Strategy changes between versions...

[furrykef]

It's been remarked that the game can sometimes change almost completely between versions, whether it's from patch to patch or when you throw in an expansion pack. The result of course is that strategy and tactics must change to accomodate the game changes, which is fine so long as you can keep up with everything.

And that's the problem. Keeping up with everything is hard unless you're an oldbie! I've been around since Vanilla Civ3 (and Civ2:ToT and SMAC/X for that matter) but I only play the game extensively on the off occasion, and the same went for browsing through the forums here. So I don't hear about every single little change as it happens.

But some of the best threads and documents on strategy are the older ones, such as Vel's thread, and we can't just ignore them just because some parts are obsolete and some aren't. But it's hard to know just what is obsolete, and why it's obsolete.

Here's a specific example. Pop rushing in despotism has interested me, and while recently I've found some tips on the way it's used in later versions of the game, I also come across older information. One of these was an apparently infamous doc on CivFanatics...I looked through its associated message board thread to see if anybody has information on how to adapt it, if possible, to newer versions. But the last thread I found said something to the effect of, "Can we lock this thread? Recent patches change everything and the information is obsolete." Well, that's just great. Go ahead, tell me it's obsolete and leave me without a clue what makes it so.

Of course, I could always refer to the Civ3 patch readmes. So I looked through the PTW readme for "pop" and found nothing relevant. Then I looked through the vanilla Civ readme and found a bugfix (first citizen 20 shields now), a change of unhappiness from pop rushing from 20 to 40 turns, and a mysterious and unhelpful line, "Eliminated 'Despot Pop Rushing'." What eliminated it? The unhappiness change? Something else? Did it really eliminate it or did it just make it a less effective but viable strategy?

So, I propose we use this thread to document the changes in strategy and tactics from vanilla to Conquests...I'm looking more for "big things" (pop rushing) than "little things" (map making doesn't allow trading maps anymore), but anything that can have a significant impact on strategy would be fine...it'd also help anybody who, for some reason, may have been playing vanilla all this time and on a whim buys Conquests, but my main goal is to help us keep our facts straight when consulting older documents and ultimately give us a more solid "knowledge base" for this game.

- Kef

[furrykef]

OK, here's a starting point. I'll update it as we get more information here. The items I'm least sure about are marked with a question mark. Much information was inferred from the patch readme's. The points that seem to be the most important are in bold.

Civilization III (henceforth 'vanilla')

• Only one kind of Great Leader. Military Great Leaders (MGLs) can rush Great Wonders; Scientific Leaders were not introduced until Conquests. As a result, MGLs were usually used to rush Wonders rather than build Armies.
• Palace easily exploited for pre-building. The player could build it in more than one place at a time, whereas starting with vanilla 1.16f, it could not, so only one Wonder could be pre-built with the Palace at a time.
• Corruption/waste was severe. Forbidden Palace had higher priority. (?) Commercial trait didn't help fight corruption much. (when did Commercial become better?) The situation improved in vanilla 1.16f and vanilla 1.21f.
• Ring City Placement (RCP) exploit existed but wasn't yet discovered. Cities equidistant from capital suffered less corruption. (This did exist from the beginning, right?) This was effective until Conquests.
• Despot pop rushing extremely powerful. Sacrifice population by forced labor repeatedly to rush-build lots of stuff, because food is cheaper than shields in outlying cities with higher corruption and waste. This strategy was effective until vanilla 1.17f; see there for why. It was sort of the ICS of vanilla Civ 3: if you didn't use this "unnatural" strategy, you would be crushed by somebody who did. (Of course, multiplayer was only a pipe dream at the time...)
• Clear/plant forest exploit. Forest could be cleared for 10 shields, replanted, then cleared for another 10 shields. Fixed in vanilla 1.16f.
• Can trade cities with the AI. This led to some cheap AI exploits, for example, removing all the troops from a city, trading it to the AI, and recapturing the city with the removed units. Such exploits were fixed in vanilla 1.16f at the cost of no longer allowing the player to trade cities with the AI except in specific circumstances.

Vanilla 1.16f

• Corruption alleviated. Some felt it wasn't enough; corruption-fighting ability improved again in vanilla 1.21f.
• Communal corruption in Communism for number of cities is now flat.
• Cost of palace now affected by number of cities. Now moving the capital is harder, but the palace can be exploited for pre-building in a large empire.
• Palace pre-build exploit fixed. A player can only build one palace at a time.
• Size 1 cities without culture will be auto-razed upon capture. Seems to be a way to make it harder for the human player to simply take newly-built enemy cities and use them for production. (Was there maybe another reason?)
• Culture-flipping happens more frequently. This helped "culture-bomb" strategies (rush-building cultural buildings in frontier cities), but made it harder to hold onto captured cities. (Did this have the indirect effect of making propaganda more effective?)
• Clear/plant forest exploit fixed. A given forest can only be cleared for the 10 shields only once; replanting and clearing has no effect. (Can a planted forest that wasn't there before still be cleared for 10 shields?)
• AI can trade on player's turn. This was either a bug or a deliberate attempt to fight "tech-whoring" (selling techs to everybody as soon as a new tech is obtained). This proved to be too cheap and the player cannot initiate trade on the AI's turns, so vanilla 1.17f fixed it.
• AI can no longer see invisible units.
• Cannot trade cities with the AI. The player may extort cities from the AI when signing a peace treaty immediately after war, and the player may give cities as gifts (and only gifts; the AI will not even pay 1 gold for a city), but that is the extent of city trading.
• Great Lighthouse weakened. It no longer allows safe travel into Ocean squares, making this Great Wonder a serious consideration mostly just for archipelago maps.

Vanilla 1.17f

• Despot pop rushing strategy severely limited due to resulting unhappiness being cumulative (and also transferable if city is abandoned). In addition, the unhappiness penalty was increased to 40 turns until vanilla 1.21f.
• Infinite city growth with size 6 city and granary exploit fixed. (How did this work?)
• AI no longer trades on player's turn. This of course makes "tech-whoring" a viable strategy again.
• Military advisor bug fixed. The advisor previously would include Workers (and Settlers?) in its power calculations. (Did this go for the AI, too? The old "give 'em a ton of Workers and they'll declare war"?)

Vanilla 1.21f

• Pop rushing unhappiness back to 20 turns again, but heavy repeated pop rushing still not very effective. Limited pop rushing can still be effective.
• Corruption/waste situation improved. Courthouses, Police Stations, and We Love The King Day are all more effective.
• AI more hostile toward non-military units in its territory.

Vanilla 1.29f

• Seems to be mostly bugfixes and other changes and additions that don't alter strategy. However, the player can now disable the preservation random seed for a new game, thus allowing the old "if you don't like what's in the hut, reload the game" cheat for little wimps. You know who you are.

Play the World

• A whole bunch of new Civs added. Well, duh. However, new traits were not introduced until Conquests, so they were just like other civs but with different leaderheads and unique units.
• Civil Defense, Commercial Dock, and Stock Exchange improvements added. Wall Street also requires 5 Stock Exchanges instead of 5 Banks.
• The Internet Great Wonder added. Be like Al Gore and claim you have invented the Internet...
• Workers can build airfields, outposts, and radar towers.
• Probably a bunch of stuff I'm not in the mood to look up yet.

Play the World 1.01f

• Mostly unit and improvement tweaks. (Did any have any major impact?)
• Bombardment can destroy Airfields, Outposts, and Radar Towers.

There are no major strategy changes in any subsequent Play the World patch that I'm yet aware of; they are largely bugfix/tweak releases that don't change strategy.


Conquests

• New Civs, traits, Wonders, and a bunch of other stuff added. I'll come back to this post later to add the details...in the meantime, RTFM.
• Free tech with Philosophy. This idea appears to be a nod to Civilization II. It also turns out to be powerful to the point that many players almost always beeline straight for Philosophy (after getting Pottery if needed).
• Statue of Zeus added. This Wonder has been shown to be extremely powerful by pumping out Ancient Cavalry every 5 turns as long as you have ivory.
• Great Wall changed. Previously, it would strengthen city walls; now, it adds city walls to all cities. This makes the Great Wall much more powerful and no longer so much a "booby prize".
• Secret Police HQ added. While it was supposed to work only under Communism, it acted as a second Forbidden Palace even in other governments (fixed in 1.13 BETA). However, you wouldn't want to actually build it: see next item.
• Forbidden Palace and Secret Police Headquarters actually increase corruption. Not true in every case, but true often enough to make them nearly worthless.
• Scientific Leader (SGL) added. Military Great Leaders can no longer rush Great Wonders, but can rush small wonders. A direct result of this is that Armies become more desirable to build.
• Curragh added. This early-game boat is often used by human players to explore the map more quickly and contact faraway Civs much earlier than they could before.
• Barbarians much more docile. Often they will fortify unless they see a stray Settler or Worker.
• Barbarians seem to pop out of goodie huts much more often. This makes early hut-popping not so good an idea except for Expansionist civs. Thus, the relative strength of the Expansionist attribute is increased.
• Double GPT bug. Atari's hastiness backfired and gold-per-turn agreements pay double gold. This often made AI players very powerful. Fixed in Conquests 1.10 BETA.
• Corruption system overhaul. One effect is the Forbidden Palace and Secret Police HQ can actually have negative effects, especially in Communism. Building them was now seldom a good idea. In other news, Atari's HQ has gone into civil disorder and Atari has fallen into Anarchy. Players charge that Atari is switching to Fascism. The corruption system is seemingly finalized in Conquests 1.13 and matches the original planned design.
• RCP unexploit. Now cities equidistant from the capital suffer more instead of less corruption, encouraging non-ring city placement. See previous comment about civil disorder and fascism.
• Agricultural trait added.
• Seafaring trait added. However, seafaring civs did not start next to the sea reliably until Conquests 1.10.
• Traits for vanilla and PTW civs reassigned. This changes the playstyle of some civs.
• Industrious trait weakened. Industrious workers work only 1.5 times faster than normal workers, down from 2 times faster. Enslaved workers no longer work faster. Forest clearing is not faster for Industrious workers. (Presumably this applies to swamp and jungle terrain as well?)
• Unit upgrade cost increase. A direct result is that Leonardo's Workshop is much more important now. (Umm...I forgot how much the increase was...)
• Zero-range bombard added for units such as Archers. This means they will get a free shot at attacking units, boosting the longevity of the archer and making it a reasonable possibility for garrisons in dangerous territory.

Conquests 1.10-1.12 BETA

• Double GPT bug fixed.
• Corruption system improved. Sort of. (Does this fix the anti-RCP problem?)
• Forbidden Palace and Secret Police HQ are weakened. (This is true in Conquests as well, but they didn't work properly at all; they're working in this version.) They no longer act as a second and third Palace, because the distance from the true capital is still factored in.
• Seafaring trait fixed. Seafaring civs start near the sea as they should.

Conquests 1.13, 1.15 BETA

• Corruption calculations tweaked again.
• Max corruption is now 90% instead of 95%.
• Cities with Palace, Forbidden Palace, or Secret Police HQ all suffer very little corruption.
• Forbidden Palace and Secret Police HQ add 50% to Optimal City Number (OCN). This provides a large boost to both the AI and the player, and the SPHQ may make Communism more desirable. Time will tell...(EDIT: We're not sure if in practice the boost is in fact 50%.)
• Secret Police HQ works only under Communism as intended.

Not sure what version yet

• "Zero-turn war" exploit fixed. Previously, the player could declare war on the AI (via renegotaiting peace treaty, then canceling it), then suing for peace before leaving Diplomacy. Now declaring war will boot the player from Diplomacy.
• Fixed huge cities exploit. Workers cannot be added to starving cities. Previously, a city that was starving would only lose one citizen per turn, so hordes of Workers could be added to a city to make it huge (despite it having a massive food deficit) to boost score in order to "milk" the game.
• Free techs from Scientific trait are truly random. Before, they were heavily weighted toward Monotheism (Middle Ages), Nationalism (Industrial), and Rocketry (Modern). Therefore, Scientific players cannot count on getting these specific advances for free. (This was in a PTW patch, but which one?)

To be added/fixed when I'm feeling less lazy:

• Info on Jaguar Warrior rush (when did it become less effective? I think this had to do with the retreat fix, but is it the whole story?)
• Details on retreat fix for fast units (their retreat is much less reliable)
• Any other major unit changes (help me out here, guys! )
• Clean up the list

- Kef

[vmxa1]

Yes there was a proposal to redo all the key thread after C3C had settled down. It is a good idea if anyone will pick up the ball.

[furrykef]

Yeah...I think before we do, we should figure out what exactly has changed between each version, which is another reason why I started this thread. And another reason is to simply provide "a look back", so to speak.

- Kef

[DrSpike]

The more important factor regarding despotic pop rushing that changed was that the effects stacked from 1.21 onwards, and were transferable if you disbanded the city.

Prior to 1.21 regardless of the length of unhappiness you could keep 1 person content forever with a luxury or 1 MP.

Also worth adding is that the AI tech trading rates changed in all the initial patches (btw wasn't 1.07 the release version?) whilst they were finding the balance.

The 'granary exploit' you refer to was probably the Size 6 city producing a worker every 2 turns, exploiting the food box size change from 6 to 7.

And no IFE!

[furrykef]

Hmm...you sure the cumulative pop rushing effect was introduced in 1.21f? I can't find anything in the vanilla readme about pop rushing in 1.21f except for an unrelated bugfix. Perhaps it was in 1.17f, and what is meant by "Eliminated Despot Pop Rushing"?

- Kef

[DrSpike]

I wont swear to it, since it was a while back, but I thought it wasn't properly fixed until 1.21.

[furrykef]

BTW, that size 6 city/worker per 2 turns exploit was what I was referring to, but I still don't get how exactly it worked...

- Kef

[Theseus]

Great idea!!

If this gets going, it should be categorized and topped.

[furrykef]

OK, updated the post. I glossed through the patch notes in order to touch upon what I think are most of the major points, although I'm not sure of the strategic ramifications of some of the changes/additions. Go ahead and add your own refinements/interpretations/whatever.

- Kef

[DrSpike]

The tech trading on the players turn was introduced in 1.16 to combat tech whoring. 1.17 removed it, but that patch had the highest used AI tech trading rates under the original (pre 1.29) tech model. You might also want to mention the tech model change in your list.

Plus I think the release version was 1.07.

Also I thought about it some more and remember (after making posts to point out there was no point at all extending the unhappiness caused by rushing, since it did not address the issues at all with temp cities) making a post questioning why they had fixed pop rushing correctly but had also taken the step of increasing the length. I also remember a 1.17 egyptians deity game where I didn't pop rush at all, so your conclusion from the readme that the change happened in 1.17 is no doubt correct.

[MrWhereItsAt]

Originally posted by furrykef
BTW, guys, when was the "free tech with Philosophy" added? I can't find it anywhere, and it definitely deserves mentioning.

- Kef

This was introduced with C3C - right from version 1.00

[furrykef]

That's what I thought; just wanted to make sure because I couldn't find it in the manual. I thought I read somebody hinting that it was in PTW, but I probably either misunderstood what they were saying or maybe they just typed the wrong acronym. Thanks.

- Kef

[furrykef]

I feel the list still misses some of the "big picture" things, partly because what the patch readme's say tend to rarely be the whole story. Anybody have any more ideas?

- Kef

[DrSpike]

Originally posted by DrSpike
You might also want to mention the (1.29) tech model change in your list.

[furrykef]

Ah, right, sorry. What exactly was this "tech model change"?

- Kef

[DrSpike]

Well technological advancement came too quickly on the higher levels, making people spend most of the game on Deity buying tech. The AI got a bonus compared to the human so that research costs were lower. For example on Deity they were 60% of the human tech costs, which didn't change across difficulty levels

In 1.29 they tweaked it so that the AI's costs were constant, but the human's costs were dependent on the difficulty level. This meant tech progress was slower, and that the AI was meaner in selling techs, requiring more research than previously on the higher levels. It was an important change.

[Aeson]

Infinite city growth with size 6 city and granary exploit fixed. (How did this work?)

This was because the size 7 granary is twice the size as the size 6 granary. When you grew to size 7, the granary would save all the food from size 6 (20 food), you'd start the next turn with a full food box at size 6.

So in any city which had a Granary, fresh water source (or Aquaduct), and produced 10 shields at size 7, you could build 1 turn workers in. Even if your food intake was very low (it may have even worked with no food surplus at all). So you could maximize production in these cities and still get the equivalent of +10 food per turn.

They fixed this by emptying the granary when dropping from size 7 to 6. Then later (one of the PtW patches I think) fixed that so that it instead you only keep the food from the size 6 Granary when growing to size 7.

---------------

Some things for your lists:

1.16f changed the AI to never trade cities except for stand alone items with peace treaties.

1.16f modified the Great Lighthouse to no longer allow safe movement onto ocean tiles for naval units.

1.17f? Changed the AI trading to allow for them trading during the player's turn. Changed back in 1.21f? (may have been 1.16f and 1.17f respectively)

1.17f changed it so the military advisor no longer viewed Workers (and I assume Settlers) as military units.

1.21f changed the way AI view non-military units in their territory, making them much more likely to demand their withdrawl. Based on proximity to city centers and number of units.

1.21f? got rid of the ability to renegotiate peace treaties.

Other significant changes that I'm not quite sure which patch they were introduced:

Cities may no longer exceed size 256. (I think it's 256?)
Cities may no longer have workers added to them when running a food deficit.

PtW introduced the new less agressive (random) barbarians.
PtW patch: Free Scientific era techs modified to allow for any of the first tier of techs, instead of being heavily weighted towards Monotheism, Nationalism, and Fission(?).

[Arrian]

Minor correction: The weighted tech in the modern era was Rocketry, IIRC. Why I felt this worthy of a correction I don't know.

-Arrian

[ducki]

1.21f? got rid of the ability to renegotiate peace treaties.

Did they add this back or is this something different from what I'm thinking of?
I can currently negotiate w/ an AI, go to "Active", choose Peace Treaty, confirm I'm sure, and then demand stuff from the AI for my continued pacifism.

[furrykef]

1.21f? got rid of the ability to renegotiate peace treaties.

What do you mean?

Cities may no longer exceed size 256.

Umm...was getting a city that big ever a realistic idea? How the heck would you get the food?

Cities may no longer have workers added to them when running a food deficit.

Was this ever a good idea? I wouldn't want to increase the size of a starving city and just make it worse...


I'll add the more important ones soon; I'm kind of in a rush at the moment so I can't just yet.

Thanks guys!

- Kef

[Aeson]

Minor correction: The weighted tech in the modern era was Rocketry, IIRC. Why I felt this worthy of a correction I don't know.

It was worthy of correction because it was wrong!

Did they add this back or is this something different from what I'm thinking of?

Something else. This was where you'd actually cancel the peace treaty, but stay in diplomacy and make a deal for signing peace afterwards. It had the effect of the AI thinking you were at war with it, but circumvented the few turns the AI wouldn't talk to you. It just ended up being free cities for the player who had a powergraph lead. (Which was very easy to have considering Workers counted as military!)

Now when you cancel the peace treaty, you are booted from the diplomacy screen and you have to wait to contact the AI normally again.

0 turn war would be less confusing I suppose.

[Aeson]

Umm...was getting a city that big ever a realistic idea? How the heck would you get the food?

You only lose one pop point per turn from any city, no matter how large the food deficit. So if you add 2 Workers a turn, then you'll grow each turn.

This was a way to circumvent the maximum population limit dictated by claimed tiles. "Worker Dogpiling." Basically a scoring exploit for milking games.

[teaster]

I've been playing Civ III since it's initial release and until you see all the changes listed out you don't realise how drastic the changes have been. Many strategies that were sure fire winners in the beginning have been minimised or eliminated.

And this doesn't even touch the changes in the units and their ablilities that have been toned down (fast unit retreats), improved (some bombarding units being able to kill units and armies extra move) or added (enslavement).

The latest addition I've enjoyed the most as a warmonger wanna be is the addition of the Temple of Zeus. Though you have to have a bit of luck acquiring some ivory. It can drastically change the strategy of a builder type like myself who would love to be a warmonger but can't seem to stop building long enough to pump out a hoard large enough to roll over my neighbours. Now I can insert a few swordsmen and spearmen into my build queues to support the ancient cavalry and have some fun.

[bongo]

Furrykef (kinda hard trick to pull without taking off your shoes and socks..)

Thanks for starting this thread

I have played civ3+ptw almost to death now, and feel I know everything that is to know about them. Sadly I will not get the time to play c3c to the same level, so I am really looking farward to reading about the changes here.

Much of it is covered in the manual/readme but not in the level of detail I desire for such a game

Things I want to read about:
Traits: Both how the new one works, and in what ways the old ones have been changed (industrious workers are now only 50% more effective, and captured workers only work at half speed of normal workers, regardless of being industrious)
Zero range bombard: Some units(archers..) get one shot at attacking units. It's supposd to work differently in MP than SP(?)

[furrykef]

Originally posted by bongo
Furrykef (kinda hard trick to pull without taking off your shoes and socks..)

And if you don't have opposable thumbs on your feet

Things I want to read about:
Traits: Both how the new one works, and in what ways the old ones have been changed (industrious workers are now only 50% more effective, and captured workers only work at half speed of normal workers, regardless of being industrious)
Zero range bombard: Some units(archers..) get one shot at attacking units. It's supposd to work differently in MP than SP(?)

I'll try to add this and the other stuff I've been meaning to add today or tomorrow.

- Kef

[bongo]

Oh suddenly remembered, unit upgrade cost has been increased, by 50% I think. So that 'mass upgrade of warriors' strategy has become more expensiv. A side effect is that Leonardos Workshop is a stronger wonder now.

[MrWhereItsAt]

Something minor, but can make a difference over the course of a game - Forests now take 4 turns to clear FLAT. No bonus for Industrious workers here (at least, IIRC).

[UnityScoutChopper]

Hiya,

My bit of grist for the mill:

CONFIRMATIONS FOR "QUESTION MARKS"
-- "Commercial trait didn't help fight corruption much. (?)" -- True.
-- "Ring City Placement (RCP) exploit existed but wasn't yet discovered. Cities equidistant from capital suffered less corruption. (This did exist from the beginning, right?)" -- True.
-- (Can a planted forest that wasn't there before still be cleared for 10 shields?) -- Yes. This leads to the micromanagement-happy strategy of systematically planting and chopping forests on every (forestable, originally unforested) tile. However, it's so boring that even the exploit-happy use it but rarely AFAIK.


CHANGES TO ITEMS:

-- "Non-military units viewed more negatively by AI" change -- the top post in the thread (which I assume should evolve as more information is contributed) states that this is so, but not why it's so. The reason for the change is that, before then, on a sufficiently small map, you could be a real scumbag and deprive the AI of resources by placing non-military units on them. Even without a ROP, you would but rarely be asked to remove them. Thus the AI workers could never reach them to place roads on them. In short: "This was done to remove the exploit of denying resources by abusing the AI's high tolerance towards non-military units on its territory."

-- Despot Rush: I'm amazed nobody's mentioned the other change to the model: the release version just plain gave too many shields to pass up! The first citizen killed was worth 39 shields, rather than 20. Imagine: found a city, build some stuff while reaching size two, then invest a single shield in a barracks, and BAM two turns later, you've got all 40 shields of it!

-- MGLs can't build Great Wonders -- extremely important change, as it beefed up the builder style relative to the warmonger style. Yes, armies were improved, but war now brings just the fruits of war itself -- not the fruits of peace as well.

-- Size 1 cities without culture will be auto-razed upon capture... -- no matter why Firaxis did it, this change led to an unnatural strategy (still valid to this day) called "capital chasing". Sir Ralph wrote some excellent work on the matter. Note that "having culture" was not enough; a border expansion was required. This changed either in PTW or in one of its patches. (I never bought PTW, so I don't know.) Now even a single point of culture suffices.

-- "Curragh added..." -- the curragh takes an existing semi-exploit, the Suicide Galley, and shifts it much further forward, especially for Seafaring civs.

NEW ITEMS:
-- one or several items for the change or changes to the cost of buying workers. The first change was in PTW, but I think there was more than one. The Vanilla price was appx. 25-30 gold per worker (depending I think on the selling civ's Attitude); the current price is, I believe, 125-130. This change removed the exploit of carefully watching all known civs every turn ("SHIFT-D, ENTER, check, SHIFT-D, DOWN, ENTER, check, SHIFT-D...") for worker offers and taking all offers -- the old price was highway robbery in the human's favor (30 gold max in exchange for you GAINING the result of 10 shields and 10 food, and them LOSING it.)

-- retreat chances: unfortunately, I don't know offhand when the changes were (and don't have the readmes at hand), but the REASON for them was that with the old chances, fast units retreated so often that their lower attack didn't matter -- you could for example trade speed for invulnerability, stack them with defenders, and attack, auto-retreat, leave enemy territory, and continue merrily along far too much of the time. There was also a significant change to the impact of unit quality on retreat chances. This meant that you can no longer just build e.g. regular horsemen and use the retreat-sometimes, win-sometimes system to bring them to elite, because regulars now have dismal retreat chances. Originally the impact of unit quality was low or none -- I don't remember.

Take this and add the fast unit's, well, fast-ness, plus cheap upgrades for horses all the way to Cavalry, and plus the former non-existence of Medieval Infantry, and you can understand why Arrian has a horse for an avatar. Which brings me to...

-- the addition of Medieval Infantry. Although technically not a bug fix, it is another tremendous balancer between the Horse path and the Sword path. Just ask Arrian.

-- the addition of Guerillas in PTW. Likewise a balancer for the Archer path.

-- starting in C3C (I think), the AI can accidentally attack invisible units, starting a war. This is annoying on the one hand, and exploitable (for a war free of a rep hit) on the other hand. It is not the only flaw exploitable for a "free war": you can also demand things from an AI repeatedly during one diplo session until they are ultra-ultra-furious, then demand they leave your territory, which will usually work. But it may be a more reliable exploit. In any case, we can assume it will soon be patched out.

-- Vanilla: Min. and Max research times changed in one of the patches. Can't remember when or even for certain how much, though I seem to recall a change from 32 turns to 40.

-- PTW release: new worker actions added: Radar Tower (defense booster) and Outpost (vision booster). Radar tower arguably has real strategic impact; I guess we can agree that Outpost doesn't.

-- Conquests release: Shakespeare's Theater boosted to also act as a hospital, making it only a booby prize instead of The Mother of All Booby Prizes.

-- Conquests release: Temple of Artemis added, bringing a fantastic wonder for warmongers and fast expanders or a hey-this-sucks-I-want-my-bonus-temples-back wonder, depending on who you ask.

-- Conquests release: Map Trading and Communication pushed back to late middle ages (Printing Press for maps and Navigation for contacts or vice versa, b-b-bl-bl stupid amnesia...), removing the "MapMaking Slingshot" effect where a sneaky human who had anything to offer at all could leap forward once somebody got MapMaking, since humans have more valuable maps to start with and are better than the AI's at acquiring and whoring better and better maps. Writing (contact trading) served as a similar, but smaller, slingshot.

-- Conquests release: Max research time upped to 50 turns, weakening the max-turn research semi-exploit.

-- Conquests release: massive increase in resource scarcity, occasionally annoying for humans but arguably deadly for AIs facing a tricky human.

----

By the way, is it REALLY true that Industrious civs don't have faster slaves? I SWEAR they still do, but this is the umpteenth time I've seen somebody write that. I've been using them less than in the past for obvious reasons, and likewise have so maybe I've just overlooked the


USC

[bongo]

Originally posted by UnityScoutChopper
-- retreat chances: unfortunately, I don't know offhand when the changes were (and don't have the readmes at hand), but the REASON for them was that with the old chances, fast units retreated so often that their lower attack didn't matter -- you could for example trade speed for invulnerability, stack them with defenders, and attack, auto-retreat, leave enemy territory, and continue merrily along far too much of the time. There was also a significant change to the impact of unit quality on retreat chances. This meant that you can no longer just build e.g. regular horsemen and use the retreat-sometimes, win-sometimes system to bring them to elite, because regulars now have dismal retreat chances. Originally the impact of unit quality was low or none -- I don't remember.
USC

Here's the retreat possibilities:
conscript:34%
regular 50%
veteran 58%
Elite 67%
Units never retreat when fighting another fast unit or when the defender is down to 1hp

As far as I know this percentages are for vanilla civ3 and ptw, last patch. Found it on my HD, originally copied from a post made by jaybe