The Medieval Pack II Beta: The Gathering (Part 2)

hey guys, just to say that i played with the medmod(and ctp2 actually) again after some time.

first of all, i think there is an error in the historical gl field for the temple of karnak.
second, the new tech tree and units look great! i love it that you can have a noble early on
third, playing on hard level with 8 civs on a normal map, i dont get much of a resistance. i havent had many attacks from an ai, but there are wars between the ai's with mamny cities changing hands.
fourth, on the diplomatic side, it's fun to finally see the ai sending you messages asking you to stop researching something. i would still like to see the ai doing more counter proposals, even impossible ones

keep up pthe good work!

[janilxx]

Wow, one day off and the weekend and what we have here. Tons of new posts!

I think that we have now had the best discussions here in last 4 days. Many, many posts and all are full of good arguments!

Couple of comments about your discussions.

Terraforming:
If AI builds always farms on grassland and mines on plains and that can not be corrected(it should!!!) then I would propose that AI would terraform some of its grasslands to plains or some plains to grasslands. I suppose this would destroy mines/farms on these squares. Then he maybe would build new mines on new plains and new farms on new grasslands. (I suppose that squares have to be empty that AI would build improvement on that square).

No settelers to mountain:
There should not be goody huts in mountain too(I am not sure is it possible now). If I would found settler or city from goody hut we would have problems. Settler could not move or there is a city in the mountain even it should not be possible

My own comments:

I saw that there is now a file MM_const.txt. I have used modswapper and I saved medpack to be the used mod. Then I checked const.txt. AI_GOAL_TIME_SLICE, AI_MAX_TIME_SLICE and AI_TOTAL_TIME_SLICE values are not changed(increased) in const.txt. But in MM_const.txt they are increased. So how does modswapper work? If I start medpack using modswapper does it copy mm_const.txt over const.txt when I am strating game?

What I mean is that would those new(increased) AI_GOAL_TIME_SLICE, AI_MAX_TIME_SLICE and AI_TOTAL_TIME_SLICE values from MM_const.txt be used in medpack definitely even the const.txt says that those values are not changed.

I had an embassy with green AI and I checked intelligence screen. There were in third line a text: "Embassy requred to access priviledged info". But I had an embassy and I could see all his infos even that text was there. A little thing that would need to be corrected.


Funny thing. As I said before AI-AI relations were always or There were actually 2 wars ongoing. AI1 against AI2 and AI3 against the AI2 again. So AI2 had two wars ongoing. And AI2 was diplomacy-minded peacemaker


I had an ally and I built forbidden city. That ally had an noble next to my city. That noble tries every turn establish embassy with me and he fails every time Should there be some kind of check that if I(destination nation) do have forbidden city that is not obsoleted he would not try to establish embassy.


Siege engine's icon is too big! It is ugly on map. It's size could be reduced about 20%


It would be good that last changes in docs would be written in different color. For example the last changes would be in red etc. Now you have to read all docs through every time you get new version.


Is it ok that when I am using spy and he is captured but that spy still stays next to AI city?!? Quite stupid?


It is quite stupid that slaves always goes to the nearest city. Or am I right? That city grows fast when I am in big war against AI. And when slaves will be someday free my city will be in big problems! In my opinion slaves could diverge for example to three nearest cities. So one city would not grow too much with slaves.


I have war agaisnt AI and I took three of his cities. He have not done anything! I have not seen any of his units(of course there were defenders in his cities) trying to come and capture his cities back. And one of his city I captured was his capitol! This is the same subject that was discussed a lot in last days.

Jani

[quote Arquebusier's [/quote]

Wes I did the research for these guys, however they do not show up in the building queue.

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

[This message has been edited by joseph1944 (edited May 21, 2001).]

[Dale]

Locutus:
How you described the diplomacy tips is how I've always played diplomacy. Basically, my diplomacy mod was designed for just that method. As I said ages ago in a thread about diplomacy, put your neck out diplomatically, suck up, get the AI happy and go for diplomacy win. It's just a lot harder with the diplomacy mod since now the AI can "think" about it's diplomacy correctly.

Also, I'll check the embassy code tonight. I didn't realise there was something going on there. I haven't changed it since it was working last, so it might be a conflict with all three scripts somewhere. Will check it out though.

------------------
Rommell to a sub-commander outside Tobruk: "Those Australians are in there somewhere. But where? Let's advance and wait till they shoot, then shoot back."

[Locutus]

I like the terraforming ideas of Peter, I was trying to accomplish the same thing by largely removing swamps and forest from the game and replacing them by grassland and plains, as I described earlier. Great minds think alike, I suppose My method doesn't need SLIC and feels less like cheating, but is also not quite as effective.

Another terraforming related issue: would it be possible to make 2 kinds of forest? One that could only be terraformed to grassland and one that could only be terraformed to plains, that would make more sense than the current setup. If you cut down the trees, the soil underneath it is determined in advance, you can't just choose it. I don't know how well this would work when generating random maps though...

Harlan brings up another point though: the city limit. The more cities a player has, the more powerful he is. But I hardly ever see the AI go over the limit while the human doesn't mind, esp. not since the punishment for this in the MedMod is (deliberately) fairly low. This setup may be more realistic but it hurts the AI. I for one have at times had empires with up to about 25 cities where 10 was the limit (you'll need to invest a lot in happiness but it's quite possible), but you'll never see the AI do this, 13 is the most I ever saw (and that was due to exceptional circumstances).

Harlan, I think you need to build Castles first before you can build a Bastion, could that explain it?
Glad you liked the diplomacy tips though, now I know that my diplomatic successes aren't just all in my head

David,
Wow, you created a DL just to tell us that, you really needed to get that off your chest! (You could have just emailed Wes ) I don't know why you got banned in the first place but I hope the mods were lenient on you for that...

Hex has a point. That suggests that there's simply a lack of PW and the AI prefers to build Farms over Roads. The Unit Repair code now eats part of the PW, so if that isn't compensated properly in the AI strategy files, the amount of PW for Terrain Improvements will shrink so much that there isn't enough left for Roads. Maybe upping the PW levels a little (further) will help?
[This message has been edited by Locutus (edited May 20, 2001).]

[Immortal Wombat]

quote: Originally posted by Peter Triggs on 05-19-2001 10:40 AM Who do we know that's good at graphics?

Harlan , but it doesn't matter, because we can't yet add tile improvemets or terrain graphics to the game.
It would have to be a trade-good, and then the AI would trade it, and it would get stupid.
Has anyone written to Activision about this? They have no reason to keep the *.til editor anymore, can't they release it?

[hexagonian]

Regarding roads

Try dropping Farms to later in the game - the AIs in my mod are building roads (although they tend to go overboard sometimes)

[preacherman]

Wes, all I know is that when I got radio no subs, went to cheat mode remove Electro-magnetism from advance list I had subs. Quit game without save and changed MM2unit file to obsolete at nuclear power, reloaded game and I had subs.

------------------
Hear the word and live refuse and fry

Wes:

Trust me, the trade values do always drop upon reload. You just weren't looking hard enough.

If you need any more convincing, please visit http://members.aol.com/davidwamurray/tradebug.htm

They don't drop by quite half, but they certainly drop, and it is the most stupid of bugs.

[Harlan]

Oh, one last thing. Locutus, your diplomatic tips really help. Wes, you should include those in one of your readmes, cos otherwise many people won't see. And hopefully we'll soon have the AI kicking so much butt that people are gonna need all their diplomatic wiles to survive!

[Harlan]

Terraforming! Arrgghh! (its hard to type while shaking a fist in the air ). I'm glad some people agree with me about terraforming, even if it is only to get me to shut up about it. You're right, Locutus, this has been a pet peeve of mine for quite a long time.

I really like Peter's idea of having the forests automatically cut down, to help the AI. One could extend this to the jungle and swamps too. It wouldn't be so bad if this was a freebie, AI only thing, since the AI needs every advantage they can get. Give the human player the opportunity to do it for not a terrible amount of PW (though draining swamps is expensive). Extra nice bonus would be to have the AI forest replaced randomly by either grassland or plains, since the AI only knows how to put irrigation on grassland and mines on plains.

As for other types of terraforming, they should be banished to the future - grrr! (more shaking of fist :0).

One other pet peeve this reminds me of is the AI's tendency to build cities in very poor geographical spots, just cos its empty. Without terraforming, the human isn't likely to build, say, in the middle of the desert, and with Settlers unable to walk on Mountains, no one is building cities in mountain ranges. The final piece to this is getting the AI to have more of a negative value (maybe even literally a negative value) for glacier, tundra, and desert. With the AI pushed into more intelligent behavior, its cities will do better, and big deserts, glaciers and so forth will remain as they should be- empty. Given the 10 city limit early in the game, its important that the AI is as smart as placing its cities in great starting spots as the human. I haven't seen that in my two recent games.

I'm glad people agree about the few AI units. It might be interesting to speed up AI production, while slowing down their science. This way, while their cities will become more developed, and they'll have more units, they're less likely to pull ahead scientifically, making it impossible to catch up. Some playbalancing is clearly needed here. Anyone want to try changing the two lines I mentioned in the previous post, and see what happens? As for me, I'm gonna wait till Wes has things in working order before continuing with playing, and hopefully FINALLY finish the Alex mod. (Its so close to completion, but there were some SLIC problems that made me lose enthusiasm. With Locutus back in the picture, hopefully those can be overcome)

I would love to be able to change terrain, but alas. Maybe someone should try contacting some ex-Activision folks about it now, cos why should they care anymore? I hear the whole company might go out of business.

Locutus, by the way, I'm 100% sure in my last game that I developed the Military Engineering tech, and didn't get Bastions, though I did get Battlements.

Maybe the AI's inabliity to build roads has to do with it never getting to that priority item, since it has its hands full with other things? So changing their production might fix that as well.

Also, Wes, you neatly sidestepped my comments about Catapult vs. Archer, by not addressing it at all. I won't let you slip past so easily! Why not have an archer track of development, and an artillery one?

[Peter Triggs]

Hi, all

As Locutus said:

the AI analyses its goals and executes them with the highest priority first, moving down the list and skipping items which don't force-match (i.e. for which it doesn't have enough units).

And check out the Barbs force matching settings: they're in Kamikaze mode, they attack with whatever they've got.

As regards terraforming/tile improvements, I agree that something should be done about this. The game algorithm that handles this stuff is only half exposed. It was the same with CTP1, you could never get it to do what you wanted it to do.

For example, the AI will not touch forests. (David Murray has aptly called it a 'tree hugging AI'.) In the version I play, I've added a couple of handlers to deal with this problem. When a city is founded, all forests or jungles in it's radius that do not have trade goods on them are immediately converted to plains. When a city reaches a population of size 6, you get a message saying "Sire, the city of {city[0].name} is bursting at the seams. We have sent squads of workers to scour the surrounding countryside and collect building materials for it's expansion." and most of forests/jungles in the squares it's about to expand out into are converted to plains. This is a big help for the AI; you no longer find underdeveloped cities in the middle of a forest/jungle.

It would be easy to adapt this code so that whenever one of it's cities expands, the AI does any terraforming/tile improving that you want (although there might be a bit of a problem making sure the AI has enough PW reserve to pay for it and I agree with both Harlan and Locutus that some of the existing possible terraforming is just unrealistic). It's been in the back of my mind for some time to bring the Cow back, not as a unit but as tile improvement (Dairy farms or Ranches?) for hills. On the other hand, sheep might be better or there's Harlan's suggestion about irrigation. For forests, I seem to remember that someone suggested lumber camps or paper mills. It's really just a question of deciding what we want and getting someone to come up with the necessary graphics. Who do we know that's good at graphics?

[GreatDane]

I agree with Harlan on the few AI units problem. In my games I have not seen many AI units either especially later in the game when the units are much more expensive in upkeep. Giving the AI the resourcers to build and maintain more units would certainly result in a more aggressive AI.

Henrik

[Locutus]

Wes,

I basicly agree with Harlan on the terraforming, most of it is quite unrealistic and it only unbalances gameplay. So some of the terraforming should indeed be more difficult. But OTOH, it is true that making it too difficult pretty much removes the whole terraforming feature from the game as it's unavailable or too expensive for too long, which is a pity because terraforming in a sense is and will be part of human history. So some sort of compromise may be needed.

One kind of terraforming that's still missing from the game however and that *does* have a history is land->water and water->land terraforming, the Dutch pretty much shaped every square centimetre of their country through terraforming: they (we) turned land into water, water into land and swamp and forest in to plain/grassland (and on a smaller scale plain/grassland into back forest and into swamp). Maybe you could replace the terraforming of terrain into mountain/hill with terraforming things into shallow? For making hills/mountains more productive food-wise, farms would indeed seem like a more suitable solution to me. OTOH, though historically corrrect, terraforming (into) water would unbalance gameplay just as much or even more so than terraforming (into) mountain/hill.

In any case, some more drastic changes in terraforming would be appropriate, even if it was just to finally get Harlan to shut up about it (I still have some vague memories of this same discussion going on when I first came to Apolyton a loooong time ago (long before I registered and started posting))

I hope you find my tips useful. If Dale doesn't do it before me, I'll look into the re-establish embassy code, it's quite possible that it's affected by the SLIC savegame bug (let's call it that and let's call the other, older, savegame bug the scenario savegame bug to prevent confusion).

Huh what, you don't have the Capital code? I finished that a week ago... Doh! Must have forgotten to email it to you or something Presuming I didn't overwrite the code when I reinstalled the game a few days ago ( ), I'll email it to you later today. I'll look into the enslaver disband code too, hopefully I'll finish it this weekend but can't promise it yet as I have some schoolwork to do as well.

The changes you made all sound very good to me.

Harlan,

Actually, the Bastion *does* come available with Military Engineering, I've researched and built it many times already. Must be a mistake on your behalf. You're right about the Bombard and Arquebusier's movement points though, just checked the files. Must be because Wes originally envisioned both of them as militias.

Well, if you had read Richard Myers' explanations on how the military AI works and esp. the forum discussion that sprung from it ("Conversations with Richard Myers (Azmel)"), you would have understood that (more units -> better AI) already. He explained that the AI analyses its goals and executes them with the highest priority first, moving down the list and skipping items which don't force-match (i.e. for which it doesn't have enough units). So, for me at least, this is rather obvious. The problems I'm assuming Wes is having are more along the lines of getting the AI build enough strong and useful units while still keeping up a good economy. I thought Wes already gave the AIs a production bonus but if not, doing that after all would indeed be likely to improve AI performance.
[This message has been edited by Locutus (edited May 19, 2001).]

[Harlan]

Aha! Thinking about the problem some more, I think I may have figured out, or partially figured out, the anemic AI problem. I was thinking about the Alex scenario (revised edition, still unreleased), and comparing the AI play there to their play in the Wes mod. In Alex, the AI are way more aggressive, even though the settings for them are little to no different from settings in the Wes mod. Why? It hit me- they build tons of units, cos they don't have anything else to build or do, really. I've given them extremely limited options on improvements, no wonders to build, forced them to keep PW low, science low. So all that's left is building units. And with enough units, even an idiot can conquer, esp. since they know how to stack.

The reason the barbarians are doing well must be the same. They don't have to worry about anything else but attacking. They don't develop the cities they conquer in any way. In Civ2, the AI is really stupid, but they also churn out a ton of units. If you walk into their territory, you see them wandering all over the place, in their aimless, stupid style. Not so in the two games I've play of the Wes mod. I've sent spies and such into AI territory, and its very rare I run into any unit wandering outside of towns. Probably the AI is having a hard enough of a time keeping their towns defended, plus all the other development, that they don't have any extra units to go adventuring with.

Civ2 AI had lots of units, cos the AI got a cheat on production, being able to produce things with only 75% of normal production. I noticed that in the const.txt of CTP2, there is a line:

RELATIVE_AI_PRODUCTION_CHEAT_COEF 1.0
RELATIVE_AI_SCIENCE_CHEAT_COEF 1.0

This hasn't been changed in the Med mod. Perhaps if it was, and the AI had a big production advantage, they would be much more competitive, and aggressive. But then again since I've been away for so long, maybe someone else has thought of all this and why this line of thinking doesn't work, and I'm way behind the times. Highly doubtful this is the whole solution, but maybe it will be a part of it.