Buffed AI for BNW

I'm amazed that you are able to avoid the AIs from being able to get significantly ahead on tech. In my continents game, the competitive AIs are pretty much always the 0-4 techs ahead as they hit and exceed the funnel limit on their own.

The great thing about the game I've got going now, is that it never from turn 0 onwards, felt like I was going to walk away with an easy game.

Next game which I'll start on the weekend, I'll try to test pangaea maps, since the only report of weirdness in the buffing system has come from Joosegoose playing emperor pangaea.

EDIT: You watch Maya make a swift tech come back when they get to atomic/information. They can pull back 6-8 techs in 12 turns on quick. It will be a last minute thing for them. It won't help them much, but it can be useful to have a late game diplomatic partner show up. It really depends on how long the game lasts in the endgame. Mine is going the full distance.

Excited to try out the beta 10. Which maps do you think would produce a playable endgame?

Hey Thelemon - So far it looks like any map reasonably normal map is fine because the buffing system controls how the AI progresses on an ongoing basis. If you randomise everything in the map, it can be too much for this mod to adjust. Main thing that will kill off the endgame is if the AI kills off you first. If you want a reasonable chance of making it all the way to the endgame, stay on your normal difficulty setting but compared to vanilla, suggest focusing more on the bottom half of the tech tree, build more military and focus more on population, less on science.

Joncnunn is testing this mod on islands maps and me on continents. The mod doesn't change game play except buffs it in a different way and a few other things like getting the AI to guard its workers.

Unknowns remaining in this beta V18 #10:

1. Does the buffing system change what the AI builds?
2. Does the AI improve its plots at least as good as the base game?
3. Does the buffing system change the way that the AI expands?

So far the answers are:
(1) - no
(2) - yes
(3) - not much

Would appreciate your help testing it so that I can get onto merging this with Various Mods Component and have a stable platform to work on for starting to slowly make targeted improvements to the AI that focus on improving the quality of the game experience. Already in this mod, there are some basic improvements (Smart mod, less easy to bribe etc).

EDIT:
The most difficult map type in this mod is probably pangaea, and it has not been tested by anyone I know of. The more difficult the map, the less chance you have of reaching the endgame because of AI competition. Pangaea will be most difficult because of the AIs combat buffs, multiple neighbours and because barbarians don't steal AI settlers, only other types of civilians, but not settlers. So the the AI should expand uninterrupted which will make pangaea pretty intense.

The link to beta V18 #10 with more information is here:
http://forums.civfanatics.com/showpost.php?p=13442529&postcount=398

glider1 said:

I'm amazed that you are able to avoid the AIs from being able to get significantly ahead on tech. In my continents game, the competitive AIs are pretty much always the 0-4 techs ahead as they hit and exceed the funnel limit on their own.

Click to expand...

I normally only look at techs on the turns I complete techs. A snapshot of number of techs I had the turn before reaching modern era would be me with two techs fewer. (RA came due in addition to free one from Oxford)

The population value of my capital is probably the major reason and can be seen in those screenshots.

thelemon402 said:

Excited to try out the beta 10. Which maps do you think would produce a playable endgame?

Click to expand...

I'd avoid Duel size maps. You'll probably win a domination victory before they can really come into play. (Especially if playing the Huns)

Also avoid the Amazon script and any other script that makes most of the map jungle as if would give every AI the same problems you see as the Mayans (all of this base game related)

Walls for cities likely to see combat are much more of a priority with the mod to offset the human ranged unit penalty.

Also, do NOT select ragging babs. Your human raged units have the penalty fighting them as well and it appears that the AI bonus is also given to the barbs.

glider1 said:

• Does the buffing system change the way that the AI expands?

So far the answers are:
(3) - not much

Click to expand...

#3 However, this also includes AI Settlers immune from being attacked from barbarians, which does result in the AI actually being able to found it's far flung cites on different landmasses instead of getting captured by the barbs all the time that happens in base game.
If you look at the bottom right for global view of my screen shots, you'll see how the Greeks & Spanish colonized a land mass that was previously barb country heaven. Spain has derived a lot of value out of those (you can't really tell from global view, but the Spanish colonies include natural wonders)
For Greeks its a bit of a mixed bag since they had happiness issues from over-expansion earlier. (But the AI can recover from happiness issues from overexpansion a lot better than it can losing settlers to barbs. And at no point did Greece drop behind the pack in number of techs.)

glider1 said:

You can substitute with #10, but there will still be game legacy left over from #9. I'll go and delete #9 from this forum now.

EDIT: Also, if you are using info-addict, don't focus in only on AI science. Take a look at the overall picture of AI production, GNP, happiness, tech difference etc etc, not the AI science rate (like I said a couple of posts up).

EDIT: Your combat mod could interfere with this mod I'm not sure. Also, if you are a deity player, suggest, playing this mod on deity.

EDIT: Any mod that alters handicaps is going to be a problem with this mod.

Click to expand...

The screenshot I linked to in post #404 showed the AI science rate but I assure you they were laughably behind in everything, I was a clear #1 across the board.

The combat mod is seriously nothing more than a small siege promotion for melee units and a small siege penalty for ranged units (just like mounted units). Considering there were no major wars in the game I was playing I doubt it affected anything. By giving a slight boost to melee and slight nerf to ranged I imagine it would help the AI out, if anything.

I was using no mods that affected handicaps.

glider1 said:

Alex in my emperor #10 game had 11 cities, but just lost one to China.

Click to expand...

In the emperor game I referenced, Alex had 3 cities but just lost one to China. That's kind of funny, if oddly completely different.

I'm going to scrap the game I was playing and start a new one with beta #10 on Immortal. I'm going to ignore the AI's performance in that game and attribute it to me using an old beta and/or an isolated occurrence. This'll be another Pangaea and I'll probably restart the map a few times even to get a fairly blobby mass that should help the AI out. I'd go to Deity but I consider myself a pretty weak Deity player to begin with, and part of what gets me excited about this mod is that it can hopefully produce a competitive game while retaining empire choice and flexibility, e.g. having the option to construct some wonders, whether they are worth it or not. Judging from what this mod makes important, I would figure the ToA and Hanging Gardens (and anything that raises your happiness cap) to be top tier and worth rushing, which is what I did on the previous Emperor playthrough.

joncnunn said:

Also avoid the Amazon script and any other script that makes most of the map jungle as if would give every AI the same problems you see as the Mayans (all of this base game related)

Click to expand...

I don't think I follow: what exactly is the issue the Mayans face? They've always done poor in my games (base game), and were exceptionally poor when I used this mod, with a science rate in the single digits on turn 150 standard speed. Is this what you mean, and does jungle affect AI performance?

Edit: @glider1-- Just to make sure I've been installing the mod correctly, what am I supposed to do with the CvGameCoreDLL_Expansion2 folder included in the download?

joosegoose25 said:

The screenshot I linked to in post #404 showed the AI science rate but I assure you they were laughably behind in everything, I was a clear #1 across the board.

Click to expand...

You're not really supposed to look at the AI per turn science rate with this mod but instead compare the number of techs they have to the number that you have. (Mouse over score for overall, and use trade routes available to tell how different your paths are)

joosegoose25 said:

The combat mod is seriously nothing more than a small siege promotion for melee units and a small siege penalty for ranged units (just like mounted units). Considering there were no major wars in the game I was playing I doubt it affected anything. By giving a slight boost to melee and slight nerf to ranged I imagine it would help the AI out, if anything.

Click to expand...

There's already a human ranged nerf AND an AI attack bonus in the mod.
In base cases, they change the base value of the units you see on mouse over and is why even the base values normally aren't integers with the mod even when both units are fully healthy. (Basically the mod implemented it that way instead of as a "promotion")

joosegoose25 said:

I don't think I follow: what exactly is the issue the Mayans face? They've always done poor in my games (base game), and were exceptionally poor when I used this mod, with a science rate in the single digits on turn 150 standard speed. Is this what you mean, and does jungle affect AI performance?

Click to expand...

Maya's base game issue is mostly their Jungle starting bias.

Yes, the AI has major problems with Jungle start. I think I have a screenshot where it was obvious they hadn't even researched Bronze Working to start clearing jungle from resource tiles. (It actually is a very high AI priority to improve the first unique copy of luxury tiles)

Note that even the human has some problems with a Jungle start compared to most other start types.

The AI attack bonus on the mod appears to also apply to the barbs. The Mayans have an early ranged unit as a UU and so is going to spam it. They probably didn't keep their UU in safe tiles when attacking the barbs. (Jungle another factor on this as it blocks visibility) So I would guess the barbs are using the AI attack bonus in this mod to overwhelm Mayan units.

On the folder containing the source code for the DLL, unless you can read C++, ignore that folder entirely. If you can read C++, then you could find it makes a good horror show if you have any training as a computer programmer. (The DLL has to include the base game's code as modified by the mod, all uglyness is from the base game)

joncnunn said:

You're not really supposed to look at the AI per turn science rate with this mod but instead compare the number of techs they have to the number that you have. (Mouse over score for overall, and use trade routes available to tell how different your paths are)

Click to expand...

My point was although I only posted a screenshot of the science rate, I had in fact checked all the other factors and the AI was lagging behind in each of them. They were way behind in everything. The only reason I put science rate in the screenshot was because the Maya had 6 beakers per turn at turn 154 and I thought that was pretty extreme. I can't imagine getting past turn 30 with that low of a rate.

joncnunn said:

There's already a human ranged nerf AND an AI attack bonus in the mod.
In base cases, they change the base value of the units you see on mouse over and is why even the base values normally aren't integers with the mod even when both units are fully healthy. (Basically the mod implemented it that way instead of as a "promotion")

Click to expand...

I know about this; what I did was lower the human ranged nerf so it was 15%, lower lower the AI attack bonus so it was 5%, and completely eliminate the overall human nerf. I did this because I'm not particularly strong at combat and felt I didn't need as large of a nerf. After I did this, I added a small bonus vs cities for ALL units of the warrior-infantry line (both human and AI) and a small penalty vs cities for ALL ranged units (-20% iirc). I think it works nicely to give ranged units more of an anti-personnel purpose, while melee and siege are good for taking down cities (the free cover promotion does well for this too).

My understanding with the AI attack bonus is that it's applied whenever an AI initiates combat of any sort--whether they are invading an opponent or defending their lands. If this isn't the case, I'll probably rethink the additional changes I made. I'm just not a fan of a flat bonus simply for being an AI, especially if it will apply when they are defending their lands (again, because I'm not that strong at combat) but I still want to make them dangerous on offense. Since AIs think melee units are the only way to invade a civ, I figure a bonus vs cities should help to still make them a threat on offense.

joncnunn said:

On the folder containing the source code for the DLL, unless you can read C++, ignore that folder entirely. If you can read C++, then you could find it makes a good horror show if you have any training as a computer programmer. (The DLL has to include the base game's code as modified by the mod, all uglyness is from the base game)

Click to expand...

Thanks, and good to know--pretty sure I have it installed just fine then. I can understand Lua and Sql somewhat, but that's about as computer literate as I get so I have zero desire to read into that lovely C++

Glad that you will try deity out Joosegoose that is great. I took your comments on board from a few pages ago. On pangaea/deity, the AI declare war prob is 140%. What this means is that in the very first opening war declaration, they may not actually bring a massive army like you are used to on standard deity. The war declaration might just prevent you from establishing trade.

What tends to happen is that the AI DOWS then does nothing for 10 turns or so (except for intermittent skirmishes), then may bring in a major army. They may also ask for peace depending on their assessment of threats.

So what the AI Declare War prob does is not quite right in this mod yet. The game code for AIDeclareWarProb was designed from the perspective that the AI already has that army in place (as tends to happen on vanilla deity). In this mod, the AI has to do the work itself. So the timing of declaration and army construction is a bit out in the ancient age. Later on, the AI is better at coordinating it, because by then the sync between the army construction and war declaration is easier, because the AI has already built quite a few necessary units.

So I could reduce the AIdeclareWarProb back down to 100%. Then the AI would only declare once it has an army already. However it also means that they won't have approached the possibility of war with you as much either, which makes the game less threatening and easier.

If you turn down the combat buffs in this mod by changing the variables, that will make the game easier because of the AIDeclareWarProb setting. Because the AI is weak in combat anyway, you can use the AI's extra inclination to DOW, to take land from it and pillage it without diplomatic penalty. That is one reason why the combat buffs in this mod are set as high as they are, to make it difficult to offensively tackle AIs, even if they have not got the army in place yet.

My next game will be on pangaea too....

EDIT:
I'll give thought about what can be done about the AIs non-aggressive war declaration in the ancient age. I personally don't mind it, but I understand it is weird to have an AI declare war without immediately sending in an army. It would put vanilla players off. I could try to fix it in the code.

Beginning of Atomic Era in mine:

Turn 256 (entering Atomic Era via Atomic Theory)

Me: 59 techs. Freedom Dissenents wanting Order, overall +14 happiness #1 pop, GNP, approval, and science

Maya: 46 techs. Industrial Era No ideology. Worst pop, crops, man goods, land, and science.

Greece: 54 techs. Modern Era Autocracy with dissidents wanting Order. overall -8 happiness. Worst GNP

Songhai: 53 techs. Modern Era Order, content. +22 happiness. Worst troops

Spain: 59 techs. Atomic Era Order with dissidents wanting Autocracy. overall -8 happiness. #1 crops, man goods, land, troops

Morocco: 56 techs. Modern Era Order, content. +10 happiness.

Netherlands: 57 techs. Modern Era Order with dissidents wanting Autocracy. overall +7 happiness.

Persia: 54 techs. Modern era Freedom with civil resistance wanting Order. overall -8 happiness. Worst approval.

It seems the Demographic screen approval does NOT include ideology unhappiness.

Greece just denouced me, but with them being the only civ in the world with Autocracy, I think that actually just improved my relations with most of the AIs.

Morocco killed a city state near them.

Spain captured some Songhai cities moving their border NE. They also entered the Atomic Era a few turns before I did via bypassing some techs.

The Greek City of Corinth revolted and joined the Songhai. (Just south of the Greek main land mass)

I'm seeking Science Victory. To help keep the AIs happy I proposed International Games during the most recent Worlds Council. My tourism includes Effiel Tower. Rest of the wonders helping a tourism victory are spread out.

Screenshot attached

Look like this mod might be unable to hold up the AIs on your immortal islands game, they are starting to slip behind.

glider1 said:

Look like this mod might be unable to hold up the AIs on your immortal islands game, they are starting to slip behind.

Click to expand...

A lot of the AIs have been fighting wars that predate ideologies.

Spain's been most successful in it's wars.

Songhai lost some cities to Spain but gained one from Greece via ideological revolt.

Morocco's conquest of a Greek city state ally was very unpopular with the other AIs.

This AI loathing of each other is actually a really good thing for me since Persia is the only AI sharing my ideology,

And we have the recent ideological unhappiness. Unlike it previous games where it was my own tourism doing this, it's the combos that have resulted in so many civs only having plurality support for a single ideology instead of majority.

Edit: Beginning of Information Era update

AI on AI wars continuing, some of them making good progress against each other. My screenshot as I ended the era even includes Spain capturing Songhai's capital.

Either the Atomic Era AI bonuses are stronger than the Modern Era ones or me having fewer RA partners from most of the world going order is having affect. One of the AIs constructed Apollo before I even had the tech for it.

Songhai reached Familiar over me shortly after I reached Exotic over them. But Songhai just lost its original capital (to Spain). I'm sure this is a coincidence, but Spain's acting kind of like a human would to prevent an AI from winning cultural victory would. Spain may be after a domination victory itself.

Persia revolted to Order & Maya adopted Order

I've so far managed to keep the world happy even with my Freedom. Greece being a jerk to the other AIs on top of their Autocracy is probably helping.

Greece conquered the city back from Songhai that had earlier revolted to them.

The international games is the world project in progress (International Space station wasn't an option back then)

Turn 276

Me: 64 techs. Information Era Freedom with dissidents wanting Order. Overall +4. #1 population

Maya: 52 techs. Modern Era Order. +17 happiness. Worst land

Greece: 60 techs. Atomic Era Autocracy with dissidents wanting Order. Overall +6.

Songhai: 62 techs. Atomic Era Order. +31 happiness. Worst in most categories.

Spain: 65 techs. Atomic Era Order. +6 happiness. #1 in most categories

Morocco: 65 techs. Atomic Era Order. -2 happiness. I think they are in the middle of razing a recently captured Persian city down so it will recover soon.

Netherlands: 64 techs. Atomic Era Order. +11 happiness.

Persia: 60 techs. Atomic Era Order. +4 happiness

Oh wow, great stuff. Looks to be a really happening game. Spain has managed to expand widely and maintain its happiness.

Order again however....

In my game, ideologies oscillated widely between all three. I was forced to switch from order->autocracy->freedom. I enjoyed that.

Did I just see that correctly? You have a 37 size city as the Irq?

Here is a question
How much did you have to use food trade routes in your game? Was it an absolute requirement, or could you diversify trade route types?

glider1 said:

Oh wow, great stuff. Looks to be a really happening game. Spain has managed to expand widely and maintain its happiness.

Order again however....

In my game, ideologies oscillated widely between all three. I was forced to switch from order->autocracy->freedom. I enjoyed that.

Did I just see that correctly? You have a 37 size city as the Irq?

Here is a question
How much did you have to use food trade routes in your game? Was it an absolute requirement, or could you diversify trade route types?

Click to expand...

Yes, it's size 37 and will soon be 38, will probably be 40 or 41 when the game ends.

For most of the game I run 3 food cargo ships to my capital. (Part of my standard start is the first 3 cargo ships provide food to the capital so it can grow) Shortly after I adopted Freedom, I converted one of the food cargo ships routes to a hammer cargo ship route. The other two are still food even 20 turns later.

With Civil Service + a Hospital, my capital could still feed itself without any food cargo routes but would never grow again.

Yes, base game cargo ships are way overpowered compared to caravans.

Problem is whether that food trade route strategy is the only viable one. If so, the mod is reducing diversity, not increasing it.

EDIT: Perhaps it is for a peaceful play-style? What about wide play-styles on islands? What about warfare play-style on islands?

Small update to the ranged combat human nerf.
Beta #11 is save game compatible with previous betas

I neglected to limit the range-3 or greater penalty against units to just domain land and sea. It meant that air units fighting at range 3 or greater were getting both the ranged combat nerf and the range-3 opportunity fire nerf when firing at units. Range-3 combat against cities was not affected.

That should make air combat less futile when attacking units at long ranges.

I just finished a classic emperor level quick game as Arabia (plenty of oil) using only the modified timed VC in this mod that allows AIs to build the SS for points rather than ending the game. Dido built the SS one turn before me, and I was also beaten by 5-10 turns to a timed victory by her. For the first time ever, I tried to use XCOM units in an attempt to capture a 180 strength city to win enough points for a timed victory. The XCOMS are a great addition to the end game, a lot of fun. The biggest mistake I made was that I tried to build a huge cash economy to pay my way to a rapid victory, but Alex controlled the WC and managed to embargo my critical trade partners. It made it difficult for me to reassign my trade routes profitably enough. He also forced me to switch ideologies to stay happy which was a big impost on my economic strategy as well. At times toward the end, Dido was teching on her own without the tech funnel science buff, just the passive buffs and a bit of adaptive buff. She ripped through Assyria and controlled an enormous state managing to stay happy as well. I actually think she deserved to win the game. She played it really well (with a lot of buffing help). I finished 0-4 techs behind the major competitive civs.

glider1 said:

Problem is whether that food trade route strategy is the only viable one. If so, the mod is reducing diversity, not increasing it.

EDIT: Perhaps it is for a peaceful play-style? What about wide play-styles on islands?
What about warfare play-style on islands?

Click to expand...

The 3 food cargo ships running to the capital is specific to Tradition which I independently invented in one of my first BNW games (it's so obvious that I suspect many long time Tradition players also independently invented it) is designed entirely around synergy with the Monarchy sub policy of Tradition.

Self founded Wide Liberty empires don't have the global happiness for a super capital and in addition it takes more than 6 self founded cities to reach the point in which growing the capital 1 pop point boosts city connection income more than another city growing. (Maybe 10 cities?) From what I've seen they use food routes in the other direction, 1 route each to one or two of their newest self founded cities which will be cancelled when it comes up for renewal.

Most warfare tends to be in the keep the cities that are useful but raze the ones that aren't rather than slash and burn everything to the ground and refound in the correct locations.

Doing this, the Tradition tactics seem best (including the 3 cargo ships to the capital). Of course doing this you'll want to build zoos in addition to having a religious follower happiness belief in your cities.

(For the slash and burn everything and refound, liberty tactics seems to work best, including supplying food to the new city one or two hexes away from the old one.)

And my Irq game is now complete. Science victory turn 314 (1888 AD). Abe Lincoln level. My capital reached size 41.

International Space station was a waste and may have slowed down my victory as I actually completed the tech tree. I'm thinking of making this project available with Rocketry instead of Satellites.

Most of the AIs built at least 1 space ship part, but none more than 3. (I guess you can't actually force the AI to build the space ship parts even when they have the tech. It's still a major improvement from the AI not reaching any space ship tech part in time to build any.)

Maya actually managed to catch up in techs, they were only missing 1 tech when the game ended.

The Songhai who lost their capital to Spain while making it to the Information era was 4 or 5 techs short of completing the tech tree,

All other AIs at a minimum completed the tech tree.

The Greek AI had 10 copies of future tech, they lost a couple of cities, but with them being small far flung cities it didn't hurt them any.

There was officially a 3 way tie towards cultural victory as well with me & 2 AIs both having reached influential over the Maya. At the end of the game nobody was suffering ideological pressure not even the Greek with their autocracy.

On the Battleship nerf, while it's a good idea when they first come out, it makes them ineffective against cities with maxed out defense buildings in the information era. Right after I saw that Greece had built a Rocket Artillery inside that maxed out city, I gave up and sent my fleet back home. (Greece was willing to hand that city over to me for peace though simply because they were fighting so many other civs.)

This may have been in factor in how earlier the Spanish AI conquered quite a few Songhai cities but in the late game Greece was successfully able to defend itself against 4 AIs at the same time.

I'm not really concerned about the Greece AI getting 10 future techs ahead at the very end, it doesn't do anything other than raise city defense values.

It sounds like you had a really good game. I'll take a note of what you say about battleships. At range-2, their nerf is no different than the basic nerf. At range-3 the nerf is potentially significant. Was your problem that you could not get battleships into range-2? Should I cancel the entire nerf to battleships in the information age? Did you bring carriers and missile cruisers? Would that have helped?

I'm trying out a change to the adaptive buff.

For the three yield types (gold, hammers, science), instead of this equation:
buffedYield = totalPlayerYield / totalPlayerPopulation * this AICityPopulation

I'm trying:
buffedYield = totalPlayerYield / totalPlayerLandArea * this AICityLandArea

LandArea = number of plots available to a city.
TotalLandArea = sum of land area.

So instead of the player having to maximise food to minimise the buff (and therefore the food trade route strategy),
the player would maximise land area to minimise the buff.

I think this might help wide more than tall, which I think could be good for diversity.

EDIT:
Any thoughts? In your last game, how did you do in land area relative to the other AIs? (you can get a rough idea in the score breakdown)

glider1 said:

It sounds like you had a really good game. I'll take a note of what you say about battleships. At range-2, their nerf is no different than the basic nerf. At range-3 the nerf is potentially significant. Was your problem that you could not get battleships into range-2? Should I cancel the entire nerf to battleships in the information age? Did you bring carriers and missile cruisers? Would that have helped?

I'm trying out a change to the adaptive buff.

For the three yield types (gold, hammers, science), instead of this equation:
buffedYield = totalPlayerYield / totalPlayerPopulation * this AICityPopulation

I'm trying:
buffedYield = totalPlayerYield / totalPlayerLandArea * this AICityLandArea

LandArea = number of plots available to a city.
TotalLandArea = sum of land area.

So instead of the player having to maximise food to minimise the buff (and therefore the food trade route strategy),
the player would maximise land area to minimise the buff.

I think this might help wide more than tall, which I think could be good for diversity.

EDIT:
Any thoughts? In your last game, how did you do in land area relative to the other AIs? (you can get a rough idea in the score breakdown)

Click to expand...

With the pop size 7 AI city strength being above 150 defense value I couldn't afford to have Battleships end their movement within 2 hexes of the city.
I'm afraid my designated military unit city was too busy building a spaceship part during this war. (And I was saving the gold to cash buy the last part) My first bomber (really a stealth one) wasn't ready until one turn after the war ended.

I had Destroyers with City Raider III in the area, but at no point did I get the city weak enough for them to move in. (They would have taken about 33 points of damage for reducing the city by about 25 points and then one shot by the city.)

Hum, I wonder why Battleships don't upgrade to Missile Cruisers? Missile Cruisers have the same range as a Battleship but much more powerful combat and bombard strength. For that matter, I'm not seeing where in the XML the normal unit upgrade path is set at all.

If a Battleship upgrade to Missile Cruisers can be modded in that would be preferred (and eliminate the information era too ineffective battleship bombardment issue.) I'm not aware of any UUs replacing either of these units so if the correct file can be found, that's only one upgrade needing added.

I was also trying to find the XML file in which the base food & hammer yield from land & sea cargo routes were set but wasn't finding that to experiment with bringing cargo ships into balance with caravans.

Land Area: I probably ended the game 5th or 6th in that.

Songhai was last (reduced to an involuntary OCC by Spain)
Maya's was low due to slow as molasses start.
Greece lost one of it's colonies (to Spain) so it may have fallen below me.

Let me know if you would like to try the land area buff concept for your next game. I'm so far not finding any problems with it. It would make it less necessary for food trade routes.