Making Cradle 3+ fully compatible with the Apolyton Edition

[Kull]

Upgradeable Gov-Specific "Special" Unit issues (con):

"Pezheteroi" to "Praetorians": The problem here is that each unit individually exists for a short period, and the length of the upgrade chain (L6 Republic to L10 Theocracy) isn't much better. The solution involves adding the Swordsman to the chain (thus Swordsman>Pezheteroi>Praetorians) and shifting his starting Advance to Copper Smelting, thus extending the upgrade range from L3 to L10. Moving the Swordsman makes a LOT of sense, because swords definitely utilized metals, and those were not the materials used by the Neolithic weapons of the L1 timeframe (largely flint, obsidian and fire-hardened wood).

However, we still need an offensive unit for the early game and that role is now filled by the new Club Warrior (previewed in Post# 370). Although the base sprite resembles a feather clad warrior from the Americas, the feather headdress also has antecedents in the Middle East, as depicted in ancient Ur and the later Sea People warriors seen on the wall carvings of the Egyptian Temple of Medinet Habu. As the Swordsman replacement, the Club Warrior is first available with L1 Agriculture, can be promoted to Elite, and is does not disband until the Oligarchy government.

- Files changed: Units.txt, UnitBuildLists.txt, elite.slc, Advance.txt, Great_Library.txt, gl_str.txt, 2D unit art (multiple), sounds.txt, newsprite.txt, updater3.slc & Tech Tree

[Kull]

Elite Unit Upgrade Bug plus revised times and costs:

1) There's a bug which happens when the same Advance triggers a "Feat of Wonder" AND an Elite Unit upgrade. For example, when L12 Gunpowder was researched, you saw the Feat but none of the message boxes which allow the player to review the various Elite units eligible for upgrade. And the same bug hit L18 Mass Production, which also had both a Feat and an Elite upgrade. The solution was to move the two Feats of Wonder to L13 Cannon Making (a better link anyway, since the feat grants a bonus when attacking city walls) and L16 Industrial Revolution (the Feat is a bonus to Production, and thus equally relevant to this Advance)

2) Elite upgrades followed a single path, which often produced a nonsensical result. For example, the L9 Dark Age promotion upgraded the Elite Legion (Attack 35 & Defense 25) to the Elite Pikeman (A25 & D20). Not only is that a statistical downgrade, but it costs the player 1800 Gold per unit. Talk about "insult to injury"! Accordingly, Cradle 5.1 has altered this progression so now there's a "Defensive Unit" sequence (Spearman>Hoplite>Pikeman>Halberdier>Infantryman) and another for Offense (Club Warrior>Hypaspist>Legion>Man-At-Arms>Arquebusier>Infantryman). The sequences merge with the first modern riflemen at L15 Flintlock, and proceed in a single line from that point forward (Infantryman>Machine Gunner>Marine>Hover Infantry)

3) Elite upgrade costs were also quite exorbitant, and in most cases it would cost much more to upgrade units in sequence instead of skipping to the final upgrade. For example, the string for "Elite Hoplite" in Cradle 5.0 is Hoplite>Legion>Pikeman>Halberdier>Infantryman and the total cost of doing each upgrade in sequence is 7600 Gold (1000+1800+1800+3000). However, going directly from Elite Hoplite to Elite Infantryman is only 4900. That's obviously a problem, as there shouldn't be financial incentives for skipping upgrades. Accordingly, all of the upgrade costs were revised downward (roughly by 33%) and for every string, the cost of upgrading is exactly the same, whether upgrades are skipped or not.

- Files changed: Advance.txt, feat.txt, feats.slc, feat_str.txt, updater3.slc, updater4.slc & Great_Library.txt

[Kull]

Wonder Unit Upgrade issues:

As described in Post# 216, the Cradle Wonder Units in Cradle v5.0 upgrade automatically at L14 Age of Reason and again with L16 Electricity. There are some issues, however:

1) "Age of Reason" Upgrade Slic Issue: When my v5.0 Rome campaign reached the Age of Reason, I discovered the "upgrade" only works when the units are located on unowned or "Roman-owned" tiles. All the units which were in foreign territory did NOT upgrade! In addition of course there is no associated message, which is disconcerting. Eventually the problem was traced to a single line in the wonderunits2.slc file. Extensive testing confirms that the fix works. In addition, a pair of "upgrade notification" messages have been added.

2) Brigadier & Colonel attributes: All four of the new units were given the "Plunder unit creation" capability, but that's too many. It's OK for the Field Marshall & General, since most of those were "Great House" units which already had this ability prior to the upgrade. But testing shows that it's overpowered for the Brigadier and Colonel to get it too, so it'll be removed and replaced by different attributes:
* Brigadier: Gains CanHarvest (the collection of terrain gold when fortified on a tile)
* Colonel: Acquires AlwaysHeal (the ability to heal in the field even if the unit moves that turn)

3) Hernan Cortes Upgrade: By the time Magellan's Voyage is complete, the player has probably researched "Age of Reason", or if not, it will happen very soon thereafter. Cortes is scheduled to upgrade to Brigadier when that happens, but if missed, he'll remain in-game forever. Neither situation is ideal, so his Brigadier promotion has been scrapped and he'll remain as is until the upgrade to Colonel (the existing plan for US Grant).

4) Genghis Khan Upgrade: The Mongol Empire soon broke up into multiple "Khanates", two of which lasted into the late 1700's. Accordingly, it's not unreasonable for Genghis to remain in-game until the final Wonder Unit Upgrade at L16 Electricity, especially now that his Victory Enslavement attribute has been removed. Keeping Cortes, Genghis & Grant around a bit longer adds some variety, which is not a bad thing.

5) Sun Tzu Upgrade: Changes from "Field Marshall" to "Brigadier". Only Great House Wonder Units (and the Great General) should have the FM promotion.

- Files changed: Units.txt, plunder.slc, wonderunits2.slc, info_str.txt & Great_Library.txt

[Kull]

Buildings & Increased Production:

Seven of the Buildings in CtP2 provide a Production Percentage Increase to each city, which reaches 110% by the end stages of the game (see attached). By contrast, Cradle 3/4/5 has 13 such structures which max out at 155%. That has a very powerful effect on accelerating the overall pace of the game. In particular, by the early Middle Ages (L11 Mills) Cradle cities have a 60% increase to production, compared to 10% in AE. During test campaigns it was not unusual for AI civs to construct every available building, even during the early and middle periods, leaving them little else to do. Accordingly, Cradle 5.1 has reduced the increase in the four earliest structures so that by L11 Mills, the cumulative increase is only 30% (half that in v5.0) and just 115% by the end of the game.

- File changed: Buildings.txt

[Kull]

Various Changes and Fixes for v5.1 (con):

Post# 371 summarized 10 of the new changes in v5.1, and those continue here:

11) The "Traits and Golden Ages" entry in the Concepts section lists the per-civ traits granted by the Feat system. I mention this in the "Feats" entry, but didn't provide a link. Fixed. File changed: Great_Library.txt

12) Appian Way image: The 2D graphic for the "Appian Way" Wonder is a modern picture of the French Arc de Triomphe (there's even a car in the background, lol), which is totally inappropriate for this Classical Era wonder. Instead, we'll substitute a photo showing an actual section of the Appian Way (attached). File changed: UPWP111L.tga

13) Pikemen Unit name: Cradle has a "SpearMAN" unit which eventually upgrades to a "PikeMEN" unit. In addition, there are also "SwordsMAN" and "InfantryMAN" units. Having a single unit referred to as MEN when all the others use MAN looks and sounds weird. Rather than change the underlying code name, the visible Pikemen text will be altered to "Pikeman" so the four units match. Files changed: Great_Library.txt & gl_str.txt

14) Pikeman Sprite: When the sprite was rebuilt, the Shield locations weren't set properly and "switch sides" to the Left when the unit moves R or L. There's a similar issue on the battle board where the shield repeatedly shifts behind the unit and then moves to the front, all of which is rather disconcerting. Accordingly, the sprite was rebuilt using the Halberdier's shield location settings, and the issue has been resolved. File changed: GU117.spr

15) "Upgrading" Concept: The Upgrading Concept entry in the GL did not mention the Pikeman-to-Halberdier upgrade which happens with L12 Gunpowder. In addition, the 2D art showed a single star, which has been replaced by a new graphic showing a sequence of three units, Spearman-Pikeman-Infantryman. Files changed: Great_Library.txt, uniticon.txt & UPCP087L.tga

16) Halberdier Upgrade Issue: The Halberdier unit becomes available with L12 Gunpowder and expires with L13 Cannon Making, a very short period indeed. It also temporarily breaks the chain of Defensive units which began with Spearman, and leaves a gap until the discovery of L15 Flintlock. This was clearly an error (mine), but is easily solved by changing the Halberdier's expiration to Flintlock. File changed: Units.txt

17) Khalid ibn al-Walid: When this unit was changed from Muhammad, all the stats stayed the same. That's fine for most of them, but the movement value of "2" should be more in line with that used by the other mounted Wonder Units of this era, so accordingly it will double to four. File changed: Units.txt

18) Great Library Expiration: The Wonder currently expires with "Greek Fire", but the Naval Advance restructuring project moved that in from L12 to L10. Instead, the expiration will change to L12 "Printing Press". Although the actual structure was destroyed earlier, the idea is that while a "Great Library" was extremely valuable when books and scrolls were written by hand, this new technology makes them much more common. Files changed: Wonder.txt, Great_Library.txt, & Tech Tree

19) Wonder Expiration Advances: The Wonder entries in the GL tell you which Advance will cause them to expire, but there isn't a corresponding mention in the Advance sections. The game pops up a notification when the Wonder expires, but that's after-the-fact, so it's hard to "plan ahead" and there will be times (especially with "Happiness Wonders") that pursuing a particular Advance could cause problems. Especially since as many as three Wonders can expire with the discovery of a single Advance! There are 36 Wonders with an expiration, and the Great Library has been updated accordingly. File changed: Great_Library.txt

20) GL Error: The Great Library says the Caliphate Advance enables building of the "Mecca" Wonder, but it's actually the "House of Saladin". Fixed. File changed: Great_Library.txt

[Kull]

Various Changes and Fixes for v5.1 (con):

21) New art for "Ancient Age" and "Classical Age" Concept entries: The Ancient Age had a modern picture of the ruined Parthenon (which is actually a structure from the Classical Era) while the Classical Age used a still from the CtP2 opening movie. Replaced the first with a view of the Ziggurat in Ancient Uruk while the second is now a painting of Classical Athens and its surroundings (attached). Files changed: anc.tga & clas.tga

22) TIMP Costs in the Great Library: Although most of the TIMP costs have been altered in Cradle 5.0, I forgot to update the GL records for Air Bases, Advanced Undersea Mines, Drilling Platforms, Early Mines, Farms, Fisheries, Forts, Maglevs, Mega Undersea Mines, Nets, Ports, Railroads, Stone Roads, Shipyards & Undersea Tunnels. Fixed. File changed: Great_Library.txt

23) Wonder Unit Enslavement: The Great Library entries state that Wonder Units "Enslave", but it's actually the "Victory Enslavement" attribute. Changed the text to "Enslaves enemy prisoners in Battle". File changed: Great_Library.txt

24) Amphibious Assault Concept: Units with this attribute can attack directly from ships (as the name implies), but they can also use their movement points to land, move, and attack. None of this is explained in the Great Library, so a new "Concept" Entry has been created to describe how it works. Files changed: Great_Library, gl_str.txt, uniticon.txt & concept.txt

25) Reform City pop-up message: When a city is reformed, a message appears which tells you that the city is "no longer under foreign influence". Completely absent is a notification that on the next turn the city will experience significant unrest (-5 happiness) as a result. And should the player kill the unit which made the conversion, the effects will be even worse (an additional -3 to happiness). The "Reform City" description on the "Orders menu does mention part of this, but it's also incorrect (stating the effect lasts for "one turn" when it's actually four!) Additional information has been added to both the pop-up message and the GL entry. Files changed: Great_Library.txt & info_str.txt

26) Plunder Unit Shield Cost: The "shield cost" (the amount of production points needed to build a unit) escalates dramatically as the game progresses, but a similar escalation was not applied to the Plunder units, in particular those which are supposed to disband in order to assist production in a city. Plunder I (Goat) is 100 points, which is OK for the early era in which it appears. For example, the Goat and Spearman are roughly contemporary and the ratio there is 100 points to 160. Plunder III (Cow) first appears during the Medieval period so a similar ratio is 300-to-650 (Halberdier). It's even worse with Plunder IV (Truck) with a ratio of 400-to-1400 (Machine Gunner). The bottom line is that Cradle 5.1 will increase the shield costs of the Cow (from 300 to 400) and the Truck (from 400 to 600). File changed: Units.txt

27) Slaver Unit Card: The Slaver unit card had the "red background" which is reserved for Ranged units, so a new card was created with the grey background used by civilians and agents (to include a drop-shadow effect). Files changed: UPUP004L.tga, UPUP004A.tga & UPUP004B.tga

28) Shakespeare's Theatre GL Link: There was a typo in the "Vernacular Prose" GL entry for Shakespeare's Theatre, so the link between the Advance and the Wonder did not work. Fixed. File changed: Great_Library.txt

29) Undersea Mines: In Cradle 5.0, these can only be placed on Continental shelves, but that's an extremely limited area. It makes more sense to include "Shallow" water terrain (but not Kelp or Reef). In basic CtP2, all the undersea mines were associated with undersea cities and thus restricted to deep ocean-type terrains, but this particular TIMP demonstrates a precursor technology, so it make sense to extend the availability to Shallow Ocean tiles. Files changed: Great_Library.txt & tileimp.txt

30) Destroyed TIMP Bonus: Cradle includes a slic file (from Pedrunn) which gives back 1/5 the value of a TIMP that's being destroyed while being upgraded (f/e going from "Farm" to "Advanced Farm"). It also gives the player anywhere from 0 to 1/3 the value of a tile that's being pillaged. The problem is that I didn't change the values of each TIMP in the slic file, so in most cases the "original cost" numbers are too low, sometimes MUCH too low. Fixed. File changed: tilerefund.slc

[Kull]

Various Changes and Fixes for v5.1 (con):

31) Improved art for the L4 Drama & L2 Astronomy Advances: Cleaned up the outline and added a drop-shadow effect for the "Greek Mask" which is the centerpiece of the Drama 2D pic (attached), and made a similar change to the "Babylonian Star Disk" which is used by Astronomy. Files changed: UPAP217L.tga & UPAP274L.tga

32) Arquebusier & Infantryman Great Library entries: The GL describes both of these as "Ranged Units", but they are categorized as "Attack" and those values are much higher than the Ranged (which means they will always default to combat in the front row). Descriptions have been changed to "Infantry". File changed: Great_Library.txt

33) Arquebusier Unit Card: As noted above, the Arquebusier is Infantry, not "Ranged", so the background color of the unit card needs to shift from Red to Yellow/Brown. Interestingly, the "elite" card already used the correct background color, so it was an easy fix. Files changed: UPUP161L.tga, UPUP161A.tga & UPUP161B.tga

34) Unit Attack & Defense Values Descriptions: The values themselves (Attack-Defense-Ranged-etc) are listed in the per-Unit database section of the GL, but they aren't explained anywhere. There is a "defense" concept, but it's primarily focused on city defenses, not units. There is a GL entry for "Ranged Attack", and although that description is unit-focused, it's the only one. Instead of adding a number of new entries to the Concept Database, an edited version of the unit attributes description from the Game Manual has been inserted into the existing "Units" Concept entry (which previously told you exactly nothing). File changed: Great_Library.txt

35) Caravan/Merchant Trader/Freight Descriptions: A Merchant Trader is worth TWO Caravans and Freight is worth THREE, but it doesn't say this explicitly in the Great Library. The number of Caravans a particular unit creates is based on a unit having the "IsTrader" flag AND the value of the "MaxMovePoints" flag. Thus, for example, the Merchant Trader has a "MaxMovePoints" value of 2, and thus generates two Caravans when it is built. All three unit descriptions were updated to include this important information. File changed: Great_Library.txt

36) Marketplace Upkeep: One of the Cradle 5 changes was to reduce the cost of all "economic" buildings (i.e those which "give Gold") to zero, in keeping with the changes made in Cradle 4. However, I missed one - Marketplace still has Upkeep of 8, and will also be reduced to Zero. File changed: buildings.txt

37) Military Unit per Slave: Spontaneous Slave Revolts are possible when there aren't enough military units in a city. The Cradle default is 12 slaves guarded by 1 military unit (vs. the AE default of 3). On the one hand, that's a good way to help reduce the number of revolts, which are a proven source of CTDs. However, that Cradle number was established before the Source Code team changed the Enslavement code, allowing for distribution among multiple cities, rather than accumulating only in the nearest one. In addition, the AI does a much better job of maintaining city garrisons. Accordingly, I'm going to drop the number from 12 to 7, which will force the Enslavement-focused Human player to maintain a larger garrison in his most heavily slave-infested cities. File changed: Const.txt

38) Trade Route Bonuses: The bonus per international trade route is very low (10) especially in light of how few of these trade routes there are (most are internal to the nation) and especially when the water trade route (which is almost all of them) grants 15. Accordingly, we're going to swap the bonuses so that "international" is 15 and "water" is 10. Affects 3 Wonders: Lighthouse of Alexandria, Magellan's Voyage & East India Company. File changed: Wonder.txt

39) Ramayana Wonder expiration: The Ramayana is enabled by L4 Philosophy and expires with L9 Dark Ages (along with two other Wonders), which is a fairly short period. A better expiration would be L11 Islam (which ties in to the conflict between Hindus and Moslems that existed in both ancient and modern India). Files changed: Wonder.txt & Great_Library.txt

40) Spartan Philosophy expiration: This Wonder currently expires with "Age of Reason", but a lot of slic is happening with that Advance, so I want to minimize the number of "pop-up" windows associated with it. In addition, that is a VERY late expiration for an Ancient Military training system that was certainly finished by the time of Alexander the Great. Currently it becomes available at L5 and lasts until L15, ironically the same as the Cyrus Cylinder. Since we've changed that Wonder and it now expires with L10 Feudalism, that's a good termination Advance for the Spartan Wonder as well. Files changed: Wonder.txt & Great_Library.txt

[Kull]

Various Changes and Fixes for v5.1 (con):

41) Emancipation Act effects: This Wonder has the worldwide effect of causing all Slavers, Slave Traders, and Abolitionists to disband. However, it doesn't say this anywhere in the Great Library. Words to this effect have been added to the Wonder and all three unit entries. File changed: Great_Library.txt

42) Civ Traits: Civilization-specific traits are awarded via Feats to all civs on Turn #5. Quite a few had to be adjusted because their 50% TIMP cost reductions did not reflect changes made in that area. In addition, "Veteran Unit" bonuses are now awarded to several civs for two of the new units (Club Warriors and Early Galleys), while some others were tweaked for historical accuracy (f/e Rome gains veteran Quinqueremes while Phoenicia - which never built them - loses the bonus). Files changed: traits2.slc, feat_str.txt & Great_Library.txt

43) "Rational Thought" Advance: "Rational Thought" is an Advance which is granted (and removed) as a necessary part of the Slic coding in Cradle 5. Occasionally the player will see this Advance in the list of civ-owned Technologies, but it should be ignored as it does nothing else and will be removed in the following turn. Players are curious of course, so a GL entry describing this process has been added to the "Advance" category. As an aside, the actual code name is "Subneural Ad". Although not part of the visible game, slic coders use it to perform various "Advance-linked" activities (like building Visible Wonders). Files new and changed: Rational.tga, uniticon.txt, gl_str.txt & Great_Library.txt

44) New Gilgamesh unit: When the human player builds the Dynastic House, instead of receiving a "Great King" unit, you will get "Gilgamesh. The two units are nearly identical, but Gilgamesh can generate Plunder (which the AI doesn't know how to use) while the Great King can perform siege attacks (which only work when deployed by the AI). Files new and changed: Units.txt, uniticon.txt, buildingunits.slc, plunder.slc, wonderunits2.slc, gl_str.txt, Great_Library.txt, Tech Tree, 2D art files (AgkingL/A/B.tga)

45) New art for "Medieval Age" and "Renaissance" Concept entries: The Medieval Age used an image of the Hagia Sophia basilica (somewhat duplicative with the Wonder), while the The Renaissance Age and the new Sistine Chapel Wonder were using the same picture. Replaced the first with a feudal-era Castle (see attached) while the second is now a painting of Medici-era Florence. Files changed: medie.tga & rennisan.tga

46) City Celebration Concept: For players of Civ-series games, it can be confusing when you receive the "City Celebration" messages, because you assume it comes with some benefit such as improved growth, extra gold, or more production. However, it actually does nothing. Accordingly a new "City Celebration" entry has been added to the Great Library (in the "Concept" section), which explains this to the player. Files changed: Great_Library, gl_str.txt, uniticon.txt & concept.txt

47) "Bonus Gold" Wonder Benefit: In general these values are pretty low, especially in the case of Wonders that grant only this particular benefit. For example, The "Temple of Solomon" gives 75 extra Gold per turn, but that's the same value as one would gain from an average Trade Route. It takes 2100 shields to build the Temple, but you could have 5 Caravans at the same price. Anyway, there are 6 Wonders with the BonusGold attribute, and all of them will provide more Gold than before. Files changed: Wonder.txt & Great_Library.txt

48) Pezhetairoi Advance: To ensure a "gap-free" chain of Advances between Swordsman and Praetorians, the Pezhetairoi unit will now be available with Oligarchy instead of Republic. In addition, the unit name was spelled incorrectly, and that has been fixed (from Pezheteroi to Pezhetairoi). Files changed: Feat_str.txt, gl_str.txt, Units.txt, Great_Library.txt & Tech Tree

49) Merchant Specialist Advance: With the Pezhetairoi unit moving to Oligarchy, a corresponding change is to shift availability of the Merchant Specialist from Oligarchy to Republic. As a reminder, this is not a unit but rather the ability to use the City Window to assign citizens to a specific resource gathering role, in this case that of Gold. Files changed: Pop.txt, Great_Library.txt & Tech Tree

50) Elite Units upgrade in Colonies: As described in Post# 303, Elite units can be upgraded when they are stationed in Cities, Forts, or Airbases. That now includes Colonies. Added 6 lines of code in 2 slic files and edited several entries in the GL so the player knows about this. Files changed: updater3.slc, updater4.slc & Great_Library.txt

[Kull]

Shadow files for Organizer, Captive & Assassin Sprites:

The Organizer, Captive (Plunder II) & Assassin units (all taken from the Forever Future mod) did not have shadow files, and that detracted from their on-map appearance. Fortunately, the Cyber Attorney is another female sprite which has a similar walking motion, so I decided to utilize its Shadow files, and to reposition them so the feet match the location of those in the 3 FF sprites. Accordingly 55 Cyber Attorney shadow images were modified, and all three female sprites now have on-map shadows during their Movement animations.

Idle is the other highly visible animation, and even though the movements are not similar (either to the Cyber Attorney or each other), the number of images is small and there's not much actual motion. Accordingly, 2 of the sprites only required minimal editing of 4 images. By contrast, the Organizer was a real pain, and the shadow file (fortunately only one) had to be created from scratch. The end result however, is that all three sprites now have on-map shadows and all of them look much better (see attached). Shadows were not created for Attack & Death (that required 100's of additional files) but those animations appear briefly and infrequently, so the lack is not really noticeable.

- Files changed: GU128.spr, GU 129.spr & GU203.spr

[Kull]

Shadows for Trawler & Advanced Trawler TIMPs:

All the vessels in the two levels of Trawler TIMPs look a bit artificial when placed atop the Ocean Tiles. There was a similar problem with all the new Goody Huts, but adding shadows made them blend nicely into the landscape. Those were created manually, but it was a bit easier (and more natural) using the "Drop Shadow" effect. Long story short, that was done for all 9 of these vessels, and while the effect is subtle, they definitely look better. In the attachment, the 4 vessels on the bottom have shadows, those on top do not.

- Files changed: gtset555.til & gtset565.til

[Kull]

AI doesn't build Club Warriors:

Adding new units to the game is really a two-part exercise. Creating the sprites and all the associated documentation and artwork is time consuming, but equally important is to test them in-game to make sure the AI is using them as intended. And the vast majority of the time, it's not an issue. But in the case of the new Club Warrior (see Post# 370), tests showed the AI was not building them. I made several of the usual changes to costs and stats, but to no effect. For historical and design reasons discussed earlier, this new unit had replaced the Swordsman (exact same position in the Tech Tree), so for comparison purposes I ran some tests in the Cradle v5.0 to see why the AI built Swordsmen but not Club Warriors. And to my surprise, they didn't! I'd never noticed this in earlier test games, probably because Swordsmen are a common Barbarian spawn unit and thus were seen frequently on the map. But the AI civs never built them.

I looked at every file that seemed likely to have an effect, and made tweaks or adjustments to many of them, but in the end had to confess defeat. For whatever reason, the AI was more interested in Ancient-era defensive units like the Spearman, ranged units such as the Slinger and mobile units like the Javelin Cavalry. Fortunately this soon changes, as the AI is quite happy to build (and use) Classic-era Offensive infantry such as the Pezhetairoi, Hypaspists, and Legions. And in terms of "resource usage", that actually makes a lot of sense, since the early game is more a period of exploration and expansion than conquest. In addition, the AI is quite willing to use Spearmen as attacking units, so while that's less efficient than using Club Warriors, it means they aren't operating with a purely defensive mindset.

All told, this was an interesting but somewhat disappointing exercise, except for one wholly unexpected side effect. While digging into the problem, one of the files I looked at was Goals.txt, which is one of 11 text files in the aidata folder. As noted earlier (see Post# 80), this is where you find most of the user-editable files which control AI behaviour. This particular file defines the various goals that the AI can pursue, which are then implemented in Strategies.txt. One benefit of spending literally years looking at the various CTP2 files is that you begin to recognize code words and acquire an innate understanding of what they mean. So for the first time, I noticed that several of the "goals" are assigned to units with the "HasZOC" attribute. Specifically: Siege, Chokepoint, Patrol, Harass City, Retreat & Paradrop. Most of those are things the AI is HORRIBLE at in Cradle, and on further reflection I could see why. In CtP2-AE, almost all the military units have a "Zone of Control", but those in Cradle do not.

Simply removing that attribute wouldn't be enough, however since now the code would be looking at EVERY unit, including naval ones and civilians. Thinking a bit more, I replaced "HasZOC" with "CanCaptureCity", which is an even better way to make sure that it's only selecting LAND combat units. I made the changes but didn't think too much about it, especially since the AI continued to ignore the building of Club Warriors. What did change is that my next test game had an early CTD, and the crash log pointed toward the "Rally Troops Goal". Except it's not a goal per se, but rather a portion of the code within the six goals I'd altered earlier.

On a hunch I went through the "Units.txt" file looking for any unusual cases where a unit was supposed to be able to capture a city, but really couldn't. There were a few late game Naval Units (Morey Striker & Kraken) so I changed those but they aren't available in the early game and thus wouldn't be causing the CTD. HOWEVER. All the Cradle Militia units had the attribute, but they cannot move and thus could never capture a city. They also don't have a ZOC, so what if the modified Goals.txt was trying to round them up and make them move? So I removed "CanCaptureCity" from all of them, and then re-ran the last test game. No CTD!

Which was good of course, but it was a CTD of my own creation and hadn't solved the original problem, so....who cares? Right? Wrong!

[Kull]

AI Aggression:

Since CtP2 was first released back in 2000, players have been tweaking the game settings in order to improve the AI's aggressive capabilities. This culminated in the development of "Frenzy", a slic file which gathered waves of enemy units and sent them surging toward the human player. Unfortunately it was a blunt force mechanism, and had the side effect of crippling Diplomacy since, almost by default, all the AI civs hated the human player. And it didn't do much to improve the AIs ability to engage in successful warfare against other AI civs.

The first Source Code version of CTP2 (released in 2011) included a significant improvement in AI performance across all these areas. I've played several campaigns using the 2011 base game, and those benefits are real - the AI is a talented and dangerous opponent. But to my deep regret, transferring those gains to Cradle 5.0 was an abject failure. Even a modified version of Frenzy didn't move the needle by very much (see Post# 148). The Cradle AI is very good at deploying Religious, Slaver, and Espionage "special attack" units (even more so than AE-2011) and does quite well at building large military stacks and using them in defense of their homelands. But when it comes to aggressive military warfare - specifically in gathering large unit stacks and using them to invade and conquer the cities of other civs? Awful. Even when a city revolted deep within an AI empire, they were usually incapable of re-conquering it.

That doesn't mean Cradle 5.0 was a complete failure. In all other respects the gameplay was quite engaging and human-player efforts to invade a powerful AI neighbor always presented a daunting prospect. However, aside from fending off pirates, pillagers, and Special Attackers, defense of the homeland was not a concern. But that was then. Things are VERY different in version 5.1!

[Kull]

AI Aggression in Cradle v5.1

In shocking fashion, the few simple changes made to the Goals.txt file have completely transformed the Cradle AI. You can see this in the attachment - more AI cities changed hands in this single playtest than in ALL Cradle 5.0 test campaigns combined! Let's go through some of them:

1) Upper Left Circle: The Phoenician civ (purple) was decimated, primarily by Carthage (maroon) between 1 and 1250 AD. They lost 13 cities and were soon destroyed a few turns later.

2) Center Circle: In the same time frame, Egypt (yellow) has conquered half the Minoan civ (light blue) in the east and is likewise capturing Shang (light green) cities in the northwest - but has lost territory to the Shang in the southwest.

3) Bottom Circle: The Indus civ (salmon) has driven a wedge through the heart of the Greek nation (light grey).

4) Upper Right Circle: One of the Mayan cities (dark blue) revolted and became the Aztecs (green) but the Aztec civ was soon destroyed by Carthage (maroon).

Looking elsewhere on the map you can see many instances of expansion and loss, affecting every civilization. Cradle 5.1 now features a dynamic and ever-changing world.

[Kull]

AI Aggression in Cradle v5.1 (con):

Taking a deeper look at one of those examples, the attachment below shows that prior to 1 AD, the Egyptians had already taken 4 of the 5 Persian cities on their continent (bottom of attachment) along with 2 from the Minoans (top)

[Kull]

AI Aggression in Cradle v5.1 (con):

Attacking neighbors on the same continent is certainly encouraging, but what about "cross-ocean invasions"? Yes, the AI performs those as well. In the attached example from a different test game, it is 1000 BC and we can see that Carthage sent an invasion force south across the seas to a Mayan-controlled island, has captured one of their cities (left circle), and is unloading troops to attack another (right circle).

The concept of "island safety" is an illusion in Cradle 5.1.