To Kraellin

[Edratman]

I like what you accomplished in that mod. I've got 6 factions plus myself trading now. Of course I started first so I have a big treasury edge on the AI. I going to try adding a little to the ship figures for the AI. I also gave the higher shipyard and trade figures to rebels. Livonia builds ships and they are the Pirates of the Baltic. No trading, just attack anyone. Cool.
I also jacked the build numbers 100 for all the + happiness buildings. It really quiets down revolts for the AI. I'm only up to about 1270, and I only see revolutions in the provinces with a revolt factor higher than 0. (Probably try taking several provinces to 1 next time around).
I also changed the prerequsite for chapter house to castle 9. It got rid of those weak assed crusades and allows a lot more trade by eliminating those wars with the best trading areas.
There is definately a lot more money in the game and the AI seems to play a lot smarter with money.
By the way, I used my base mod and added your build factors. It is similar to yours (still has the land bridges), but I have more trade goods than yours: in fact all inland provinces have either 2 or 3 trade goods. This made Poland and HRE much tougher than usual.

[Kraellin]

edratman,

yeah, the ai trading thing numbers were purely experimental and i wasnt totally happy with the ai still not building enough ships, but what i got did make a better game, so it was better than a poke in the eye.

i love the idea of a pirate faction. making the rebels as that is a great idea. i may have to try that also. i had hoped by making the danes 'barbarian raiders' that they would fill that role somewhat, but it didnt work quite the way i thought it would. still, the danes are better. i'd be interested in hearing how you tweaked the livonians. also, wouldnt it be cool if pirate raiding got you income for doing so.

yeah, jacking the happiness stuff up for fewer rebellions is definitely an option and particularly helpful for the ai. i can go a whole game and almost never have one for myself, so it's sorta unfair that the ai has so many. i like that idea as well.

interesting idea on the chapter houses. i've not fooled with them, but i like the crusades and jihads. also, i believe you can alter the crusade strength, so you might wanna look around in this forum for the data on how to do that.

yup. i think the more money thing definitely helps the ai out. even with no other mods, i think that alone would be enough to increase the difficulty in the game and that was one of the complaints early after the release, that the game, even in expert, was too easy. adding more money overall makes it tougher to conquer someone completely.

btw, i dont know what the limit on how many trade goods and resources for a given province is, but i suspect you could have one of each goods and each resource in every province if you wanted. i had up to 5 trade goods in my earlier versions and several resources in some, so i suspect it's only limited by how many actual resources and goods that there are in the game.

i've also one question for you. you said you have 6 factions doing trading by ship now. have you found any of them starting wars frequently by attacking other faction's ships? i found the ai doing this a lot to me and other factions and thought, boy, that's stupid. why ruin your good trade routes and income by all this warring? it makes me curious if some of your other modifications are affecting this or not, like the happiness buildings; are they lowering the need for war as well as rebellions or are you also modding the base ai behaviors?

K.

[Edratman]

Kraellin

The limit on number of trade goods is definately 5. I've tried 6 and it would not work.

I cannot honestly say if the AI is attacking ships. My experience is that if my ship is attacked by the AI, I'll stay away from that sea zone until I can put at least 2 ships in it, preferably 3. By then the war has evaporated from the diplomacy screen. I don't have enough time to check everything every turn.

I also seldom, if ever, have a revolt so that is why I thought of tweaking the + happiness buildings. It seems to be working.

I didn't specifically tweak Livonia. Your changes to port and shipyard left the rebels values unchanged, I merely put those factors for rebel build numbers.

I like the crusade change. Earlier Giljay said the AI is driven to achieve its GA goals, and crusades are the biggest points on the board. Now it seems the AI is very protective of its home provinces, probably because it is the most points.

Also I came up with a quick way to play test a mod. I set myself up as Sicily, but keep only Malta. I put 6 late era ships at Malta which makes the island off limits to the AI. I don't bother doing anything but hitting end year and watching the game using matsaori(sp?). I can knock off 100 turns in an hour or so (depends on how many cold ones I consume and have to get rid of).

By the way, the campaign I'm playing now has Polish ships trading, never saw them before. Must be led by the famous Waterski clan.

By the way, salt is also a trade good. I tried the other resources as trade goods without success. Are there any new ones you found? A couple of new ones would be good to minimize the non-trade effect.

I'm also thinking of eliminating iron for the metal smith. If anything iron would have been the most valuable trade commodity and I think everyone would have been at a similar position regarding that. Besides, I'm tired of moving units around for upgrades, just for the time factor.

I did some minor changes to base AI behavior. All Catholics are one of the trader choices. The Orthodox and Muslims have less options and I've left them at the original settings.

By the way, my base mod puts a fort3 in every province also. Helps HRE mostly it appears, but it also toughns up the rebel provinces.

[Kraellin]

hi ed, (name shortened for ease of typing :)

good to know, the 5 instead of 6 trade goods. i had only tried as high as 5.

what i was trying to see was, if by raising the overall happiness of your provinces through making the 'happiness' buildings easier to build, if the ai would lighten up on the number of other faction ships that it attacked. my burgundians, for instance, were a small faction to start with but would build ships, but then they'd almost immediately attack another faction with ships, thus starting a war, and it was pretty much a suicide move for them. so, if you observe any changes there, let me know.

yeah, the ai needs a break on rebellions. i'll most likely do something similar in my next tweaks.

roger re the rebels.

yeah, i'd forgotten about the GA points thing carrying over into the conquest game, so, if altering the crusade thing helps the ai protect itself more, this is going to be an interesting game balance thing. where i mostly like having the crusades is when i get overly aggressive against other catholic factions and the pope steps in and excommunicates me and then calls for crusades against me. that's a pretty cool item to relegate to almost never happening now because you cant build chapter houses. but, if it helps the ai protect itself more (one of the things it tends to be poor at) then, i guess it's going to be a trade-off thing of being more offensive or more defensive. nonetheless, it's an interesting catch on your part and does open the door on yes another way to balance the game issues and factions.

yeah, i was going to do something similar, just hitting the end turn button, but i end up always seeing opportunities that i want to take advantage of as a player and jump in anyways :) i also tend to turn the matteosartori on and off just to keep things somewhat in perspective while testing.

waterski clan? lol. i've seen turkish ships, burgundians, aragonese, french, german, danish, english, spanish, egyptian, byzantine, italian, novgorod, and even the papacy, but never the poles. fun, isnt it :) i did limited tweaks on the whole shipping thing before, but the next one i'm going to go hog wild and turn it on for everyone....way up and see what happens.

yeah, there is one reference in the files to salt being both a trade good and a resource. it's the only one i know of that swings both ways. there is also a reference to 'saltfish', but that may just be a typo where someone left the space out between 'salt' and 'fish'. the trade goods i know of are: butter, cotton, dyes, fish, furs, gems, grain, glassware, hides, honey, ivory, linen, oliveoil, pottery, saltfish(?), salt, silk, spices, sugar, wax, wine, wood, and wool. the resouces i know of are: copper, iron, gold, salt, and silver. oh, and there may be a resource of 'forest(s)' also...not sure...untested, but i've seen a reference for it somewhere.

removing iron for metal smith? interesting choice. you'd certainly get more factions having stronger units. what i'd like to see CA do with all this, is strengthen the whole trade thing so that you could trade with factions for the things you dont have, but that you need to keep your economy going or tech tree rising. it would add just that much more to the strat side of the game. and, it could force some interesting diplomatic situations as well.

yeah, i didnt tweak all the behavior stuff either, but apparently i must have tweaked something for the egyptians because they ended up being a very strong shipping faction. also bear in mind that by only giving the catholic factions stronger trading, you are depriving the muslims of the potential involved in trade routes and that's going to hurt them down the road pretty badly.

a fort3 in every province? wow. in a way, that actually helps the human player. when i bribe a rebel army, i always bribe where a fort exists if i can, so that the newly bribed unit, if attacked, can retreat to the castle and hold out till i bring in reinforcements. and, even if it's not attacked, that also means i can start building units right away in that province. and, i can sometimes bribe the unit residing in the fort, but not the one standing in the province and again bring in reinforcements while the guy in the fort holds out. i do a lot of bribing in my games :)

a lot of this stuff is just personal choice stuff. your way is your way; my way is my way. neither is right or wrong more than the other. it's all mostly fun. i do like hearing what other folks are doing with this stuff. someone else always comes up with something i never thought of before and it often inspires changes in my own mods, like your pirating rebels. i love that one and will definitely be trying it myself.

my biggest problem right now is that i've got all these ideas and not enough time or knowledge or patience to implement them all at once and now they're talking about an expansion pack and even the next major TW game. gah

K.

[Edratman]

I can agree your your "GAH". I've got the same problems, too many good ideas, not enough time to implement or even try 1/2 of them. Every time I make a change I have a nagging thought that I forgot some other great concept that I really wanted to try. (When you get to my age you become accustomed to that.)

You are right, it is all a matter of preference. A person gets a thought track and likes to follow it. I've got over a dozen mod choices on my "Early" startup menu, but I never go back to old ones, I just keep trying new ones. I hope everyone keeps trying things and posting them so I can "borrow" what I like.

I like your concept of trading for goods. If iron was a resource and a trade good which could then be counted as a resource, then trading would have a real function other than being a cash cow. (CA- HINT, HINT. Probably too complex a change for a expansion, but worth considering for a future game.)

The fort3 idea turns out to help me with adjacent rebels only, and then only in the earliest stages. I find bribing rebels without shipping contact to the king is a sure revolt and sometimes I will bribe a rebel province to put a stonger force in it to oppose the AI. The AI now likes to invade rebel provinces quicker because it doesn't have crusades as a point gaining option. And I'm probably going to make all initial rebel held provinces revolt factor 2 on my next change. The thinking is that I view them as independent kingdoms and they should have an impetus to revolt, so bribing a small garrison will almnost surely bring on a revolt unless you can reinforce immediately because neither I nor the AI can build + happiness buildings quick enough for revolt factor 2 provinces.

As for non-catholic factions, they only have 3 behavior choices, none of which includes a trader option. I did not do anything to limit them, CA did that.

I didn't turn off crusades, I just made it take longer for the AI to initiate them. I'll let you know what happens when it gets to that point, if I get there and don't start new mods. (Don't hold your breath. Hopefully someone else will experiment with various options and report back.) I actually tried it because I don't crusade because I learned it always cuts off major trade options. That is the big secret in the game, money is everything. Once you learn to trade and amass a treasury, some sort of victory is assured. I don't play the conquer the world game, mainly because the turns take too long to accomplish and I want to try a new mod anyway by the time I get there.

By the way, the + happiness building boost takes assassins out of the game, everyone has the second watchtower. I just put an assassin in every province for defensive purposes. Don't know how to change that. It would be nice if stars would help an assassin avoid watchtower death. I think that stars have no effect on avoiding that.

Another option for the Polish fleet is the Fishing Poles. (my surname ends in "ski").

[Edratman]

Another change that I found beneficial was reducing the support costs of early royal knights by 33%. It helps the Danes. I think the abundance of heirs and their early royal knight troops shifts the Danes to "poverty sticken" or "close to support limit" or one of the other alternates listed in the build files. So it looks like the lower support cost allows them to stay within their expansionist_trader designation and they then expand into Sweden and Norway reliably. My thinking on the AI is that just compares numbers of troops, not quality, so it thinks 6 units of peasants are better than 4 units of royal knight, for example. Thus the Danes with all those heirs and royal knights cannot build numbers of troops because the income/support ratio kicks them into some other action/decision tree. (Reducing the number of heirs should work but I don't know how to do that.) That is why I think the Danes are so easy for the player but performs poorly for the AI.
Nothing seems to help Aragon. They are surrounded by 4 large powers and seem perpetually doomed to self inflicted suicide by attacking one of them. They don't have the heir problems of the Danes, but Spain as the AI almost always jumps into Cordoba initially, limiting Aragon to one rebel province which it must compete with Spain, England and France for possession. It has a decent income, so it seems to be able to amass a fair sized stack which it throws at an adjacent province where it has numerical advantage at the first opportunity it gets. But unless that province is owned by someone already in a balanced struggle with someone else, this move almost always fails from what I have seen. I guess CA put them in for some historical accuracy and like history they are doomed to be absorbed. (I made a mod to play them and after 10/20 turns decided I didn't like them and quit. I'm sure that is a common result of modding.)

[Kraellin]

the aragonese are a difficult faction to tweak so that they're worth anything. however, that's not all bad. i dont want all factions being the same so that the player can play a tough game or an easier game. to get them a bit tougher when the ai is handling them, i give them an additional province to start, though not navarre. and part of the trick to getting them going is exactly what you touched on; you need spain and the almo's to be in conflict and france and the hred to be in conflict, so that nobody looks to the aragonese. the other thing i tried to do with them was to make them big ship builders so that they could invade to other places through the sea, and i've had some success in that regard, but not quite enough. the same is true of the burgundians, though they tend to last a bit better in my mod and maybe only because they get a little more income from the provinces i give them initially.

yeah, for the danes, i simply gave them more starting money and a bit of extra units or buildings, i think. this gives them enough extra money that they can start raiding a bit. oh, and i made them barbarian raiders and tweaked the barbarian raider ai stuff up more for ship building. so, i've had them take norway and sweden, or invade england, or invade livonia or some place like that. this works fairly well when the ai is handling them, though i still need something a touch more for them. they are still somewhat weak. what i really wanted to discourage with the danes was that silly attacking of the hre where the danes always end up losing, and i've accomplished this somewhat by adding forces to the hre in saxony or whatever that province is that adjoins denmark. the danes, as the ai, are still somewhat of a wildcard. they seem to be just too aggressive to be able to maintain what they capture, but then, in my mod, that's ok. i like the variety of game play this brings. a nice wild card faction that can influence other factors, but isnt necessarily a world conquerer.

i'm currently staying away from all unit tweaks, including support costs, so your tweaks there are of interest.

yeah, i played with the revolt factor on a couple provinces also. this is also a nice wild card thing. you can also use it with non-rebel factions to somewhat keep the ai from expanding too quickly, particularly if you've tweaked the happiness buildings up for everyone, this is a good way to province by province tweak it down again. i had to do this with the swiss when i added them. the swiss are really a later era faction and they were just killing everything by being in the early period, so, i make switzerland and the other province i gave them, high rebellion provinces. thus, if the swiss move their units out too early, they get backstabbed and it slows their expansion down considerably. i also gave them very few units to start and if someone does attack them and take their provinces, then the attacker is also most likely going to get a revolt and it will be a loyalist revolt, so the swiss are right back in the game again. it makes a nice offensive and defensive counter.

yes, the muslim factions only get the 3 behaviors. i modded the expansionist one to include the extra ship building stuff but didnt give that behavior to all, just some. thus, if given the chance, the egyptians will be pretty good sea traders, where the turks might build one ship after 100 years :)

hehe, one of these days you're going to have to actually play your own campaign and see what you've got ;) that's where i currently am and it's become a really tough game in the later years. i finally managed to wipe out the egyptians through assassination, but then the almo's re-emerged and took a TON of the old egyptian holdings before i could scoop things up too much, so i'm almost right back where i was with the almo's as i was with the egyptians, except for the shipping. i now AM the king of the seas and that's going to help.

the trade game vs the conqeust game is one where i'd like to see CA develop things more as well. it would be nice if there was a game mode for, like the GA mode, for building a trade empire, instead of trying to conquer everything. you'd still have battles and conflict and so forth as raiders and opponents tried to bust up your empire to prevent you from winning, but the emphasis would be on building an empire through a different sort of conquest, the monetary conquest. in this mode, i'd include things like allowing for vassal states, huge bribes to bribe entire provinces to your side immediately, much better shipping for everyone, land trade routes that not only work province to province, but also include routes off the board to simulate the old spice and silk routes and so forth. add in the actual value of things like iron and salt (didnt the romans actually use salt as currency at one point?) and that one buys and sells and tie all that into the tech tree and even influences and diplomacy, and you've got a completely new game mode without nearly the amount of coding it would take for a lot of other interesting additions to the game.

the whole, do i trade or do i fight or do i try to talk to my opponents to influence them is a badly neglected aspect in this game, and a lot of other ones as well. it actually amazes me that CA would move on to a completely new era and emphasis as opposed to expanding mtw into its true potential as a strat game. the possibilities in mtw are pretty incredible. i could easily see 3 expansion packs for this game just to expand into all the potential this game has and THEN move on to another era and faction emphasis.

hmmm, takes assassins out of the game? not sure i like that one, but, assassins can still be used effectively, at least by the player. you simply move them in groups and accept that you're going to have attrition each turn. also, my observation is that the level of the assassin does make a difference in avoiding watch towers and such, but it's hard to say for sure. it's more observable when you get some 3 and 4 star assassins going. they do seem to slip through the defenses better.

your overall game seems to be going more towards the peaceful trade game. higher happiness, fewer revolts, fewer crusades and so forth. i've somewhat done this as well, but not to the extent that you have. keep me informed on how it goes, and do try and play a game for more than 10 years ;)

K.

[barocca]

Quote[/b] (Kraellin @ Dec. 28 2002,14:06)]... tweaked the barbarian raider ai stuff up more for ship building...
...the muslim factions ... i modded the expansionist one to include the extra ship building stuff but didnt give that behavior to all...
Kraellin,

(no i haven't "sorted" the stats into campaigns and mods yet - when i have a little more time?)

BUT

please tell what you mean by "tweaking" the behaviour ai,

PERHAPS write a quick guide for it?

[Kraellin]

bar,

that's a bit of a mis-statement on my part. what i'm talking about there is tweaking the data the ai uses to do what it does. in the building and unit stats are areas that refer to the various ai 'behavior' types. the ai behavior types are simply those things like 'catholic_expansionist_trader' and so on. there's several of these and the ai relies on the data associated with these behaviors as set in the building and unit stat files. tweak that data and you change the ai behavior. most of that is simply altering the priorities of that which the ai builds, units and buildings and ships. that's how i got the ai to build more ships and do a few other things.

i think DOC was the first one to really exploit all this. you can see some of the discussion of this between him and me in the 'ok, i've started modding' thread.

K.

[Edratman]

I finally had all my holiday company leave and got to spend some time playing. This campaign is up to about 1290 and I'm playing England, I'm dominating everything and it is time to remod.

You asked several posts ago if the AI ships attack when they are all trading. The answer is that facet is unchanged. Whenever they have a numerical advantage they attack my ships, but I'm not seeing any evidence that the AI attacks other factions. (Could be I missed that.) It must be hard coded, the AI treats sea and land forces identical, and whenever it sees a numerical advantage it attacks.

There was no way I could maintain numerical advantage in ships everywhere initially, so I ended up in sea wars with one, then another, then another. But I had a vast trade route and a better treasury so I was eventually able to sweep the seas, leaving me right where I tried not to be, the sole major trader. Five factions have a couple of ships in a few sea zones, but now I have them outnumbered so they are not attacking, but the AI cannot catch me now, moneywise or otherwise.

The shipping attack issue will have to be resolved by CA. Until it is fixed, I really do not think the trading issues can be resolved.

You are right, I play a fairly passive trading game. As I said before, money is the secret to success. No matter how good a battlefield general you are, if you are getting outearned 5 to 1, you will lose. (WW II was a real world model for that). Even still, by 1200 I had taken France, HRE, Aragon and Italy, in retaliation for AI attacks on me and the stupid AI refusal to accept cease fire. Could have swept up more, but I wanted to play out the game. But now the outcome is inevitable, so it is time to start anew.

You are right about the Romans and salt. Rome paid the legions in salt, a commodity that they could then trade for anything. Hence the word "salary".

My game playing style is derived from my being an old dog. I'm 51 and been playing computer games since the very first day, i.e. text Zork games. (Probably been playing longer than many Org members have been alive, at least judging by their writing and spelling.) There were no WEB or games magazines available then, so a person had to figure out how to beat them alone. So I learned how to figure out how to beat every game, and this I guess involves exploiting the weaknesses in the programs. I don't make apology for this, I'm an engineer and I do the same to everything. They are only games and mastering one does not make me healthier, wealthier, younger or put hair back on my head.

Barocca,
Kraellin is correct about modifying AI behaivior. They only option is adjusting values in both of the crusader build files. I learned about it by following the thread he cites and have just applied my own thoughts to their ground breaking. And most of my thoughts are adapted (stolen) from other posts, i.e. lowering early royal knight support costs came from Wes W.

Bottom line so far, AI tries more trading by adjusting crusader build factors for trade buildings but screws it up attacking ships. (I set all factions, rebels included to the Catholic-Trader settings Kraellin used in "Russia Q" mod.)It does this regardless of behavior setting, so I am at a loss here. AI revolts are reduced considerably, but not entirely but adding 100/150 points to the build factors on the + happiness buildings. Delaying weak-ass crusades help reduce some senseless early wars and helps promote some trade, but the AI screws it up eventually. Putting more trade goods and resources around helps the AI because it plays better with more money, but until it learns to amass a treasury, it is exploitable.

I'm going to make all initial independent provinces revolt factor 2 next mod. The revolt factor 4 of Portugal seems a bit extreme to me, but I'll experiment.

Need to get more new ideas. I don't have any tweak ideas that I consider viable.

[PanthaPower]

Guys, please don't stop writing more about this. This is some good stuff

And Edratman, although I'm "only" 27, I've been around since the Vetrex console and had my share of consoles and pc's. What I noticed over the years of gaming is that 3D didn't bring better games. I have the best memories from my old commodore 64 and Amiga where gameplay still was everything. Ah, the good old times... http://www.totalwar.org/forum/non-cgi/emoticons/smile.gif

[Edratman]

Pantha
Thanks. I'll wager that a lot of Org members think you are unimaginably old at 27. LOL
I'm sure games are better now, but not in ratio with graphics changes. Just taking Civ I for example, I would say MTW is 2 or 3 times better as a pure game, but the graphics are 10,000 times better. I think MTW is the finest game I've played to date, within the genre.

Everyone, IMHO Pantha is right. This is a pretty good thread so please don't respond to this post and turn it into a game/graphics debate or an age discussion. Let's stay on topic. There is plenty of room for new posts in the forum.

[Kraellin]

ah, ed, yer an old fart too. i'm 52

well, i'm a bit further into the years of the campaign that yours, but i've just about hit the 'ok, i'm pretty certain to win' point also, so almost time for my next revisions.

i finally managed to assassinate the egyptians out of the game, and the golden horde and i finally crossed paths. they had me split and somewhat pinned at first, but with my vastly superior sea power i managed to brace up my lines and win one VERY decisive victory, so the horde is drying up also. they were pretty much the last major hurdle and i see almost nothing left on the board that can oppose me now. i pretty much own all of the middle east, all of the isolated island type areas and most of africa, so the game is pretty much in the bag. just a couple of small things to deal with in a couple more turns and i'll prolly call it quits.

one of my goals with this mod was to give the player a harder game. i think i've done that, but it's still a touch too easy. the removal of the land bridges is tricky. you know they are gone and how key that is. the ai knows they're gone, but under evaluates the importance of this. this is going to be difficult to compensate for. i was hoping by adding more shipping to the game that this would keep the player from having all the island nations, but so far, it's not enough. if i beef up those nations, and the player then decides to play one of those, then he's gone even more of an advantage, so that's what i'm going to be working on next. that, and maybe a few other tricks and options.

to me, i can live without the viking expansion. give me more editing tools, a better ai, more diplomacy, more shipping, some land trading routes, and more stuff like that. i can also live without real time ship battles, though that might be kind of interesting. there's also a few stats and formuli i'd like to see made available to the player, like the whole faction leader stuff and heirs.

the game still has my interest but i've been veering off to a little jedi outcast lately. cute game.

as for the ai ship battles, yeah, that one may be hard wired. it would be nice if that could be set and altered as part of the stat mods. that shld be at least part of the ai behaviour routines, but i'm not seeing it...yet. there's still some tricks to try, so we'll see.

K.

[cugel]

" can live without the viking expansion. give me more editing tools, a better ai, more diplomacy, more shipping, some land trading routes, and more stuff like that. i can also live without real time ship battles, though that might be kind of interesting. there's also a few stats and formuli i'd like to see made available to the player, like the whole faction leader stuff and heirs."

Amen to that brother http://www.totalwar.org/forum/non-cgi/emoticons/cool.gif That's what I'd like to see most. I might be interested in an ancient/Roman expansion, but not much interest in Vikings. Better AI is the holy grail for me anyway. I never have understood why chess programs can be made that can beat the world champion, but computer game AI is still so poor. I guess most of the limited development time and dollars goes into graphics and game features, but I wish better gameplay was more of a priority. http://www.totalwar.org/forum/non-cgi/emoticons/rolleyes.gif In fairness MTW is better than most game AIs. Many companies seem to take the attitude that if you want a real challenge, you should play a human online and the AI is for show. Since I don't play online (connection problems) that doesn't appeal to me.

A couple of general observations in light of my experience with this mod:

1. One should remember in adding resources to every province that it is impossible to trade with a province that has the same resources (naturally they don't need to trade for olive oil if they already produce it). Thus, trade income for trading with a given province may actually decline if everybody has that resource. When modding, every effort should therefore be made to vary the resources to avoid duplication as much as possible, in order to maximize overall trade.

2. The AI does seem to attack the player's shipping much more than the other AI factions (which it should in order to make things more difficult). Partially, this is influenced by the AI behavior set (the AI does seem to defend its home waters most aggressively). However, the catholic_trader behavior doesn't appear to do this IMHO. The Italians don't seem to act as aggressively when I, as the Sicilians place a ship in the Adriatic or off Genoa, but the Egyptians and Spanish reliably attack me when I do this. I have not found a way to make the AI keep it's trade lanes open. Even the Italians (Catholic_Trader) don't keep a ship in the Adriatic on a regular basis, which dramatically reduces Venice's income. The Byzantines (Orthodox_Defensive) are worse, they now produce more ships, but I've never observed them to keep a ship in the Marmara. Thus, Constantinople, which should have the greatest income in the game, doesn't. I would be interested to see if your experiences differ or if anyone has a solution for this.

3. I believe that there is no way to keep the AI from suicidally attacking your shipping unless they are set to pure Catholic_Trader. Has anyone found differently? Playing as the Sicilians, the Italians (Catholic_Trader) have never attacked my ships unless they had a decided advantage in numbers. The other factions always attack sooner or later, forcing me to sweep their fleets from the seas. Unfortunately, there seems to be no way to get the AI factions to accept a truce, even when continuing the war is disastrous for them. (Ex: the Egyptian sultan was in Crete. They attacked me without provication and I wiped out their fleet. Now their Sultan is blockaded in Crete so that rebellions break out all over their empire, but he won't accept my repeated peace offers. Naturally, the other AI factions sensed a weakness and attacked and seized the Egyptian provinces.)

4. Increasing the happiness building preferences seems like a very good idea, and one I'll have to copy. I've already set the rebellion factors to 2 for all provinces (except where it was already higher). This makes rebellions break out all over the place (negative result), but it also makes it much harder for AI factions to coalesce into 1 super faction and a few minor ones within the 1st 75 to 100 years(which seems to regularly happen otherwise). More interestingly, eliminated factions always (in my experience) stage at least one and often 2 comebacks via rebellions. I like this result very much as it keeps more factions in the game for longer periods.

4. Changing the Chapter_House build requirement to Castle9 is also a very good idea. Crusades are historically accurate, but don't work very well in terms of gameplay because the AI doesn't handle them well. We all know about crusade pathfinding, but even without that, it is often a bad idea for the AI to launch a crusade because it leads to a disastrous war that they absolutely didn't need. I've modded to make crusader knights and order foot available to all catholic factions (this is also historically accurate, since these knights weren't simply limited to crusades), so they don't need crusades to get these valuable troops. I made the build requirements for them chapter house, horsebreeder's guild and spearmaker's guild. Now, however, it will be a bit more difficult to make them, so there won't be too many in the early period.

5. Perhaps the most successful change I've found is following DOC's suggestion to lower the build times for all buildings by about 1/3. This dramatically speeds up development and reduces what I belive is the biggest problem with the game "DESERTIFICATION". As the player I seldom lose a province, but the AI factions trade provinces back and forth like poker chips, largely because they don't place enought emphasis on leaving a large army at home to protect their highest developed provinces. Every time the province changes hands more buildings are destroyed. After 150 years of this the player generally has a huge developement lead. That lead is insurmountable so long as rebuilding takes 50+ years (the AI doesn't place enough emphasis on rebuilding its provinces anyway). I might even lower the build times more (to 1/2) to see if this further helps the AI. True, the player can now build much faster, but I generally max out several provinces by about the mid 13th century and with about 350,000 florins in the bank can afford to build anything I want anyway. I think the AI should get more of a benefit from this change.

[Edratman]

Kraellin, no wonder we seem to take a similar approach to the game, we are both old farts. The game has lost much of its shine for me also. I no longer have any optimism that a mod will toughen up the game enough.

The AI needs to have a better threat evaluation system. It is why Aragon always loses, except when it has exceptional luck and why the ships attack. The AI only looks at one province and not the total forces of the faction it attacks. The sea attacks of my current game point that out. Byz shared 7 sea zones with me, in 4 zones I had a 2 or 3 to one advantage, in 2 others we were even and in one Byz had a 2 to 1 advantage. So it attacks me in the one sea zone where it has the advantage, but does nothing in the other 6 zones. Now all 7 sea zones are red to me and I have no choice but to attack and take advantage of my superior numbers. So the AI only saw the one zone with a local advantage and ignored the fact that overall I had a nearly 2 to 1 numerical advantage. I am at a loss to undo that aspect.
Small empires do the same thing on land. If they have a slight numerical advantage in an adjacent province they will attack an immeasurably stronger power. I think the pope and excommunication are intended to moderate this, but this is an empty threat once the player becomes wealthy and strong enough.
Also the AI attempt to get a numerical advantage probably explains why it likes to clump ships. I don't ever remember the AI clumping ships without a second faction present, there are always two factions involved. I now think both AI factions are trying to establish a numerical advantage and the clump is the result. Sometimes another AI faction sneaks a single ship into a 2 faction cluump, but never a second, probably because it can recognize that it cannot match the existing power in that zone.
So I think the AI definately needs to be strengthened to evaluate the overall strength of an an opponent, not the local situation.
I was thinking of trying to reduce the attack strength of all early era ships to zero, but in one of my few lucid moments I realized that would enable permanent blockade if the AI makes land war. So that game plan was scrapped.
A new ship, called a merchant ship that permits trade but does not exert zone of control or war making capability might be an option, but again that is a hard code issue. (Civ games have merchant ships with zero attack.)

I also lowered my build times by 1/3 a couple of mods ago. It was fun being able to build quicker and I agree the AI built quicker also, but as long as it has no significent trade income the player wins that race. To see this, all you have to do is watch the weakest provinces in an AI faction. They stop building new buildings, a sign that the AI is short of money and is allocating resources. So then AI is upgrading 2 or 3 provinces at that point, but I am upgrading all of mine because I have a positive income of 8 to 10K a turn. No contest. It is a marathon, not a sprint.

Cudgel, you must have been fortunate not to have Italy attack your ships. I have not had that experience, ever. Because Sicily and Italy both start with ships I find them to be the very first to attack and my recollection is that Italy is first more often. I think all behavior modes attack the very same way. All it needs is local numerical advantage in a sea zone adjacent to its own land and it attacks. But it should at least carry that inital attack over to more sea zones, using the Pearl Harbor strategy. A futher tetimonial to the inability to add more trade to the AI, in my last game even the Pope's ships attacked me and the pope only had 2 leaky rowboats and I had over 30 ships blanketing every sea zone. Once again, a failure on the AI to make a successful overall threat analysis.

I think CA also tried to cover the AI shortcoming with the Horde. While its historical precendent was handy, the massive Horde cannot afford its support costs and are beatable because of the limitations therein.

Only two possibilities remain to my thinking unless Kraellins wishes are granted:
1. Give the AI factions a gazillion florins at the beginning and handicap the players faction. This in my opinion will merely make the game harder, not more fun.
2. Drop the support costs for all units. This may have possibilities, but I'm sure there are associated pitfalls also.

I agree with Kraellin, don't give me a MTW expansion with the armies sporting slightly different uniforms and buildings. Make a better decision tree for the AI.

I had 4 months of fun with the game. It was worth the money.

[Edratman]

I think I've got it An example of turning a finely tuned Engineering mind (sic) to solving a problem.

How about delaying the merchant1 build? Make the prerequisite Castle9, or what ever you choose.

Think about it. We can't get the AI to trade, so limit our trade. This would level the field. The implications are enormous. Probably can't afford a world wide network of agents, or ships, so you'll play blind. No more bribing of rebel provinces whenever you want. Spend every florin like it was real money.

I'll wager the AI is good enough under that condition. I know my playing style would go down the toilet.

Got to run.

[Action]

Why not just take out trade goods all together and rebalance the farm incomes and mines available.

Right now all trade does is unbalance things in favor of the player. With just Sicily Naples Malta and Crete I was able to get the largest income of any faction in your mod.

[Kraellin]

action,

yes, we may have to go that route. my mod was mostly an experiment for several things, the land bridges, extra factions and to try and get the ai to make more money thru shipping. the problem, as you can see from the this thread, is that more shipping equals more ship fighting, so the ai essentially ends up shooting itself in the foot by having more ships. i'm beginning to believe that it might almost be better to either turn off ship fighting completely or, make water battles non war-causing, such that a ship battle doesnt precipitate war; it's just one isolated incident that isnt considered towards creating wars. but, we cant turn that aspect on or off at all in the game.

and, the ship stuff is something we asked for from the shogun days when we had these instant teleportations from port to port without any representation of the lines of travel and it always seemed a bit goofy that you couldnt block these lines at all. so, we dont want to go back to that either.

if the ai didnt initiate any battles on the seas it would prolly be ok. those wars it initiated on land would still affect the sea stuff, but it's just too idiotically aggressive on the seas. i dont mind and even like that it uses the seas for land invasions. that's fine. but that it makes one ship and then immediately attacks all other shipping is really kind of silly.

cugel,

hehe, computer chess programs have been working on the AI for well over 20 years now. and it's only in recent years that they've really been able to best the grand masters at all, so we'll give CA another year or so to get the ai right ;)

have you checked this closely, the thing about same trade goods NOT being traded between provinces? i've not really looked at this all that closely and was never sure one way or the other. you talking here from actual observation or assumption? i'd like to know this one for sure.

also, i'm wondering if we can add and subtract events like the marco polo event where extra trade goods are added during the game. this is another aspect that could be expanded upon to add more dimension to the game. surely war would tend to destroy various trade goods in a province, and why not allow for the development of more trade goods as part of expanding your potential in the provinces.

yeah, it would be nice to have a breakdown of the ai behaviors. the italians do seem more passive on the water and the egyptians were very aggreesive near their home, but not in open waters. and yes, i've noticed a trend in most factions that they'll very aggressively attack anyone in their home port sea zones, but i'm not sure that that is limited to a given behavior.

ed,

yup, the ai doesnt seem to look at the whole picture when deciding to attack. that is one of its main weaknesses, but i'm not sure how good it would be if the ai did look at the overall. small factions need to expand, thus they are going to have to take risks at times, so how do you balance and assess when it's ok to attack a bigger faction and when is it not? you'd have to also look at not only how big your opponent is, but also what he's involved with at the time. is he already fighting 2 other factions and thus he wont be able to handle you as well, or is he just attacking rebel provinces and slowly expanding and you, if you attacked, would be his only real opponent. i do agree that the ai isnt looking at enough, though. and this is also where an expanded alliance routine would be extremely welcome. a small faction could ally with a larger one to weaken and attack an even larger faction through coordinated attacks, much like the old Risk games where if anyone got too big, everyone joined in against them.

and this is again why i say, screw the viking expansion at the moment and let's have a mechanics/ai/diplomacy/trade expansion pack first. there's just so much potential untapped yet in this game, that i dont quite understand the current move by CA to make the whole viking thing. i dont recall ANYONE asking for a viking invasion. i DO recall many asking for added mechanics, however.

hmm, merchant and war classed ships. i'm not sure. it would help some things, but also hinder some others. that one is worth pondering. there's actually no need to represent the merchant ships on the field of play at all under the current 'trade routes' system. you could do the same thing with route markers on the map; if you also had war ships. then, you just put the war ships on the sea as ships and adjust the ai for given behaviors. but that's an entirely new system and still isnt really a solution for the ai aggressiveness on the sea. i mean, if he's building war ships also, then you've still got the problem of the ai shooting itself in the foot by making too many wars at sea.

however, i suppose if there were two classes of ships, we could then tweak the building priorities for both classes and just have the ai build more trade ships and fewer war ships and cut it down that way, so it would have advantages over the current system. the simplest solution would be to just have the ai quit attacking so much at sea. leave that to the land forces.

**************

ok, i've quit my current game now and am looking at stat files again. i've pretty much neutralized the horde, wiped the almo's and hold everything south of italy, from constantinople to morocco and ALL the island stuff. nothing left that can really oppose me now. the game went to the year 1310 or so, which is longer than it normally goes, so i take the russia Q mod as a success in that regard. the trick is still to see if there isnt a way to get the ai a bit smarter on the seas.

there's also some things in the files that i just dont know what they do yet and some things that actually surprised me when i started looking at things again. for instance, there are NO maintenance costs on ANY buildings. that seemed odd. when i mentioned this to tosa yesterday, he said, try using negative numbers in the income slot. interesting idea. there are also NO loyalty bonus buildings even though there is a column for this. interesting.

lemme know what you guys are tweaking and what you're finding as a result. i'd really like to get a handle on this ai game before moving on to the next large project i have in mind.

K.

[cugel]

"have you checked this closely, the thing about same trade goods NOT being traded between provinces? i've not really looked at this all that closely and was never sure one way or the other. you talking here from actual observation or assumption? i'd like to know this one for sure."

Kraellin: I'm sure. For example, in my mod Antioch produces silk, spices and gems, providing 199 florins from each province with which it trades all three. However, looking on my econ. screen I notice that it only gets 142 florins from Flanders and the little silk icon is missing from that province. Checking Flanders I see that it produces silk, thus they don't want to import any (can't trade coals to Newcastle). http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif

I haven't noticed the AI attack other AI factions with its ships as much as it attacks me. In my current game, the Italians (though reduced to 1 province) have the 2nd largest fleet and could crush the Almohad and Byzantine fleets, but they don't. The biggest problem I observe is that the AI doesn't consider trade a sufficient priority. It's always trying to get a numerical superiority in every sea province where I maintain a fleet, then it attacks me. Naturally I'm not about to let this happen so I make sure to build more ships than the AI factions can and make sure also to keep a local superiority in every sea zone where there's another ship. If I can't do this I withdraw my shipping for a while from hostile zones until I produce more ships and can regain the local superiority. This may be why the Italians don't attack me much as the Sicilians. (They would if they could gain a 2/1 edge anywhere, but I don't let them get it). My point is that other factions aren't even that logical. I've been attacked even when I have a local equality, not to mention a crushing overall numerical superiority when considering the total size of my fleet. The Egyptians seem bound and determined to do this if I send ships into their home zone, the Italians don't in my experience. I surmised from this that the Italian AI being set to Catholic_Trader had something to do with it.

In no instance does the AI, even when set to _Trader give any noticable consideration to maintaining a continuous chain of shipping to permit building a trade network. My chief complaint to CA is that they added this "chain of ships" requirement for trade (dubious historically at best since most of these powers didn't maintain large fleets and certainly didn't spread them out across all the shipping lanes) THEN THEY DIDN'T PROGRAM THE AI TO DO IT http://www.totalwar.org/forum/non-cgi/emoticons/angry.gif
None of the factions, whatever their behavioral AI seem to bother to keep a fleet in all their home waters zones in order to enable them to have overseas trade at all. If they do have a ship there, their trade shoots up, but they don't seem to care, they just generally move it out again the next turn (unless they're trying to get a superiority over me in that zone, in order to attack me and start a disastrous (for them) war).

"and this is again why i say, screw the viking expansion at the moment and let's have a mechanics/ai/diplomacy/trade expansion pack first. there's just so much potential untapped yet in this game, that i dont quite understand the current move by CA to make the whole viking thing. i dont recall ANYONE asking for a viking invasion. i DO recall many asking for added mechanics, however.

I totally agree with you, but imagine what a marketing nightmare that would be: "Now appearing TW: Med. Expansion Pack: IT'S THE SAME GAME, BUT NOW WITH IMPROVED AI" Can you see that one leaping off the shelves? http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif About the best we can hope for is that CA is paying attention to these discussions and will change the code to incorporate some of the ideas we discussed. (Keeps fingers crossed).

BTW, I'm not holding my breath expecting much imrovement, are you? http://www.totalwar.org/forum/non-cgi/emoticons/rolleyes.gif

[cugel]

[I]"
I also lowered my build times by 1/3 a couple of mods ago. It was fun being able to build quicker and I agree the AI built quicker also, but as long as it has no significent trade income the player wins that race. To see this, all you have to do is watch the weakest provinces in an AI faction. They stop building new buildings, a sign that the AI is short of money and is allocating resources. So then AI is upgrading 2 or 3 provinces at that point, but I am upgrading all of mine because I have a positive income of 8 to 10K a turn. No contest. It is a marathon, not a sprint."

Edratman: Yes, but there's a limit to the relative advantage the player can derive from more income. In my current game it's 1215 and I have 393,000 florins, but what can I do with it all? I can only build one building at a time in each province, only create 1 troop per province per year. And, does that armor4 really provide that much advantage in battle over armor3? Generally, that extra income goes to waste. I suppose that I could go around bribing every army on the map, but I don't play that way. If someone wants to win the game that way, more power to him. He paid his money for the game like the rest of us. My point is that it's the RELATIVE AI disadvantage not the ABSOLUTE AI disadvantage that counts. The AI can put up a reasonable fight if it can just produce good troops (and for this it needs the buildings) and enough of them, a few more bells and whistles available to the player from his extra income is nice but, isn't that much of a RELATIVE advantage. I must choose between the standard build times (in which case the player wins by default because of the problem of AI provinces increasing DESERTIFICATION - after a while they're reduced to a keep in their main provinces and can't build the best troops) and reducing the build times (in which case the player can build more with his superior income).

[Edratman]

I tried the merchant1 prerequisite of castle9 yesterday (thanks to the holiday) as the HRE and I'm getting thumped.
The AI seems quite competent playing under these conditions, and actually looks better than me right now. I've got all the Catholic behavior modes set to expansionist of one sort or another and after they got bounced out of the starting rebel provinces because they all have revolt factors of 2 or 3, they turned towards the HRE. The French penchant for going after England first is probably all that has kept me going so far.
I'm going to have to restart with some easier faction to try this out.
Cudgel, you are right about not being able to use the money in the bank, but it also means the player can do everything he wants and does not have to choose. The AI is forced to select where and what to build because so much of its income is going to troop support. It will often have to build peasants or some other low level troop late in the game just for financial considerations. The player does not have to do that with money in the bank.
Another change CA could make is to at least allow land locked provinces trade with adjacent provinces. That would average about 5 trade partners, not bad, and also not blockable by a solo rowboat.
Kraellin, you are right about CA offering only "Upgraded AI" in the expansion.
You know what the problem is, we are looking for perfection. CA is looking to make money and perfection is sacrificed for that goal. And most companies prefer the glitz of graphics to gameplay. What to do?
I have no more ideas or thoughts.

[Kraellin]

well, i've gone back to newbie status as a modder. made a TON of mods to my stat files yesterday and saved them with the whole tab delimited thing in excel and it totally screwed them up and i was so cocky and lazy that i hadnt made backups first...d'oh. so i spent two hours last night removing tons of tabs and adding tons more to straighten out the files. lol.

i had really thought that my excel in office 2000 saved these things correctly, but why i think i thought that is that i saved the early.txt file in excel and it did fine, so i guess i figured it would work in the build_prod and unit_prod files ok also...heh. no joy. ah well, live and learn.

i did have one version work before i messed up the files and found some interesting things out. i've now got the almo's building a much better fleet and the byzantines. i've also found out and am still testing things about the ai behavior routines. tricky, since there is no file you can look at to see what these behaviors are supposed to do; you can only try em out and observe patterns and so forth, but i'm slowly getting a feel of it.

i'm also doing some oddball testing of things and so far i've managed to crash the game, or have the early.txt not show up in the campaign list. lol. gotta push the envelope ;)

i'm purposely being pretty vague here right now, because of what i'm trying out. i'm going a different tact than some and i either want to get it working or totally fail before i release the mod and what i've done....it's going to be a while ;)

one interesting thing that is showing up again, that showed up in the russia q mod; the spanish do build ships and they once again get into a war with the almo's, and, since both factions are now building ships, the almo's effect a bridge to spain with ships, but the spanish build ships and send single units down to attack the almo ships. this is very cool. there is little question in my mind that the spanish ships are doing anything other than that, at least as their first priority. they build up 2 or 3 in their home port and then send out a ship to the straits of gibraltar to attack the almo ships. the ai either knows the importance of the straits or knows where the almo ships are or both, because it always does this. it doesnt build a trade route there and just run into them by accident. it builds a few ships, defends its home port and then sends out a ship to go nail the almo ships. quite nice.

i'm also getting a better feel of the numbers needed to get the ai to actually produce more ships. more on this later.

K.

[Edratman]

Kraellin

Sounds like you have found some interesting developments. Patience is a virtue in my profession, so I can wait while you iron out the bugs, but if you want some help, all you have to do is ask.

I see the quality of the AI is surprising you also. That is what I am finding out without trade available. Now I can't discount the factor that my game style relies on trade income, but I am over matched at the moment. (Definite proof that I learn how to exploit the weaknesses of the programming.)

I had so much trouble with Excel mods that I avoid it now. Too bad, because it is so much easier. I just use it after the fact to see if I made a spacing change when working in the build files, another mod error I had to learn the hard way, one of which required a complete dump and reload. (I think I read somewhere that GilJay said they wrote the programs in notepad, but I doubt it. Startpos yes, but the build files had to be done in Excel, the spacing is just too touchy.)

Good luck in this new direction, I've got faith in your imagination. Besides, who else is going to uphold the honor of us old farts?

[cugel]

http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif Well, now Kraellin, I made the same mistake you did. I started out modding with early.txt and thought "well, that seems to work with excel so now I'll try crusader_unit_prod." CRASH One total system lockup later I learned humility. Now I just use wordpad, change a line, test, change a line, test . . . Definitely slower and more humble, but that's o.k.

"i did have one version work before i messed up the files and found some interesting things out. i've now got the almo's building a much better fleet and the byzantines. i've also found out and am still testing things about the ai behavior routines. tricky, since there is no file you can look at to see what these behaviors are supposed to do; you can only try em out and observe patterns and so forth, but i'm slowly getting a feel of it."

I'd definitely be interested to hear what you think about the AI behavior routines. I'm seeing a lot of different stuff come up with that. For instance, now that the Spanish build more ships they are highly aggressive with them. Same as you I found they didn't try to trade at all, they just wait until they achieve parity with me in any sea zone to attack. I had to wipe out their fleet. Then they collected 6,000 troops and invaded one of my provinces. I think that this is due to Catholic_Crusader. I believe that AI personality is highly aggresive. Sometimes it's just not evident because they are often locked into a bitter struggle with the Elmoheads and French/Aragonese and close to their support limit. In my game they expanded and took over France & Iberia so they have a bit more room to maneuver.

BTW: How did your Byzantines do trade wise? Mine never seem to do very well no matter how many ships they build, which I find terribly annoying, given their historical status (even noted in the intro text) as THE major trading power in the Med. during this era. Ah, well. http://www.totalwar.org/forum/non-cgi/emoticons/rolleyes.gif

[Kraellin]

hehe, i've got ships coming out my arse ;) i think i'm close to figuring out how to get this shipping stuff really cranking. the byzantines are now super powers because of their shipping and trade routes...quite scary.

the almo's also build a lot of ships, but get VERY protective of their home port in morocco. they'll spread out if no one is around, but if one ship moves into their home port, they'll bring almost their entire fleet back to home base to protect it. talk about paranoia.

the english and the danes are now fighting for control of the north sea and english channel, but somehow in the 2nd version of this, the danes arent building hardly any ships, so something got screwed up there.

the germans and french still seem to be shy in building ships for some reason. they do build some now, but gotta look at the stats again.

the swiss and the turks are building ships. lol. not many, but it's a start :)

my spanish give varying results. they build ok, but their deployment is quite erratic. sometimes they sit and protect the home port and sometimes they use the home port to raid from and sometimes they spread out. not quite sure what's going on there yet, but at least they build.

the aragonese are the real surprise. they are now capturing provinces by having more money from their limited trade route, but they do seem to use trade routes fairly effectively.

the italians are building again, but seem bent on holding the sea zone around genoa, which always starts a war with the burgundians, so hard to tell how effective they might be otherwise.

i turned the egyptians way back at first but now have them back and building ships. they almost always end up fighting the byzantine through ship battles in the egypt sea zone. the egyptians are set to muslim devout this time, instead of muslim expansionist and it seems to make no difference in their aggressive sea action. they ALWAYS try to drive everyone out of that one sea zone.

ok, so here's my reasoning on trade and the aggressive AI. we know that the ai tends to attack when it has the advantage, at least with some of the ai behaviors. so, let's take advantage of that. we cant change it, so let's use it to our advantage. that means, get the ai building a TON of ships and let's go ahead and fight for the seas. this will keep the human on his toes and cut down the human trade routes somewhat. and that's exactly what we need. the human can all too easily run away with the game thru trade routes and money, so let's get the ai VERY aggressive in this regard.

before, i was going for trying to get the ai to be less aggressive on the seas and expanding his trade by keeping the peace and simply spreading out. well, to do that, you almost have to set all the ai factions to a very few ai behavior types, like muslim_peaceful. and that just isnt very good. oddly, so far, orthodox_stagnant is producing the best trade route routines. the byzantine in my game are now out-producing me in ships and trade. it's quite impressive. if i can just get the others to do this, i'll have a great mod, one that folks have been wanting for a while when they noticed that the ai sucks in trading.

i've modded quite a few things, including some ships costs, ship building time, ai preferences, starting locations for factions and a number of other things i'm not going to yap about quite yet (specially since they arent working right yet:).

i've totally abandoned my 10 year rule for this. if they fight on the opening move, so be it. and, i've gotten some interesting conflicts. the aragonese sending a crusade to cordoba was a shocker. the byzantines invading tunisia by sea was interesting. the danes taking over england was cool. the german naval fleet mixing it up with the danish longboats was funny. the papacy invading switzerland was quite unexpected. lol. gotta watch that old pope ;)

there are some observable patterns emerging in the ai behaviors. most behaviors that have 'defensive' in their name seem to do the protect-the-home-base-first routine. the spanish are catholic_defensive_crusader in my mod. other 'defensive' ai's seem to operate similarly. the muslim_devout also seems to act like a defensive behavior. the muslim_peaceful will build and spread but then come home if threatened. in the last game, there were 16 stacks of ships in the straits of gibraltar with a total of 22 ships of different factions, the almo's having spread out to begin with but then brought damn near everything back when 1 or 2 other factions stuck ships in there.

ok, and here is one of the keys to this stuff. we werent really certain how those numbers after each behavior were affecting everything. the way it seems to work is that it's a ratio of any given number to ALL the other numbers of all other things. for instance, if you raise the muslim_peaceful(20) on the ship stuff to muslim_peaceful(400), the code says ok, take ALL the other units of muslim peacful, including the land units and compare these numbers, then build based on the highest values most of the time, but not always.

we also know that the ai alters its behavior based on what's going on in the game. it even states this in the early.txt file. what we dont know is, which of these ai behaviors are switchable. can a muslim_peaceful become catholic_crusader_trader, for instance, or, does the ai simply use the poverty_stricken, desparate_defense, close_to_support_limit, and rebels behaviors as the one's that can be switched to? my recent thinking is that ALL of the behaviors may be being used to switched to. remember, just because it says 'catholic' in the file, doesnt mean the code reads this as catholic only. this is just a variable name, just a tag for reference or for starting out a faction behavior. it doesnt necessarily mean that a catholic faction cant use a muslim behavior. it's just a tag. it doesnt mean anything particularly. you could have named them behavior_a, behavior_b, behavior_c and so forth. you get what i mean here?

now, to test this, i decided to change some of the starting behaviors to include those 4 oddball behaviors, poverty, close_to, rebels, and desparate. the game wouldnt show in the campaign list. so, one or more of those 4 is different in nature than the muslim, catholic, orthodox ones. i suspect it's the rebels, since if that one is set, it wont EVER change. that is also stated in the early.txt file. more testing is needed on all that yet.

now, going back to the numbers at the end of the muslim_peaceful type references. you MUST look at the land based unit numbers. some of those numbers are quite high, 600 and over, so, even if you went from muslim_peaceful(20) to muslim_peaceful(400), you may still not be giving ship building a very high priority relative to peasants(600). you MUST look at ALL the unit values. and then, it's just a matter of balance, getting more ships, but not so much that the ai becomes terribly vulnerable to land attacks.

now, what i dont know here is if the code is taking units AND building preferences into account as one thing. do the building preferences and the unit preferences get taken together or are they separately considered when the ai is spending its money? that one i dont know yet. my earlier guess was that they were taken separately.

ed,
there are a number of things that need to be tested. if yuuki was doing all this he'd very carefully test each individual stat and setting and isolate all this stuff out systematically. i'm terrible about doing this. i tend to do 50 tweaks at once to get an overall change and then end up wondering which tweak did what, but, i get bigger changes and more diverse effects this way. i find the tweak the one stat, test ad nauseum, tweak, test, tweak, test style a bit too mundane for my tastes. i'm too impatient. if you were so inclined, testing some of this stuff one by one would be helpful, like the poverty, rebels, desparate, close_to stuff. which one of those is the one that prevents the game from showing in the campaign list when it's given to a faction as the starting behavior...stuff like that. in other words, answer some of the questions i've listed above. i'm currently bent on getting all the factions positioned on the map better and getting them all to do more shipping.

oh, and that's another thing; because we know that some ai behaviors are quite aggressive, locating and positining your starting ports for the factions is quite critical now. i'm having to re-arrange some of the province ownerships to spread out the starting ports better so that the ai can at least start to produce some ships. this is helping. the last change i did was to remove the danish from denmark. lol. sounds goofy, and is terrbily non-historic, perhaps, but i go first for game play, so, if the english end up starting in georgia, so be it ;)

cugel,
i think i answered all your points in the stuff above, so nothing really specific to direct at you personally here, other than keep the faith and keep on truckin :)

K.

[Edratman]

I'll start some testing tonight. Anything real specific you want me to check?

My thinking regarding the poverty, close to support limit and critical defense descriptions is that the name is fully descriptive of the mode. They are the purchasing patterns of building and troops (not to mention hard coded decision tree activities) that the AI switches to when it is deficit spending on support or losing provinces at a certain rate. Whenever I changed the build values for anything these categories I left alone. But I always included rebel in every change.

Great thought on the possibility of applying cross religion behavior patterns. It never occurred to me. If I don't hear from you before I start I will try this. I can envision some problems here, I know the build values have zeros for the religious specific buildings when they don't apply. I think troops have similar settings. So I will try one small step and just change one Moslem country to Catholic trader and adjust the moslem church (whatever it is called) value for catholic trader and see what happens.

If that causes problems I'll try setting someone to rebel.

It's a plan.

[Kraellin]

ed,

well, i was just suggesting trying the oddball behaviors, like desparate, poverty, desparate and rebels to the starting behaviors for a faction and seeing if it worked. i did this with 4 at once and my game wouldnt show in the campaign list, so at least one of them cant be the starting behavior and i was just curious as to which one that was. it might be useful for utilizing more of the behaviors at the start to give things a twist.

K.

[cugel]

Kraellin: Don't you wish Giljay or someone from CA would actually COMMENT on what these AI behaviour routines actually do within the hardcode, so we wouldn't just be guessing and generalizing http://www.totalwar.org/forum/non-cgi/emoticons/rolleyes.gif

Oh, well. http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif

So you think that the factions change into anything but _rebel? Well, I never would have guessed that I would think that poverty_stricken coudn't be a starting behavior or the AI faction would never get off the ground. It might shift to it pretty quickly (as Denmark seems to do early in the standard game).

[Kraellin]

frankly cugel,

i'm thinking of going over to the u.k. and camping out on their doorsteps with a tape recorder, a bottle of brandy, and some cuban cigars.

as for the rebels there is a note in the early.txt file that says that this one never changes once set. now, that implies that the others do change, as we've suspected. and, one of those prevented my campaign from showing up. i suspect it's the rebels one.

every time i start a new campaign with some new tweaks, i find different things that make me wonder. for some odd reason the egyptians are now NOT building ANY ships. in fact, they wont even build a shipyard. my best guess is that they are in the close_to_support_limit mode or something. quite odd, but i dont know why.

the byzantines have also slowed production a bit, but are still ok. the danes have picked up, but that was one of the tweaks this time, to get them building more.

i've also given prussia to the poles so that they have access to the sea to start with, but so far, no polish navy ;) (i'm making no cracks here, just smiling :)

i tweaked the spanish ship building down a bit and it worked mostly as suspected, so that was good. they were spending too much on their navy and not doing much with it and thus suffering on land. i REALLY wish CA had given us access to the ai routines for determining ship movement and ai behavior breakdowns so we could alter these to get more trade going.

i am having some success. the game is a bit harder with all the extra ships on the board. in one game i couldnt get the richest player message for the first 40 to 50 years. in the current game it's gone back and forth between me and some others. this is good.

i'm still always playing as the sicilians, however, to keep myself removed from the main action. i want to be able to observe the ai but not affect the game that much.

btw, does anyone know a cheat that allows one to view spies and assassins from the ai, and see the buildings in all provinces? the matteosartori cheat only allows me to see the normal units on the board and i cant look into the provinces themselves to see what's being built. this would help tweak things a bit easier.

ok, one more little datum here. the nov's are set to orthodox_expansionist and they also seem to be one that doesnt want you in its home port, unless you have the bigger force, of course. so add that one to the list of home port aggressiveness.

one of the other things i did was to move the danish out of denmark. since i've got them set to barbarian raiders i wanted them isolated from mainland land attacks so that they could act more like raiders. this has worked pretty well. i did have to bump norway's income up a bit, though, to keep them from dropping into poverty_stricken. i also gave them a bit more starting money. but what happens is that the english, whom i have as catholic_naval_expansionist tend to take denmark by sea invasion, and this gives the english just too much income, so i'm cutting income in denmark. i also gave the danes a starting ship in sweden.

once the danes get off the ground with ships, they really do well. with their trade routes in the northern waters, sweden is making 1000 or more per turn now and norway almost 400. sweet :) this is the sort of thing i'm after.

the danes, the english and the byzantines are doing very well with shipping and trade now. i've still got a problem with the burgundians and the italians because the italians always want to rush over into the sea zone (ligurian sea?) outside genoa and that's the burgundians home port. thus, the two always end up fighting and both then suffer in shipping. i'm thinking of altering the burgunian home port as a result.

the aragonese are doing ok. they dont have much income but do build ships and so far, do expand ever so slowly, so i'm mostly pleased with them also. they do clump up a bit, but also do move outwards.

the battle around the black sea is another interesting story. the bzy do great. the turks even build a few ships. and with all the rebel provinces ripe for picking, there is always a nice power struggle going on tween the turks, the byz, the russians and the novs. even the polish and hungarians get sucked in at times.

generally, i play one game for about 40 moves or so and make notes on things i think need altering. i make the tweaks and start again. 40 moves is about what i consider the 'opening game'. i can usually get enough of a feel for things to see how the balance and game play is going with that many moves. i realize that any campaign as it is would produce different results if you played it 10 times over, but once is usually enough to see the types of things i'm tweaking, so, it works.

here's another little tidbit on the almo's, or more specifically, muslim_peaceful. they NEVER initiate a war by sea. they will respond if attacked, but they never start a war by sea. in fact, i dont recall them starting one by land either. they do wage war, but i dont recall them ever starting it, land or sea. too bad they suck at spreading out their navy for trade.

so far, if i had to pick one behavior that is ideal for trade, it would be a toss-up between catholic_trader, orthodox_stagnant, barbarian_raider, or catholic_naval_expansionist. i might drop barbarian_raider cause i think i remember them being a bit pissy around their home port in one game, but they do do pretty well at sea. i'll let you know more as i go along.

K.

[cugel]

K: Sounds like you're much more ambitious than I am with all theses revisions http://www.totalwar.org/forum/non-cgi/emoticons/cool.gif I modified my early.txt, unit_prod and build_prod files much as we discussed with DOC earlier and I've noticed some good, some not so good results. Frankly, I've been reading Philippe Contamine's War in the Middle Ages and it struck me how unhistorical this all is. "Money remained scarce almost everywhere. Rulers and powerful men had limited and intermittent supplies of cash. The maintenance of soldiers in their service was preferably assured by lands which were conceded to them, in practice in heredity, in return for performance of a certain number of duties of a military character. Certainly payment of soldiers was not an entirely unknown phenomenon, but it was a secondary, marginal one, less important in any case than the upkeep of castle garrisons at the expense of rich, powerful lords." Lack of income to pay a large standing army like the Romans had is probably the real reason why no-one succeeded in unifying Europe during this era.

Still it makes the game much more fun, I have to admit. http://www.totalwar.org/forum/non-cgi/emoticons/wave.gif

After making a few changes, I played again yesterday and I'm still having trouble with the Danes, despite my giving them Norway (from whence they immediately take Sweden) and starting them with 16,000 florins). They're still not building any ships, despite the fact that I gave them a port and shipyard2. http://www.totalwar.org/forum/non-cgi/emoticons/angry.gif Oh, well, back to the drawing board. http://www.totalwar.org/forum/non-cgi/emoticons/rolleyes.gif

You took the Danes out of Denmark?? http://www.totalwar.org/forum/non-cgi/emoticons/confused.gif That's a bit radical http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif But why did you make them Barbarian_Raiders? Why not try Catholic_Trader? Their raiding days were really pretty much over by the beginning of the game anyway and they were acting much more like other catholic powers "forming lasting settlements in lands which became recognized as their own...." (Contamine, op. cit.). In addition, they historically held Norway and the southern portion of Sweden throughout the middle ages. Personally, I like to get as much historical accuracy as I can, commensurate with having fun with the game, of course (not that you have to, of course).

A further 10 cents worth: if you're having trouble with Burgundy, why not up their rebellion factor to 4 and don't worry about their trade income. They'll survive much longer as a minor faction (which is what they were during this era) since anyone who conquers them will suffer loyalist rebellions galore. This works with the Swiss also. If they get too strong, they can cause havoc in the early era. But with high rebellion factor in Switzerland, they can't move too many troops out to wipe out the HRE without suffering a rebellion. You might also up the quality of troops in such rebellions as a means of ensuring faction survival. (The best strategy for Novgorod still seems to be to get wiped out and then have loyalist rebellions break out all over the place. (I've got them to where this doesn't happen as much, but it seemed to work for them). http://www.totalwar.org/forum/non-cgi/emoticons/shock.gif

My point is that not every faction has to try to take over the map. Most were content to maintain their independence and influence events around them. It was very tough though for anyone to permanently subdue many regions because of various insurrections.

Well, when you get things worked out satisfactorily let us know. I'd like to take a look at it.

[Kraellin]

hi cugel,

the whole money thing is indeed a potential game killer. it might be interesting to make a minimalist type campaign where that really comes into play. this could be done and might make for a better, more realistic and accurate representation of feudal times. it's an interesting idea, and if you do go this way, i'd be curious to see what comes up.

as for the danes, hmmm, how do i put this...it works in my campaign, is pretty much the answer. the danes in the game have always been a problem. if you give them a ton of cash so they can expand it tends to warp the balance. if you leave them in denmark they have a hard time doing anything and, they are in constant threat from the HRE. i made them barbarian raiders for the simple reason i wanted to test out that ai behavior and the danes were the most logical choice for this, given everything else. now, being barbarian raiders, it also made sense to isolate them a bit from the continent, since i particularly wanted them as sea-raiding barbarians. i also found that they had a lot of trouble starting on the continent. they'd take something and never hold onto it. so, it just made sense, once i made them raiders, to take them off the continent.

my foolings around with mods are rarely done for historical considerations. i just dont know that much about history, so i dont make any pretense about trying to make things historical. what i do is try and figure out how things work and make them work better from a purely game playing perspective and then let the history buffs take that info and do with it as they see fit. it's like the difference between me and a history buff when he watches something like Braveheart. i see a rousing good tale of a fight for freedom and he sees a historical boondoggle full of speculation, alteration, and historical fabrication. frankly, before that movie, i'd never heard of william wallace :) so, it's just a matter of interest and background and purpose as to why i do the oddball things i do. i mean, horses dont really fly, but i did make a flying horse map for stw in which they did :)

what's most likely happening with your danes is that they are hitting either the poverty_stricken or close_to_support_limit behavior and clamping down on spending. that's what i found too. you have to get them over this hump and bringing in trade money and then they'll start producing pretty well. i also recently found it's why my egyptians never built shipyards in that one game i mentioned. i've now fixed that and they're building again.

the burgundian problem is multiple. they will build ships. the problem seems to be with the italians. for some odd reason, and i believe it's related to the GA goals, the italian port/shipyard is in venice in my game (cause that's where i put them), but they want to sail around to the ligurian sea (outside of genoa) and think that genoa is their home port, so they always start a fight with burgundians by sea, since that's the burgundian starting port sea also. now, if somehow the burgundians get out of the ligurian sea, they tend to go west into aragonese waters and boom, another sea battle.

ok, so why not leave burgundy as a minor faction. well, it is. the thrust of this mod, though, is to get so bloody many ships out on the sea as to make it more difficult on the player and his trade route advantage. that's been one of the biggest problems with the entire campaign; the player knows to build trade routes, the AI doesnt. so, i'm not so much interested in anything but creating ships for the ai at the moment. balancing things out will come later. i dont really have a problem with having minor factions. it's nice to be able to play as a major or a minor, but the the current project is simply to increase shipping, figure out the ai behaviors better and then tweak things a bit for balance. this is also why i gave prussia to the poles, so that they'd have access to the sea to start with. oh, and btw, i now have hungarian navies :)

yeah, i did a similar thing with the swiss; the whole rebellion factor thing. my swiss start with switzerl and tyrolia. tyrolia has a rebellion factor of 3 and switzerland is 2, i think. this is a very nice way of limiting a faction expansion-wise. so i agree with you there and i may try it with the burgundians if nothing else works. i'm also toying with changing what they own and thus moving their starting port location.

yes, i used to chuckle over the nov's and their rebellious popular revolts and their re-emerging and always coming back stronger. i did finally 'fix' them, but yes, it might be interesting to leave them that way, just for the added differences it brings to the game.

the one thing that is currently beginning to make me wonder about the way i'm going is, what happens when someone wants to play as the byzantines in my mod, or the english. is it going to be too much of a walk in the part. i always play as the sicilians, and play as a pretty much neutral observer, so what's going to happen with a very aggressive player when he starts just walking all over all these ship building factions? not sure, but for now, it's enough just to get the ai building ships.

K.

[Kraellin]

alright, i'm about ready to release this thing. the major goal of increasing shipping has more or less been accomplished. it aint purty, but i have increased it.

there's a couple things i dont quite understand, and may never understand, like why, if i set burgundians and italians to muslim_peaceful as their default, are they still attacking each other at sea. that one seems odd. and, oh, yes, you can set a catholic faction to muslim behavior and vice versa. these things are just tags like i figured.

still havent heard from edratman on what he found out in trying to set the odd behaviors to start behaviors, but that wont delay my mod.

one other thing i tried to do with this mod that i wasnt disclosing before, is, i altered one faction, the swiss, who i set to catholic_isolasionist, to create tons of agents. i wanted one faction that would more or less just sit there and spy and assassinate and create havoc through spies and so on, but not be a major military power. now, i cant really tell if it worked or not because i dont know a cheat that reveals hidden units, but i will say that i had more of my emissaries killed off than in any game i recall before, so maybe it is and maybe it isnt working. nothing like blind modding to leave big questions.

i also set the papacy to create more bishops and religious units, but didnt notice any change there. i suspect they just dont make enough income to do much. one side effect of this though, was that in setting some of the buildings and so on for this, i increased the ability of catholic factions to build crusades and 20 or 30 moves into the game this seems to take effect. again, not sure though. might just be a normal amount.

i've now got the byz, the italians, the danes, the almo's, and the english as the major sea powers, with the byz and italians clearly leading this race, and the danes and english a good second. the almo's and the spanish are third, and the aragonese and burgundians are 4th. other factions will show up on the seas as well. i've seen the papacy, the turks, the polish, the hungarians, the swiss, the novs, the french and the HRE all show up at times, but quite a bit later than the others and in smaller numbers. this is close enough for my purposes at this time.

i may do more with this mod, keep tweaking this and that, but i'm figuring to move on to another project soon, so this may be the last ai mod for a while.

there's tons of things that could be done with all this. you can move province ownership around to suit, make historical models of province ownership, buildings and units, tweak single ai behaviors to make tailored factions, like my attempt at making an agent only faction, or a cavalry faction. one could make a land-based barbarian raider. one could make an inquisition faction, say the spanish or something. by being able to tweak the behaviors and give them to different factions, there's all sorts of things available.

one of the things i didnt try that may be possible, is to give any religion jihads or crusades or both, or, to turn them one or both completely off.

one could set up a head to head game with a single ai player. you could divide up the entire campaign map between two factions only, half to the ai and half to the human and let them go at it.

one could set up a minimalist game, like i talked about before, or just max everything out and make everything easy to build and easy to support.

there also seems to be the possibility to give other buildings income. i was quite tempted to give inns, taverns and brothels some income in this last mod. and as tosa suggested to me, maybe it's possible to give buildings a support cost by simply making the income column a negative number, like say on castles, for instance.

bear in mind that with this mod i was only trying to increase the early shipping, the early ships. no attempt was made to tweak the better ships that can be built later on, so your shipping by the ai may dry up as the game progresses.

this mod is also just an extension of the russia q mod. that was the base mod i used and simply took that and carried on with it, so there's still 19 playable factions, no land bridges and so forth and so on. the biggest difference now is that some of the ai factions actually do a decent job of ship building.

it shld also be noted that i can do almost nothing to affect HOW the ai distributes and uses his ships. i could do more testing with the various behaviors to see if it that is the determinant on distribution, or if it's like doc says about the choke points ont he map and simple geography that is mostly determining things.

i had thought i was being very clever in giving the burgundians and italians both the muslim_peaceful behavior so that they wouldnt always fight when they met in the ligurian sea, but the code outsmarted me by either breaking the rule against muslim_peaceful never attacking or by simply changing its behavior to something else, or in some other way completely that i dont understand currently. ah well.

i prolly wont release this today, but it shld be up within the next few days. i may yet do a touch more tweaking, so no promises.

K.

[cugel]

"i had thought i was being very clever in giving the burgundians and italians both the muslim_peaceful behavior so that they wouldnt always fight when they met in the ligurian sea, but the code outsmarted me by either breaking the rule against muslim_peaceful never attacking or by simply changing its behavior to something else, or in some other way completely that i dont understand currently. ah well."

OMG Set the Italians to Muslim_peaceful??? http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif
And I thought your other ideas about the Danes were radical

Ah, well I think I'll have a look at this. Don't know that I'm going to install it though http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif Seriously though, didn't that cause some conflicts with the Game engine? Or is that just if you attempt to build some building/unit belonging to another faction's religion? You can give Catholic factions muslim behavior sets? I wouldn't have thought of that, but: "why not?" if the AI is going to shift to that behavior set anyway (my guess is that it doesn't - that muslim and christian factions don't use each other's behavior sets). I suppose that's no worse that Giskard setting the French to eastern orthodox Heresy Heresy [covers ears] http://www.totalwar.org/forum/non-cgi/emoticons/joker.gif

I wonder if each game isn't rather ideosyncratic. If you only play 40 turns or so (I did this a lot too), you might miss some things. I was concerned about play balance, but the other day I played about 120 years as the Sicilians (just remained neutral and built ships) and I noticed that just about every faction (except the Aragonese) were still in it by 1240. Everybody who was eliminated made at least one comeback and some made two. The Egyptians built about 40-50 ships and took over most of the map. Then their assault on Scotland and Ireland failed and within 10 turns they were cut down to size (despite being helped out by an "Egyptian loyalist" rebellion in Tyrolia) (had to laugh about those Tyrolian Egyptian loyalists).

My only point about Burgundy and minor factions is that they can be made to be very independent and difficult to hold on to (which is historically true) and it also makes for better play balance. They don't have to build ships to do this - especially if it causes other problems for the AI. The player or AI that wants to hold them will then have to garrison large armies there or there'll be big loyalist rebellions there.

I've given up hoping to balance trade with the AI. It just doesn't get it. Rather like attempting to teach your poodle to read. If you can do it. Well, that's more than I can say. http://www.totalwar.org/forum/non-cgi/emoticons/cool.gif I'd say that the remaining issues for me are:

1. Crusades. The AI stinks at crusades, so next I'm going to up the build requirement for Chapter House to have fewer of them. I've made crusader knights buildable for all catholic factions so they should have a high build requirement or there will be too many such knights built in the early period.

2. It's still too easy to win as Byzantium. Nothing to do about this. If I cut down the Byzantines, the AI will do even worse with them than it does, which isn't too good. If I play as them, it's a cakewalk.

3. I might break down and stifle trade as a last resort. I'd hate to do it though because that's how I like to play and I can't really imagine any other style of play. If I did, I imagine that the AI might easily beat me, given the other advantages it has playing on expert. I'll have to see.

4. I don't think that anyone can really fix the problems that the AI has with trade so #3 might be the bet. Just make the build requirement for merchant castle9 or something. That should take the starch out of trade (and cripple the Italians into the balance). I you or I did that, it might be a good idea to make the Italian city states independent (as Paladin did in his mod), rather than "Italian" (which is historically accurate since Italy wasn't unified between the 5th and 19th centuries), and remove the Italians from playability - either that or start the Venetians out with a trade guild. and fleet?

I'd really like to see what those boys making the Lord of the Rings mod come up with. If it was good I'd play that next. http://www.totalwar.org/forum/non-cgi/emoticons/pat.gif

"Ringwraiths and Goblins and Trolls Oh, MY" http://www.totalwar.org/forum/non-cgi/emoticons/cool.gif

[Kraellin]

cugel,

this is what i was saying before, the behavior tags/labels/names are just that, they are names only and have no other significance than that, well, other than turning on that behavior. they could have been named anything. just because it says 'muslim_peaceful' doesnt mean that only the muslim can use it. this is what i was trying to convery to doc in our earlier conversations where i was talking about the ai changing behaviors from time to time based on the circumstances within the game.

now, i dont know that the ai does change to all the different behaviors, but it does seem to alter its tactics at times, so i'm fairly sure that any of the given behaviors, except maybe 'rebels' is available for any faction to use. this may also be why the italians i had set to 'peaceful' and the burgundians i had set to 'peaceful' went ahead and attacked each other anyways. they simply changed to something else.

the only reason that these names are the way they are is for easy reference purposes. if you simply drop the 'catholic, orthordox and muslim' part of the name off then it makes a lot more sense. just think of it without those religious parts of the tag on there and use them accordingly.

the one that i think you cant use, is 'rebels', at least not for a playable faction. you might be able to assign it to a non-playable faction, but that is something i wanted edratman to verify.

yeah, i almost changed the burgundians to a single province with no sea access. would have made things a lot simpler.

one of the nice things about this game is that you can make a lopsided game with the factions. you dont have to have a perfect balance. it's perfectly fine and even desirable that they arent all perfectly balanced. if the player wants to play a tough game, then pick a tiny faction and if he wants an easy game, then pick a big faction. this leaves it up to the player, which i like. my only concern when balancing is that the ai isnt walking away with things too awfully much faction to faction. i dont want one faction always dominating every campaign, at least the ai factions.

yeah, i'm afraid the trade thing is going to be difficult no matter what we do. the ai just doesnt 'think' trade routes that much. it's geared mostly to causing wars and taking provinces. it needs work. i'm sure CA is aware of this and will improve it as we go along. the whole ship thing is new to mtw and they packed so much new stuff into the game that i'm guessing that someone finally said, 'um, hey guys, this stuff is great, but at some point pretty quick, we've got to actually sell something', so threads had to be tied off somewhere and the shipping/trade route thing seems to be one of those.

this current mod is simply one attempt to try and work around the shortcomings of the ai ship stuff. i think it improves it a bit, but is far from being everything we might want in the whole sea/shipping arena. still, it's fun playing around with this stuff.

i've still got a few ideas i may try, so dunno when i'm going to release this mod now. if you really want to see it, i could bundle up whatever the current version is and post it in the test upload area, i suppose.

K.

[+DOC+]

A minimalist mod with little or no trade.... now that might be very interesting indeed and may even make the game harder in itself since trade is the main way we humans can whoop the AI.

Hi all.

Been busy with Civ3 and PTW... Got bored of MTW after i finished my mod. lol. When it gets to the stage of fighting 1 hour long 4000 v 4000 man battles multiple times every turn i get bored of MTW. Especially when you're dealing with hundreds of horse archers, man those guys are annoying to fight. http://www.totalwar.org/forum/non-cgi/emoticons/frown.gif

So, the talk of a minimalistic mod might keep the battles down to a more interesting and less repetitive level.

Any ideas? Remove trade from the sea, or even altogether? Then ships would be useful solely for naval assaults and wouldn't necessarily be so important

Might make it very interesting and the game progression more frought with danger....however, one may have to seriously chop building prices and time to produce them though? Anyway, if any of you are interested, start up another thread and we'll get the ball rolling. Otherwise, see ya in a month or two when i may be interested in another late game hour-long turn lol. http://www.totalwar.org/forum/non-cgi/emoticons/smile.gif

The only trouble i envisage with this is trying to balance the factions when trade is no longer an issue...particularly since the Italians, Sicilians, English and Danes rely so heavily on it.

Doc

[Kraellin]

i hate windows.

my harddrive crashed finally. that's meant setting up a new one and windows and all the drivers and all the oddities and tweaks and peculiarities that windows has/does/is. pain in the arse. i'm still working on it, but i've got it so that i can at least get online again. what a pain.

at any rate, doc, yeah, one of my original ideas was to do a minimalist and maximum type money thing. do one with tons of money and one with very little and compare. i think CA took a low end approach, not quite minimalist, but low. and i think they did this because of the ai trade stuff. i think they just ran out of time to tweak it right, or at least better.

hehe, i just auto-calc, doc. saves me chasing all those horses :)

well, i wouldnt want to remove trade completely, but you could cut it down and minimalize it. you'd still have provinces with tradable goods, just not all of them and to compensate for the land-locked provinces not getting as much money, there are things you can do to change that, particularly by upping the trade goods and resources of those internal provinces. this balances the land-locked with the sea provinces and changes the importance of trade. it's just a balance thing if you want to lower the trade stuff. it might also be interesting to only give a few key provinces tradable goods so that their strategic importance is exaggerated, like constantinople, for instance. but then drop all the rest to almost nothing. just scatter a few around, particularly on the mainland so that the islands arent quite so key. lots of ways to tweak it around.

civIII eh? yeah, i sometimes go back to that one for a few days, but i always get pissed off by the way they do a few things and by the lack of what could have been done. i'm afraid good ole sid has fallen behind. what is PTW?

K.

[cugel]

You have my sympathy K: My hard drive crashed last year and melted my processor into the balance. I had to buy a new computer and then the NEW hard drive crashed after 1 month. http://www.totalwar.org/forum/non-cgi/emoticons/frown.gif

Not fun. It took me a week each time to get things set up right again. Kudos all around to Microsoft for making it so easy http://www.totalwar.org/forum/non-cgi/emoticons/angry.gif

BTW, did you ever get your Byzantines to trade effectively? You said they produced a lot of ships, but I bet they didn't leave any in the Marmara for any length of time

[Kraellin]

hehe, aint windoze fun :) lol.

the byzantines kicked butt. yes, they do skip a sea zone from time to time, but remember, that trade routes are optimized by having ships in the zones with the most ports and that any zone you have a ship in is the start of a new trade route, so i suspect that the ai moves its ships around to start NEW trade routes from different places to optimize the income. still, it does seem a bit screwy at times when you cut off the black sea trade from the rest of the world by moving ships out of marmara. overall, though, it was much improved in a number of the faction's shipping and trade. the italians were the ones that did this the most. they'd move out of the zone around venice and go around the italian peninsula to the genoa zone and work from there a lot.

it's far from perfect and i had to tweak a lot of stuff to get it to work as well as i got it, but as least there are a lot more ships on the board, which is essentially all i was trying to do. a routine for training the ai on how to construct good trade routes would have been a nice file to have in order to tweak faction by faction; that would have been really nice. so, without that, the best i can do is to just make the ai REALLY prolific on ship building use large numbers to make up for poor routing. oh, and find out how each faction behavior works and then pick the behaviors that seem to promote good trade routes.

K.

[+DOC+]

As mentioned in another thread, you might want to swap the province characteristics of Genoa and Venice around for that very reason K. The Italians are crap at trading from Venice and good with Genoa.

[Kraellin]

doc,

yeah, i think it may have something to do with the GA game and yeah, i read your other post in the new thread about that. if i ever get this computer back up right i'll take a look at it.

K.

[cugel]

Another idea K might be to reduce support costs for ships, either by 1/2 or to zero. Then producing a lot of them won't cause the faction to have to change to close_to_support_limit (which sometimes seems like a dead-end for the AI - example: the Danes, who once they get stuck in a rut never seem to get un-stuck).

[Turbo]

Quote[/b] (Kraellin @ Jan. 03 2003,14:17)]hehe, i've got ships coming out my arse http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif i think i'm close to figuring out how to get this shipping stuff really cranking. the byzantines are now super powers because of their shipping and trade routes...quite scary.

the almo's also build a lot of ships, but get VERY protective of their home port in morocco. they'll spread out if no one is around, but if one ship moves into their home port, they'll bring almost their entire fleet back to home base to protect it. talk about paranoia.

the english and the danes are now fighting for control of the north sea and english channel, but somehow in the 2nd version of this, the danes arent building hardly any ships, so something got screwed up there.

the germans and french still seem to be shy in building ships for some reason. they do build some now, but gotta look at the stats again.

the swiss and the turks are building ships. lol. not many, but it's a start http://www.totalwar.org/forum/non-cgi/emoticons/smile.gif

my spanish give varying results. they build ok, but their deployment is quite erratic. sometimes they sit and protect the home port and sometimes they use the home port to raid from and sometimes they spread out. not quite sure what's going on there yet, but at least they build.

the aragonese are the real surprise. they are now capturing provinces by having more money from their limited trade route, but they do seem to use trade routes fairly effectively.

the italians are building again, but seem bent on holding the sea zone around genoa, which always starts a war with the burgundians, so hard to tell how effective they might be otherwise.

i turned the egyptians way back at first but now have them back and building ships. they almost always end up fighting the byzantine through ship battles in the egypt sea zone. the egyptians are set to muslim devout this time, instead of muslim expansionist and it seems to make no difference in their aggressive sea action. they ALWAYS try to drive everyone out of that one sea zone.

ok, so here's my reasoning on trade and the aggressive AI. we know that the ai tends to attack when it has the advantage, at least with some of the ai behaviors. so, let's take advantage of that. we cant change it, so let's use it to our advantage. that means, get the ai building a TON of ships and let's go ahead and fight for the seas. this will keep the human on his toes and cut down the human trade routes somewhat. and that's exactly what we need. the human can all too easily run away with the game thru trade routes and money, so let's get the ai VERY aggressive in this regard.

before, i was going for trying to get the ai to be less aggressive on the seas and expanding his trade by keeping the peace and simply spreading out. well, to do that, you almost have to set all the ai factions to a very few ai behavior types, like muslim_peaceful. and that just isnt very good. oddly, so far, orthodox_stagnant is producing the best trade route routines. the byzantine in my game are now out-producing me in ships and trade. it's quite impressive. if i can just get the others to do this, i'll have a great mod, one that folks have been wanting for a while when they noticed that the ai sucks in trading.

i've modded quite a few things, including some ships costs, ship building time, ai preferences, starting locations for factions and a number of other things i'm not going to yap about quite yet (specially since they arent working right yet:).

i've totally abandoned my 10 year rule for this. if they fight on the opening move, so be it. and, i've gotten some interesting conflicts. the aragonese sending a crusade to cordoba was a shocker. the byzantines invading tunisia by sea was interesting. the danes taking over england was cool. the german naval fleet mixing it up with the danish longboats was funny. the papacy invading switzerland was quite unexpected. lol. gotta watch that old pope http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif

there are some observable patterns emerging in the ai behaviors. most behaviors that have 'defensive' in their name seem to do the protect-the-home-base-first routine. the spanish are catholic_defensive_crusader in my mod. other 'defensive' ai's seem to operate similarly. the muslim_devout also seems to act like a defensive behavior. the muslim_peaceful will build and spread but then come home if threatened. in the last game, there were 16 stacks of ships in the straits of gibraltar with a total of 22 ships of different factions, the almo's having spread out to begin with but then brought damn near everything back when 1 or 2 other factions stuck ships in there.

ok, and here is one of the keys to this stuff. we werent really certain how those numbers after each behavior were affecting everything. the way it seems to work is that it's a ratio of any given number to ALL the other numbers of all other things. for instance, if you raise the muslim_peaceful(20) on the ship stuff to muslim_peaceful(400), the code says ok, take ALL the other units of muslim peacful, including the land units and compare these numbers, then build based on the highest values most of the time, but not always.

we also know that the ai alters its behavior based on what's going on in the game. it even states this in the early.txt file. what we dont know is, which of these ai behaviors are switchable. can a muslim_peaceful become catholic_crusader_trader, for instance, or, does the ai simply use the poverty_stricken, desparate_defense, close_to_support_limit, and rebels behaviors as the one's that can be switched to? my recent thinking is that ALL of the behaviors may be being used to switched to. remember, just because it says 'catholic' in the file, doesnt mean the code reads this as catholic only. this is just a variable name, just a tag for reference or for starting out a faction behavior. it doesnt necessarily mean that a catholic faction cant use a muslim behavior. it's just a tag. it doesnt mean anything particularly. you could have named them behavior_a, behavior_b, behavior_c and so forth. you get what i mean here?

now, to test this, i decided to change some of the starting behaviors to include those 4 oddball behaviors, poverty, close_to, rebels, and desparate. the game wouldnt show in the campaign list. so, one or more of those 4 is different in nature than the muslim, catholic, orthodox ones. i suspect it's the rebels, since if that one is set, it wont EVER change. that is also stated in the early.txt file. more testing is needed on all that yet.

now, going back to the numbers at the end of the muslim_peaceful type references. you MUST look at the land based unit numbers. some of those numbers are quite high, 600 and over, so, even if you went from muslim_peaceful(20) to muslim_peaceful(400), you may still not be giving ship building a very high priority relative to peasants(600). you MUST look at ALL the unit values. and then, it's just a matter of balance, getting more ships, but not so much that the ai becomes terribly vulnerable to land attacks.

now, what i dont know here is if the code is taking units AND building preferences into account as one thing. do the building preferences and the unit preferences get taken together or are they separately considered when the ai is spending its money? that one i dont know yet. my earlier guess was that they were taken separately.

ed,
there are a number of things that need to be tested. if yuuki was doing all this he'd very carefully test each individual stat and setting and isolate all this stuff out systematically. i'm terrible about doing this. i tend to do 50 tweaks at once to get an overall change and then end up wondering which tweak did what, but, i get bigger changes and more diverse effects this way. i find the tweak the one stat, test ad nauseum, tweak, test, tweak, test style a bit too mundane for my tastes. i'm too impatient. if you were so inclined, testing some of this stuff one by one would be helpful, like the poverty, rebels, desparate, close_to stuff. which one of those is the one that prevents the game from showing in the campaign list when it's given to a faction as the starting behavior...stuff like that. in other words, answer some of the questions i've listed above. i'm currently bent on getting all the factions positioned on the map better and getting them all to do more shipping.

oh, and that's another thing; because we know that some ai behaviors are quite aggressive, locating and positining your starting ports for the factions is quite critical now. i'm having to re-arrange some of the province ownerships to spread out the starting ports better so that the ai can at least start to produce some ships. this is helping. the last change i did was to remove the danish from denmark. lol. sounds goofy, and is terrbily non-historic, perhaps, but i go first for game play, so, if the english end up starting in georgia, so be it http://www.totalwar.org/forum/non-cgi/emoticons/wink.gif

cugel,
i think i answered all your points in the stuff above, so nothing really specific to direct at you personally here, other than keep the faith and keep on truckin http://www.totalwar.org/forum/non-cgi/emoticons/smile.gif

K.
I think we need to face facts that without another patch (fat chance) that fixes the AI naval trade, this facet of MTW is hopeless. If the AI would at least make a modest attempt to try to trade, building more ships might make things better. Unfortunately, the AI treats the ships as combat units. Ports and shipyards when they are built, are built in provinces without consideration of trade resources. Finding a dozen ships in one sea area and none in others is common AI "tactics".

I bought this game looking for a challenge, not looking for a walkover or blowout. Based on this, I went "heretic" and modded this part of the game.

My solution was to eliminate naval trade altogether but retain the naval combat/transport features of the ships. This was easy to do. Income now comes exclusively from farmland or base province income. This had some interesting consequences in the game as it forces serious limitations on the size and quality of the player forces. Holding provinces and keeping them happy becomes a critical part of revenue along with decent govenors. Rebellions or civil wars are budget breakers. Size of empire equates with revenue, allowing for the 'Hojo' like enemies I liked so much from STW.

[Turbo]

Quote[/b] (Edratman @ Dec. 31 2002,06:52)]I think I've got it An example of turning a finely tuned Engineering mind (sic) to solving a problem.

How about delaying the merchant1 build? Make the prerequisite Castle9, or what ever you choose.

Think about it. We can't get the AI to trade, so limit our trade. This would level the field. The implications are enormous. Probably can't afford a world wide network of agents, or ships, so you'll play blind. No more bribing of rebel provinces whenever you want. Spend every florin like it was real money.

I'll wager the AI is good enough under that condition. I know my playing style would go down the toilet.

Got to run.
I came to the same conclusion however, I eliminated naval trade. Income comes from province income and farmland just like STW. This puts the human on the same playing field as the AI and changes the nature of the game. I find myself struggling with income, and constantly checking govenors and happiness levels. I never paid any attention with this before because with the one-sided naval trade I was swimming in florins. Rebellions and civil wars are budget busters now, it is easy to go into the red. And, playing Poland is close to hopeless....

[Kraellin]

turbo,

cool beans.

fraid i'm somewhat temporarily out of the modding business for a while. since my harddrive crashed i havent put mtw back on yet. still ironing out a few bugs and i may have some hardware going bad; i keep getting little oddities that shldnt be there and some interesting clicks and pops now and again, so it may be a while.

at any rate, yeah, the minimal thing is fine for a harder game...almost always is, but i'll still be trying to work out something with the ship stuff. i hate just cutting off parts of a game to make it better, though, sometimes this does seem to be the only way.

K.