upgrades to Chasqui Scout.

The current version of the Chasqui is laughably poor. This gives the Incans a powerful super-Scout right from the start. This AI uses the first one well, but never builds more. Through testing there is some concern that this is too powerful a unit in the human player's hands. Personally I think the UU is fine as is, the real problem being that it belongs to an Agricultural civ.

Swiss Mercenary

Increase cost to 40 Shields.

Yes, we all love the Swiss Merc. But let's face it, the only reason we do is that it's super-powerful. Half-cost Musketmen without the Saltpeter requirement is too good. Hopefully the increase in Shield cost will transfer it from the "no-brainer" to the "slightly interesting" category.

Ansar Warrior

Another unit that I think is too powerful. When it comes down to it, this is a cheaper Rider. And we all know how the AI cannot defend against 3-Move units. I'm not sure how to balance this, and if we should.

Keshik

By contrast, the Mongols get the shaft in the Knight UU category. The Mongols could use the help.

Conquistador

Reduce cost to 60 Shields.
AND/OR
Swap Attack and Defense values (Attack 2, Defense 3)

The problem with the Conquistador is that, at its 70-Shield cost, most of the time you would rather have another Knight. Paying 70-Shields to send these deep into enemy territory only to perish the next round is not exactly cost-effective (incidentally, the AI never does this).

Both of the proposed changes make the Spanish UU slightly better at doing what they do best. If you're not too afraid of the ATAR ability you'll even support implementing both.

H'wacha

Lethal land bombardment needs some testing. Although it's a novel unit to use when you get around to playing the Koreans, IMO it turns games into boring bombard-fests against the AI, with a foregone conclusion.

Javelin Thrower

This unit is too powerful for the human player, as it's a simple matter to keep a few Barb camps around, for use as "Slave farms". On average this translates a couple of Javs into ~15 Slaves...well worth the investment. Needless to say, the AI never does this.

Assuming Firaxis/Breakaway does nothing about this bugger, what should we do about it?

An "evil" solution is reduce cost to 20 Shields but make them cost 1 pop to build (like Workers). Unfortunately this would hurt the AI more than the human player.

Crusader

Increase Attack to 6.

Am I the only one who is not impressed with the Knights Templar?

---

All comments are (of course) welcome.


Dominae

[DrSpike]

A nice list. Regarding the broken curraghs for seafaring civs I think they should only have the move bonus, not the different probability of sinking in sea/ocean. If you're worried about suicide runs that the AI will not attempt still occuring then I'd favour increasing the chance of sinking in ocean. Glancing at the editor I'm not sure that's possible, but probably best to increase the probability for both sea and ocean if that is the only possibility.

I'd favour those changes and keep the cost at 15.

[ducki]

Originally posted by Dominae
Curragh
Increase cost to 20 Shields.

I've only seen the AI use these on Sid. This means that in a vast majority of games they're a human-only unit. This gives the human player an important advantage in terms of early Contacts in almost any given game. The "suicide Curragh" strategy (also human-only) reinforces this problem. Although not a super-solution, making Curraghs more expensive addresses this to some extent.

I'd prefer a solution that encourages the AI to build Curraghs and explore more than trying to hamper the human.

I'm also toying with the idea of restricting Curraghs to 1-tile movement only in Sea and Ocean, and making them only available to Seafaring civs. This is quite a drastic change, however.

As one who rarely plays Seafaring(though I'm liking Carthage), I don't like this idea. If you only ever play on Pangea/Continents, it's a great idea, but 'pelago players AND Seafarers would both be penalized. Might as well not have the unit in the game in those situations.
Prefer encouraging the AI to build and explore with them. Sure wish the AI could learn to read the terrain cues and 'probe' likely areas for crossing. I rarely do suicides, but I will probe the obvious spots.

Cavalry

Reduce Attack to 5.
Create new "Improved Cavalry" available at Nationalism with old Cavalry stats.

As discussed previously, most games "end" with a Cavalry offensive by the human player. The AI needs Riflemen to effectively defend against Cavalry, but Nationalism is often a long way off when the human player beelines to Military Tradition.

The way this modification works is that with the advent of Nationalism a civ can "upgrade" their Cavalry to Improved Cavalry at no cost (nationalistic sentiment makes them more effective in battle, or whatever). A nice side effect of this is that it should give some incentive for the human player to research or trade for Nationalism.

I'm eager to see if this change works out in actual play.

Love this idea, but want to return to "As discussed previously, most games "end" with a Cavalry offensive by the human player. " below.

Chasqui Scout
...
Personally I think the UU is fine as is, the real problem being that it belongs to an Agricultural civ.

Also return to this related with "The game is over at..."

Swiss Mercenary

Increase cost to 40 Shields.

Yes, we all love the Swiss Merc. But let's face it, the only reason we do is that it's super-powerful. Half-cost Musketmen without the Saltpeter requirement is too good. Hopefully the increase in Shield cost will transfer it from the "no-brainer" to the "slightly interesting" category.

I think all UUs should be just below "no-brainer" but far above "slightly interesting". You should have to agonize over taking advantage of your Civs uniqueness versus expedience. That's just me, but I think UUs should be an opportunity to truly turn the tide, again, related to the point I'll revisit.

Crusader

Increase Attack to 6.

Am I the only one who is not impressed with the Knights Templar?

I think using the word "Knight" adds to the lack of "wow". I have to keep reminding myself that you get a super-MedInf for free every 5 turns without resources, IIRC. Not as unbalancing as Zeus, but amassing an army by only having to build half the units is a big advantage. I'd rather have knights, but if I have no horses, I need something to even the odds.

And now, back to the other point(s).
"As discussed previously, most games "end" with a Cavalry offensive by the human player. "
"Personally I think the UU is fine as is, the real problem being that it belongs to an Agricultural civ."
"Hopefully the increase in Shield cost will transfer it from the "no-brainer" to the "slightly interesting" category."

So much of the balancing and discussion of imbalance has its roots in the early game. "Winning Early - What Do YOU Do?" was my introduction to the idea that you win or lose in the Ancient Age.

This is not necessarily as it should be, I think. The game that ends with a Cavalry offensive was actually over long before that. I believe it was over as soon as the player had built his first 5 towns. The problems are not really that Agriculture is overpowered or that REX rules or that a granary before a settler beats the AI every time. It's that the AI doesn't know how to catch up, get ahead, exploit his strengths.

What I'd like to see is more Age balance. I shouldn't know if I will win or lose by the time peaceful REX is over. I shouldn't have a(n almost) carved-in-stone game beginning that doesn't vary from Civ to Civ, terrain to terrain. Which brings me back to your suggestions.

I like most of them, some I haven't seen the original yet, so can't speak to them.
But I'd like to see more AI-encouragement than Player-handcuffing. If any first-rank tech/unit/building needs a little nerfing, I'm of the opinion that it's the granary more than the curragh or agriculture. Now I'll go get my firesuit, as I'm sure noone wants their granaries messed with.

Oh yeah. Good list Dom. Thanks for getting the discussion going on specifics.

[nbarclay]

Originally posted by Dominae

Curragh

Increase cost to 20 Shields.

I've only seen the AI use these on Sid. This means that in a vast majority of games they're a human-only unit. This gives the human player an important advantage in terms of early Contacts in almost any given game. The "suicide Curragh" strategy (also human-only) reinforces this problem. Although not a super-solution, making Curraghs more expensive addresses this to some extent.

I'm also toying with the idea of restricting Curraghs to 1-tile movement only in Sea and Ocean, and making them only available to Seafaring civs. This is quite a drastic change, however.

For non-Seafaring civs, I think the curragh is fairly well balanced as is. It is one and a half times the cost of a warrior or scout, with the same speed as a scout traveling over open terrain. In exchange for its speed, however, it gives up the ability to pop huts and to do scouting inland for good city sites. So I don't regard "build a curragh ASAP" as so much of a no-brainer as to need fixed. Of course Seafaring civs can get a lot more advantage, but not overpowringly so.

I do like the idea of sabotaging the ability to send suicide curraghs across the ocean to the far continent ultra-early.

Cavalry

Reduce Attack to 5.
Create new "Improved Cavalry" available at Nationalism with old Cavalry stats.

Considering how many more cities are size 7 when players get cavalry than when they get knights, the power ratio of cavalry vs. musketmen is not out of line with knights vs. pikemen. With cavalry reduced to an attack value of 5, cavalry vs. musketmen would be weaker than knights vs. pikemen even before taking the difference in the number of size 7 cities into consideration.

The problem with this is that one of the key strategic decisions in stock Civ 3 is, "Do I attack now with knights, or wait until later to attack with cavalry?" After all, every knight killed is a unit that can't be upgraded to cavalry. Attacking with knights gives a civ additional territory earlier, but if cavalry have better odds than knights, there are also advantages to waiting.

But if cavalry can't attack with better odds than knights, going ahead and attacking with knights against pikes becomes pretty much a no-brainer in anything resembling a close game. You get the additional conquered territory sooner, and you don't give up any significant potential for more potent conquest later on (unless you were already so far ahead that you expected to attack with cavalry against pikes or advanced cavalry against muskets). I view that as violating the goal of preserving the flavor of Civ 3.

Chasqui Scout

0/0/1, ATAR, 20 Shields.
Scout upgrades to Chasqui Scout.

The current version of the Chasqui is laughably poor. This gives the Incans a powerful super-Scout right from the start. This AI uses the first one well, but never builds more. Through testing there is some concern that this is too powerful a unit in the human player's hands. Personally I think the UU is fine as is, the real problem being that it belongs to an Agricultural civ.

If we reduce the Jaguar Warrior's cost back to 10 (which I think would be a good idea, since the Aztecs have not traditionally been considered excessively powerful in SP), reducing the cost of the Chasqui Scout to 15 should make it a respectable UU without fundamentally altering its nature. It would be both faster than regular scouts on hills and mountains and able to defend itself against barbarians. (One limitation on regular scouts' exploration abilities under standard rules is that especially on higher barbarian settings, they tend to get killed off eventually. A scout that doesn't get killed by the first barbarian he runs into is limited only by the need to end turns outside other civs' borders.)

Also note that with attack and defense values, the Chasqui Scout can trigger a GA through military means, and might possibly even be militarily useful in the process. In contrast, changing it to a faster scout with no attack or defense means that the Incas can obtain a GA only with wonders.

Swiss Mercenary

Increase cost to 40 Shields.

Yes, we all love the Swiss Merc. But let's face it, the only reason we do is that it's super-powerful. Half-cost Musketmen without the Saltpeter requirement is too good. Hopefully the increase in Shield cost will transfer it from the "no-brainer" to the "slightly interesting" category.

The Swiss Mercenary's advantage over the Pikeman is smaller percentagewise than the Hoplite's over the Spearman. More importantly, the advantages of the War Chariot, Mounted Warrior, and Immortal are not only as great or greater percentagewise but also targeted toward offense rather than defense where they tend to do human players more good. In human hands, the Swiss Mercenary does not even come close to being the best UU in the game, and if it makes one of the AIs particularly tough, is that really such a bad thing?

H'wacha

Lethal land bombardment needs some testing. Although it's a novel unit to use when you get around to playing the Koreans, IMO it turns games into boring bombard-fests against the AI, with a foregone conclusion.

This would be open to serious exploitation. Imagine keeping some Hwachas around to finish off units knocked down to one hit point by more modern artillery units. Not to mention the serious anachronism of giving Hwachas lethal land bombardment but not Artillery or even Radar Artillery.

Crusader

Increase Attack to 6.

Am I the only one who is not impressed with the Knights Templar?

For someone whose strategy centers around racing to Chivalry and doing a lot of fighting with knights, the Knights Templar wonder could be a good investment. Crusaders are more powerful than knights, and 300 shields is just over the cost of four knights. Every crusader beyond four can be thought of more or less as a free unit.

On the other hand, when a top player races through the medieval era as quickly as possible and researches Steam Power first thing in the industrial era, the wonder's useful life is short enough that it may not do much more than pay for itself. Considering the opportunity cost represented by paying 300 shields up front to get not a whole lot more shields' than that worth of units later, it's not a great deal for someone who is going to be zipping through the medieval era.

Aside perhaps from questions of realism, I wouldn't have objected if crusaders had been given an attack of six. But I don't view the wonder as it stands now as broken.

[alexman]

Before you discuss any more changes for the AU mod in this thread, please take a look at the new AU mod thread.

[MrWhereItsAt]

Originally posted by Dominae
I'm also toying with the idea of restricting Curraghs to 1-tile movement only in Sea and Ocean, and making them only available to Seafaring civs. This is quite a drastic change, however.

I am puzzled by this - doesn't this mean they would sink if they ventured beyond coast? Or do you mean something else?

Crusaders may look ineffectual by being so slow (I assume this is your gripe with them. Com), but they really shine in Armies. If you are going out of your way to build a Wonder that just gives you extra attacking units (and they DO count for army maintenance purposes!), then you are planning on warring or already are. In these cases you can reasonably expect some MGLs, from what I have seen, and Crusaders in an Army means attacking with 7/4/2 stats. BLISS.

The two full games I have played of C3C I have got the Knights Templars and they have devastated my opponents if used correctly. Knights become the support units then, and the Crusaders/Crusader Armies do all the hard work.

[Dominae]

Originally posted by nbarclay
For non-Seafaring civs, I think the curragh is fairly well balanced as is. It is one and a half times the cost of a warrior or scout, with the same speed as a scout traveling over open terrain. In exchange for its speed, however, it gives up the ability to pop huts and to do scouting inland for good city sites. So I don't regard "build a curragh ASAP" as so much of a no-brainer as to need fixed.

Curraghs provide you with early Contacts. The AI does not build them. Therefore in any game where early Contacts is an important thing (i.e. any game Monarch level and above), Curraghs are worth well over their cost in Shields (just translate the cost of techs you got via Curragh-enabled trading). The fact that you can use suicide missions to amplify this advantage makes the situation even worse.

But if cavalry can't attack with better odds than knights, going ahead and attacking with knights against pikes becomes pretty much a no-brainer in anything resembling a close game.

I disagree. It's pretty much accepted that games of Civ3 "end" with Cavalry. Do you disagree with this, or think that this is not a problem? If not, why go into specifics.

In just looking at combat odds, I think there are multiple differences between Knigths and Cavalry that you're not considering.

1. An offensive with Knights occurs much earlier on in the game, and it's rare for your economy to strong enough at that point to completely overrun all opposition. The AI has the advantage in the earl-game in terms of production power (as you increase difficulty level, you extend the period over which the AI maintains this advantage). Therefore, it's hardly a no-brainer to always attack with Knights, even if Cavalry are "weaker" stats-wise. You'll never put as big a dent in the opposition with Knights as you will with Cavalry.

Now, granted, this is a problem with the AI and not Cavalry per se. However, when coupled with the points below it very much becomes a Cavalry problem.

2. Cavalry have 3 movement points. This goes a long long way in making Cavalry simply superior to Knights when they have the same relative odds in battle. How many cities you can grab in one turn is far more dependent on movement points than on A/D values (in this case).

3. Musketmen are much "closer" to Knights than Riflemen are to Cavalry. This is the big one. In a close game, the amount of damage you can do with Knights until your opposition starts to trump them with Musketmen is far less than what you can accomplish with Cavalry until Riflemen arrive on the scene. If the human player is playing smartly, there's something like a 6-tech spread there (probably more). The AI's only good defense against a human offensive is good defensive units. When those defensive units are 8 techs away (and the human player is conquering 10 cities per turn), it's game over.

As you can see, I disagree with your argument that weaker Cavalry undermines the decision between Knights and Cavalry. All the change does is ensure that the AI has a fighting chance to get into the Industrial era when the human player beelines for Military Tradition.

The Swiss Mercenary's advantage over the Pikeman is smaller percentagewise than the Hoplite's over the Spearman.

I think it's better to compare UUs to non-UUs, but let's do it your way:

1. Swiss Mercs waive the Saltpeter requirement for Musketmen. Hoplites and Spearmen have no resource requirement.

2. Swiss Mercs do not only really Pikemen, but Musketmen too. The cost reduction of Hoplites:Pikemen is 33%; that of Swiss:Musketmen is 50%.

3. The AI has better chances against Hoplites than against Swiss Mercs, because: 1) it's the early-game and the AI is better in the early-game, and 2) Swordsmen and Swordsmen-UUs do better against Hoplites than Knights and Knights-UUs do against Musketmen.

What this means is that, in the hands of the human player, the AI has no chance of putting a dent in your until Military Tradition, and even then it would have a lot of trouble.

In human hands, the Swiss Mercenary does not even come close to being the best UU in the game, and if it makes one of the AIs particularly tough, is that really such a bad thing?

The best UU in the game should not be our baseline. The AI probably most often builds Swiss Mercs at 40 Shields in the same amount of time than at 30 Shields, due to its production bonuses at Monarch and above.


Dominae

[alexman]

I'm trying to create separate threads for AU mod discussions like this. Please help me!!

Otherwise some great discussion will get left out of the AU mod documentation.

[Dominae]

Sorry.

I was trying to keep discussion threads in the same thread, but your goal seems loftier. I will help.


Dominae

[alexman]

I would like us to finalize the panel so that we can start voting on AU mod issues and start opening new AU mod threads for discussion.

nbarclay, you have been nominated to be the 7th member of the panel. Do you accept?

If not, I nominate Stuie, who has shown interest in the mod lately.

If neither of the above accepts, and if nobody wants to resign, we should set up a public poll to determine the 5 members of the panel.

[Stuie]

Thanks alexman. I'm available if called upon, but obviously defer to nbarclay's decision.

[alexman]

Actually, after a PM exchange with the other panel members, and unless there are any objections, I will open a poll today to determine the 5 (or 3) members of the panel.

7 panel members are just too many to coordinate when a quick decision is needed.

Thanks for accepting, you will be among the candidates, Stuie.

[Stuie]

Ha! I'd be happy to be number 7, but if there's only five there are others far more qualified/deserving (as I'm sure the poll will show...). Even as a 7th I felt somewhat unworthy.

But I'll definitely stick around to kibitz.

[alexman]

Poll's up!

[Theseus]

What if we kept all seven, but had a quorum of 5?