Version 1.01 discussion thread

You probably just need to add BetterBTSAI.h to both CvPlayerAI.obj lists, alternatively you could try that new Makefile.

I'm not sure. I have never had to use a makefile. I use Visual Studio 2003. I tried adding BetterBTSAI.h and BetterBTSAI.cpp to the #include list at the top of the CvPlayer.cpp and CvPlayer.h files and it made no difference.
 
Well it's CvPlayerAI.cpp, and there should already be an #include "BetterBTSAI.h" - line there. Did you add that header file to your project yet?
 
Well it's CvPlayerAI.cpp, and there should already be an #include "BetterBTSAI.h" - line there. Did you add that header file to your project yet?

Ok you were right. It already had the #include "BetterBTSAI.h"—I had just missed it. I had to figure out how to include BetterBTSAI.h and BetterBTSAI.cpp in my project since I had not had to do this before. Once I added these items to my project, I was able to compile the DLL. Thanks for your help! :goodjob:
 
Hey guys,
1st of all i want to say Good Job.
It is very interesting to see the good ingame results after trying this mod.
The AI now has a better plan for victory and thats very nice :goodjob: .

Now to the sad things.
I usually play on Deity to have most challenge.
Since barbarian spawning highly depends on the number of cities founded, on this Diff the Barbs appear very early.
However, and i think this is intended, u can deal with them in the beginning (before they have Axeman) with Warriors on Hills with Forest outside your borders.
They see u, they attack u and its fine until they come with multiple units and in that case its the players fault not to move another warrior there.
So far so good, now this mod comes into play.
Unfortunately, the Barb AI seems to be rewritten too and now that early defense concept does not work anymore.
No Barb-Warrior or Archer will attack one of the warriors on those hills.
They will move around and when they are allowed to enter cultural borders, all of them will come. And even worse, they dont go to the tile next to the city, maybe pillage and then suicide, but they will pillage every possible tile.
The only chance to deal with the barbs atm is to research BW as soon as possible and then to build Axemen. However, it is very common, that there is no Bronze nearby where u could build a 2nd city.
And even if u are lucky and get that bronze, it is still annyoing. The Archers won't attack of course, but u might have Hills in your CB and so you cannot attack them, because odds are not good enough.
You are financial and have that juicy flood plains start? Forget it! All cottages will be plundered soon. And forget it once again, because flood plains tend to be far away from Bronze.
In Summary, at least Deity (can't say about the other diffs) is totally imba. Only chance for survival is to have bronze nearby and even then u have a hard time if u dont have a coastal start with good spots for fogbusting.
This reminds me a bit on the time, the mod FFH increased barb logic and suddenly it was very difficult in the beginning. However, in those mod warriors are stronger and very cheap, much cheaper than in Vanilla.
In my opinion this game is not designed for an ubernation of pillagers in the beginning. The barbs should be a nuisance for careless players. But they should not force the player to have only one possible research path in the beginning (and to be lucky in the first crucial situations)

So Jdog, please continue your great work but make the barbs stupid again.
Atm, it is nearly impossible to use this great mod with barbarians on.
 
I disabled the normal AI odds-checking behavior for barbarians in MongooseMod for precisely this reason. Except for the Sandworm and sea monsters, the barbs behave suicidally and will always attack anything they can as soon as they can.

However this constitutes a gameplay change and thus is not likely to be something JDog would approve for inclusion in BBAI/UP. Try my mod is I guess the best advice I can give ya. ;)
 
If the barbs were stupid in stock BTS and are now smarter in Better AI, changing them back would not be a gameplay change at all, it would be reverting one.
Even without barbs/with stupid barbs, it should simply not be possible to beat the Better AI on Diety though.. or if it is, please tell me how and we will try to fix that ;)
 
If the barbs were stupid in stock BTS and are now smarter in Better AI, changing them back would not be a gameplay change at all, it would be reverting one.

Most of the improvements in BBAI don't apply to barbs anyway, since barbs do not plan wars and move large armies around like AIs do. And certain other improvements that did affect barbs on a tactical level do still apply with my change.

The change is based on a fundamental belief on my part that barbs should not be MORE powerful/damaging/threatening to a player's survival than AIs you're at war with are, in general. They are supposed to be a fantastic annoyance, a real thorn in one's llama, but when you start dealing with barb densities like Deity has (let alone the overloaded values on DarkMongoose + the Raging Barbs option heh) they are actually way more effective than the comps, even over the course of a whole game, at deterring the human from winning imo. That isn't how it should be. And I can now say from experience playing my mod that even with the change, barbs are still plenty deadly... but they're also manageable, as opposed to unmanageable... pretty much exactly b/c of the situation Schlalex described.

But it's not really a reversion, since vanilla never had barbs behaving this way so BBAI never got to affect it. :p

Even without barbs/with stupid barbs, it should simply not be possible to beat the Better AI on Diety though.. or if it is, please tell me how and we will try to fix that ;)

This part I completely agree with. :)
 
If some element in the game proves to be hard to counter for the human player, then a BetterAI mod should never immediately consider making the game element behave more stupid. It's contrary to its goal.

Improvements in the AI can however lead to undesired irregularities in difficulty during the game. It could for instance lead to an AI that at some difficulty level would completely owns the human player up to the medieval era while it couldn't keep up with the human player if the human player would survive until after the medieval era. If such an imbalance in opposition occurs, then rebalancing of the level of opposition throughout the game is in order.

In this case, the barbarians have become tougher because they don't behave suicidally stupid anymore. I would say that the behaviour in itself is an improvement as I find it far more interesting to fight barbarians now than before (playing at immortal level usually, huge world, 11 AI's, aggressive AI, epic game speed, big and small map type). It will of course cost more units to repel them and you will need at least archers to do it. The biggest improvement in my view is that beating the barbs will require an active defence, you can't sit back on hilltops or fortified cities while the barbs suicide their troops.

However, if this is the most challenging part of the game and the game becomes noticeably easier in other periods of the game, then maybe some imbalance has been created. If many players feel that way, then I'd say that maybe the barbarian strength could be toned down a little. I think this mod should never take the path of making the barbarians stupid. But it could turn down the number of barbarians a little or give the humans a small combat bonus against them (similar to the AI combat bonus vs barbs).

Personally, I haven't encountered many situations where the barbs dominated my lands. They sometimes manage to pillage a tile, but that's about all. You just need a more hands on approach to beat them instead of placing some units on hilltops around your borders. I typically need a few units on the non-fogbusted borders. The AI is normally very actively moving units through the fogged areas at the highest difficulty levels, seriously reducing the spawning rate of barbarians. You need some units in the border areas where the AI isn't doing this. And these units aren't there to defend hills, but to counter-attack barbs who try to pillage your lands. I actually prefer chariots the most if I can have them as they're able to counter-attack in a huge area if you have some roads. But I often need to use archers in the earlier period of the game.
 
I am not an expert on how vanilla plays at this point, so you may very well be right Roland.

However in addition to an extremely high barb density setting (much higher than Deity) on the custom difficulty levels, my mod makes some other changes to barb rules, including the ability for them to spawn in any tile that is fogged to humans and AIs regardless of proximity to other units including their own (which was necessary for high density values to work properly). This, combined with my belief that the barbs should be a more savage, primitive, brainless, but nevertheless deadly annoyance, makes the suicidal behavior change a good idea in my mod, in my opinion.

I also personally prefer that the need for an "active defense" be reserved for wars against humans or AIs, but freely acknowledge this is probably a matter of preference.

So I am not really pushing for this change to be made in vanilla, and as I said JDog would never approve it anyway. :p But the dood asked about such a change, which does exist in my mod, so I wanted to at least give him the option.
 
I am not an expert on how vanilla plays at this point, so you may very well be right Roland.

However in addition to an extremely high barb density setting (much higher than Deity) on the custom difficulty levels, my mod makes some other changes to barb rules, including the ability for them to spawn in any tile that is fogged to humans and AIs regardless of proximity to other units including their own (which was necessary for high density values to work properly). This, combined with my belief that the barbs should be a more savage, primitive, brainless, but nevertheless deadly annoyance, makes the suicidal behavior change a good idea in my mod, in my opinion.

I checked out your mod a few days ago. It looks very interesting. Although, it's hard to get an idea of the entire package without downloading it as your description hasn't finished yet. Descriptions aren't the most interesting part of mod making of course.

I also personally prefer that the need for an "active defense" be reserved for wars against humans or AIs, but freely acknowledge this is probably a matter of preference.

Agreed on the 'matter of preference'. I view them as less settled down people who live by plunder but still value their own lives.

So I am not really pushing for this change to be made in vanilla, and as I said JDog would never approve it anyway. :p But the dood asked about such a change, which does exist in my mod, so I wanted to at least give him the option.

Never a bad moment to promote your mod. ;)

Good luck with getting a following (it will be a bit harder with a 5 year old game and it successor just around the corner). I will be watching your mod.
 
Hi,

Just played a full game with your mod, after all these comments on immortal difficulty I feel semi silly saying I played on monarchy. My biggest beef with the diifficulty related ai cheats isnt, production or expansion related but tech related where an equal civ much surpasses the human player in tech- and tech flies by so fast u can barely enjoy the time periods even on no tech brokering, and especially on no tech trading game options. I dominated monarchy at a good tech pace and will try a game in emperor next and see how bad the tech cheats are-just dont like techs popping every 1-2 turns by the ai. Unless its an old tech ofc.

My suggestions:

- Edit some of the civ personalities so that none favor a poor military at any time, especially if their neighbors are built up with troops.

- In my game I had the 2nd largest empire early with the 8th largest military for like the first 70 turns (normal mode). Eventually became largest empire still early. I shouldve been bullied more or declared on-I only had about 4 threats and turned down half of them

- +3/+4 free hammers per city per turn for the ai on monarchy/emperor difficulty settings (I wont suggest into the higher/ridiculous game difficulties), this bonus would not apply to wonders or spaceship parts or be inherited by conquest by the human but for more/easier units/buildings for the ai. Can better simulate smart slaving/chopping by the human.

- More aggressive ai use of slavery while in peace/war (with lower unhappiness than for a human) or a trigger that does about the same. All ai cities should build early granaries. Maybe a high priority on a granary and when ai gets a granary, they get a free worker (so less non military time building workers). My land gets improved way faster than an ai does.

- More aggressive use/adopting of nationalism/drafting- or if they wont use drafting, implement a trigger which does basically the same thing for them. When I get nationalism I draft like crazy and its game over for the ai usually, since I can conquer one neighbor fairly easily then build up to repeat, usually without interference. I build mass cats at nationalism and use drafted units to mop up the easy left overs.

- Ai still doesnt seem to reinforce cities well that are under seige by the human. Ill take a city and move on and see the next city has 4 cavlary type units that couldve helped the last city.

- Slightly less ai cheating gold but reduce its unit upkeep/costs to almost nothing. Hopefully ai will make larger unit reserves like a human.

- Ai seems to send enough to take a city now (much improved) but AI still needs to bring maybe 1.5 times as much to take a city (because of last minute slaving and bringing reinforcments from other cities, while they bombard) I did lose two cities because of not seeing an army coming (army came through another country's open borders to attack me, very impressed)

Summary: AI needs to be more beefy militarily, send maybe 1.5 times as much to take a city instead of just enough, AI needs to slave/draft more so it will have more developed cities/much larger standing army. AI prepares and sends decent stack (should be more) but then doesnt seem keep up a decent trickle of new military (increase ease/frequency of ai unit production). No to high cheating tech research but yes to high production/unit stacks/city development by the ai-even if it means a bit of cheating imo.

Maybe also set higher priority by the comp of ganging up on the top points player (human or comp). In my game a faraway civ declared war on me-as points leader- and it didnt really do anything until eventually a neighbor also declared war on me and got owned by me solo. A neighbor, 2nd in points, couldve declared and made it much more difficult for me to actually attack (I did have great relations with them tho). After I beat that one civ solo, 2nd points leader asked to become my vassal lol (maybe 25 turns after I got nationalism and started mass drafting).

I'd like to feel "in danger" of dying or losing several cities past maybe the first 80-90 turns on normal game speed. I'm gonna try emperor difficulty next, I like the production boost for ai but I will have to see how much I hate the ai tech cheats.

Thanks for the mod! Very good work!

-Charles
 
Most of the improvements in BBAI don't apply to barbs anyway, since barbs do not plan wars and move large armies around like AIs do. And certain other improvements that did affect barbs on a tactical level do still apply with my change.

The change is based on a fundamental belief on my part that barbs should not be MORE powerful/damaging/threatening to a player's survival than AIs you're at war with are, in general. They are supposed to be a fantastic annoyance, a real thorn in one's llama, but when you start dealing with barb densities like Deity has (let alone the overloaded values on DarkMongoose + the Raging Barbs option heh) they are actually way more effective than the comps, even over the course of a whole game, at deterring the human from winning imo. That isn't how it should be. And I can now say from experience playing my mod that even with the change, barbs are still plenty deadly... but they're also manageable, as opposed to unmanageable... pretty much exactly b/c of the situation Schlalex described.

But it's not really a reversion, since vanilla never had barbs behaving this way so BBAI never got to affect it. :p



This part I completely agree with. :)

The current release is a material gameplay change on barbs.

Galleys will blockade, consistently, but also seem to have the rules dictating their entry into culture modified. When you play on a high difficulty with the inexplicable patch 3.17 quad galley spawn rate, all I can say is that you better not try to culture-block these things, and you better hope you either a) CAN spawn bust them away entirely or b) have no seafood to protect.

High level AI pillage party rushes are also just about impossible to block, although BBAI was designed to make an AI that is challenging w/o the bonuses so that's to be expected.

However, the changes to the AI have resulted in some gameplay changes, and the line is pretty gray. I'd not worry about that though, as there is enough broken in stock BTS that holding gameplay sacred isn't too appealing.
 
Hi Charles,

the discussion wether the AI should play to win, should "roleplay" or even should try to actively try to stop the player from winning has been going on for some time. While there are pro´s and con´s for both perspectives (does it feel right if Ghandi attacks when strong, should your all too powerful neighbour with good relations let you get a spaceship victory?), jdog has implemented very advanced (for the CIV AI) new tactics to actually make the AI try and win a game - before it was just sort of "stumbling" onto a victory, instead of actively pursuing it.

I guess tweaking things like military attack logistics is not as straightforward as one would think - and increasing one aspect could lead to really deficient behaviour on another end. The AI is - compared to a human player - still very much braindead, and it needs algorithms that work for all situations. Clearly they must be a compromise and not optimal for many strategic settings.

I agree with your feedback, and I think comments like yours are very valuable to jdog, especially coming from someone "new" to the mod with views that are still untarnished and able to see the difference to the unmodded game.

Have fun, Jan
 
I kinda take a highlander view to single player games....there can be only one...winner. So I think it would be ok to give negative relations to the human player who is pulling too far ahead or even the top computer player-as long as its noticeable winning advantage over the other civs. The computer already makes deals way less favorable when your winning, this would be military/strategic version of that.
 
I love this mod, and it makes the AI smarter in a lot of ways, but the AI is still really dumb when it comes to colonies. For example, in my current game, Mao started on a small, icy island all by himself, so he was already struggling to keep up, but then he split his already small empire in two by making a colony, and he effectively crippled himself. Colonies are hardly ever useful even for a human player, so I think the AI should just be prevented from making them.
 
Just finished my first game with BBAI (noble, normal size, normal speed, BBAI 1.01). Generally was pretty good - the AI started quicker and seemed to maintain (although I came out on top in the end). A couple of oddities:

1) At one point, Ragnar launched a surprise attack on me from the sea. Could have been effective (the target city had very weak defence, and my good units were at the far side of my territory) but the first wave consisted of only 3 cavalry (died, partly because they attacked from the sea) and four trebs (landed next to the city a turn later). The main invasion appeared about ten turns later, by which time I had riflemen and (I think) cannon in the area. Any idea why this might be?

2) Later, while presumably pursuing a space victory, Hannibal tech'd Fibreoptics before Superconductors, letting me get ahead of him in building due to Laboratories. That meant to happen?

I thought I had a third example: Suveryman (sp?) lead for most of the game, but fell badly behind in the industrial era. His tech went to pieces (espionage showed him researching "Plastics (148)"). Of course, I realise soon enough that he'd actually given up on tech to focus on culture, and it turns out that if I handed won the Space Race he'd have won Cultural a few turns later. Never seen the AI manage that before.

Thanks all for a good mod, it's really nice addition to the game.
 
the discussion wether the AI should play to win, should "roleplay" or even should try to actively try to stop the player from winning has been going on for some time.
Something also to consider:

Are the barbarians another AI player (who wants to win) or are the barbarians an enviromental obstacle (that's solely there to force players to defend cities)?

Personally, I don't see barbarians as a player, but only as environment, like terrain. As such, I prefer the "mindless" barbarians, who rather harass the player than try to actually defeat him.

The reason I catagorise barbarians as environment is that Better AI is essentially there to make the AI players more interesting, so they make smarter decisions, more like a human player in multiplayer. Or in other words: In a human-only game in multiplayer, Better AI shouldn't have an impact, since there's no "AI player" so to speak.

Smarter barbarians, however, would change an all-human multiplayer, hence it's changing the environment (instead of the AI players' abilities).

Cheers, LT.
 
Well, Lord Tirian, based in various stuff in the game structure in general ( like , for example the barb handling in WB saves ), I would definitely say that the original coders did not thinked of the barbs as a AI player. And the fact that animals are barbs tip me to the side that they were meant to be simply a enviromental obstacle ...
 
Don't even get me started on the animals, that's outside the scope of BBAI.

But the problem with barbs is that they cost the barb faction 0 hammers and respawn indefinitely until the land is busted. Having to deal with a 6 archer choke around 2000 BC isn't fun and it really takes AWAY from the focus of this mod, which is the AI. The AI, which gets bonuses and gobs of units to deal with this, is comparably unaffected (though not entirely).

At any rate, IMO this mod should not be adding "difficulty" by simply allowing the AI to abuse its bonuses more effectively. The most extreme example of AI bonus abuse would be to have the deity AI simply declare on you instantly and use all but 1 or 2 of its archers to attack your starting city, literally killing you before you could do anything. Yes, that is optimal play for the deity AI because it is a VERY cost effective 3rd city and helps its winning chances immensely, but it isn't good game design for a human player.

Currently barbs (especially super galley blockades before MC is possible and the quad+ archer chokepillage fests) in BBAI are a less extreme example of such an AI abusing its bonuses. One could argue that BBAI should make the AI do precisely things such as this and then simply tweak the difficulties, but the implied difficulty of balancing that (and making the AI actually effective with such tactics) seems a high hurdle to me.
 
True enough. Barbs do not play by the AI rules and they don't seem to be thought by the designer as more than a glorified and animated version of a terrain piss off.... so probably is better to leave them alone, maybe even save this for the Better BtS Barbs mod :D I'm not oposed on doing this change, but IMHO it falls out of the scope of the mod ( wait, better not say this... last time I said that something did not belonged here, I had to say the same thing 20 times in the same thread and actually inflamed a thread :p )
 
Back
Top Bottom