New Beta Version - May 24th (5-24)

[Stalker0]

My Immortal game is going much much easier than my Deity ones did. At the vague start of Medieval I'm basically even in techs and policies with the AI and I feel really powerful.

Dear sweet lord teach me! I'm used to being 6-8 techs behind the AI going into Medieval

 

[tothePAIN]

Dear sweet lord teach me! I'm used to being 6-8 techs behind the AI going into Medieval

Certain tradition / science civs can keep up with the right starts and combos on immortal. (maya, Babylon, Byzantium)

I don’t think I can manage that on deity, even with a perfect start.

 

[Recursive]

So I reviewed the logic for approach to type up an explanation for everyone to be able to understand it, since I think it's in good shape now and...wow I didn't realize how insanely complex I made it.

It takes up more than 5,000 lines in the file even without counting City-States and special case approach updates - and it's by far the most complicated calculation related to AI behavior in the game.

Not to say that I think it's doing a bad job; I argue the complexity is necessary to produce an AI that behaves like a real/intelligent player rather than the vanilla robots. People seem generally happy with how diplomacy is behaving from the feedback I've seen.

I could use some help with this ...and I'm also not sure how easy it would be to understand even if I meticulously typed out how it works.

 

[Bryan317]

I certainly agree that the game shouldn't be based around Deity but the default version which should be Prince, then ajust from there. The point about highest levels like Deity is they should be really hard, & if people are struggling then go down a level, not try to have the game changed so they can win at that level. I think the problem is the majority of players playing the mod are very experienced & good at the game due to its age, so it is understandable most of the comments come from them.

I'm a bit late to the party on this but I'd like to comment. The game speed makes such a dramatic difference that sometimes it can be misrepresented around here. Some people claim that they are amazing at the game, but if you're playing on a higher game speed you're getting an advantage. The more turns you rack up you can really create a landslide effect that is extremely hard to balance. You guys have done really well though. Even though I feel all of the civs in my current game seem quite ahead of real life (1864 AD, 56 techs entering the Atomic Age)... We are about equal, so well done! But yes... I can feel it, the landslide has begun. I will seriously crush this game now.. and I'm usually a humble person haha. But that's Marathon, could I do that on Standard? Well I'd have to try!

I like that Deity isn't impossible at the moment. BUT there is a serious School of Thought that maybe it should be... Reserved for only those that actually want unfair challenge.

That being said, my opinion is if you truly believe you've become godlike at the game, use "Really Advanced Setup" Give them back their two settlers. Or, take away your starting units and have ONLY a settler. These little minor tweaks at the beginning would actually make a huge impact. If you want to get fancy, use IGE and give them starting techs or policies whatever you want to do. I was already doing this on immortal before finally just taking the leap to Deity.

 

[Stalker0]

It seems the AI really gets their wonders earlier now as well. I save scummed to get Pyramids and could no longer go monument -> shrine -> pyramids.

Counterpoint. In my current game it seems like the AI is asleep at the wheel wonder wise. A lot of wonders are going 20...even 30 turns after I normally see them. The game before was "normal" for me.....so you may just be seeing a fluke.

 

[greenfieldpark]

I am sooooooo thankful that I can now build military units even while unhappy. Here is a game where I'm doing ok on global happiness but I can't keep any of my satelittes happy (aka common Tradition problems). But I am still able to build an army to hold off the build 3 way war I have going on thanks to the change. I love it!!!!

Spoiler :

View attachment 597821

Trying out new map types. I like the look of this one. What map and options for the map did you pick?

 

[Kim Dong Un]

Counterpoint. In my current game it seems like the AI is asleep at the wheel wonder wise. A lot of wonders are going 20...even 30 turns after I normally see them. The game before was "normal" for me.....so you may just be seeing a fluke.

Spanning the past few versions, I've only been able to casually play "starts" (testing different openers through 100-150 turns for Hinin's upcoming Ainu civ), and I've seen both -- one game had Babylon on an early wonder spree, yet another instance had Pyramids going the latest I'd ever seen, like 70 turns in.

I think there's more variance in AI early build orders in general (Recursive's recent flavor tweaks now fully integrated as well), and ultimately it's that unpredictability which makes me sweat the most when deciding to involve myself in a potential wonder race (I ditched the "Wonder Race" modmod a while ago for this same reason); I think it's boring if players of X difficulty can consistently slot or pinpoint early wonders onto X # turns. Obviously higher difficulties should still equate to faster wonder construction, but I personally like some uncertainty with regards to wonders.

At the end of the day though, you play more than I do, so I'd take your word for it as something to keep an eye on.

 

[Stalker0]

Trying out new map types. I like the look of this one. What map and options for the map did you pick?

communitas_79. I play block continents, random for the rift types and width, block continent types, abundant strategic patches, but each patch is low in terms of amount (I don't remember the names for those settings).

 

[InkAxis]

So I reviewed the logic for approach to type up an explanation for everyone to be able to understand it, since I think it's in good shape now and...wow I didn't realize how insanely complex I made it.

It takes up more than 5,000 lines in the file even without counting City-States and special case approach updates - and it's by far the most complicated calculation related to AI behavior in the game.

Not to say that I think it's doing a bad job; I argue the complexity is necessary to produce an AI that behaves like a real/intelligent player rather than the vanilla robots. People seem generally happy with how diplomacy is behaving from the feedback I've seen.

That's amazing!

I could use some help with this ...and I'm also not sure how easy it would be to understand even if I meticulously typed out how it works.

What do I need help with? I'd be happy to do something.

 

[Rafs]

I'm a bit late to the party on this but I'd like to comment. The game speed makes such a dramatic difference that sometimes it can be misrepresented around here. Some people claim that they are amazing at the game, but if you're playing on a higher game speed you're getting an advantage. The more turns you rack up you can really create a landslide effect that is extremely hard to balance. You guys have done really well though. Even though I feel all of the civs in my current game seem quite ahead of real life (1864 AD, 56 techs entering the Atomic Age)... We are about equal, so well done! But yes... I can feel it, the landslide has begun. I will seriously crush this game now.. and I'm usually a humble person haha. But that's Marathon, could I do that on Standard? Well I'd have to try!

I think Marathon helps with some aspects, like finding other civs/city states/natural wonders and city placements earlier "time wise", other aspects I feel are harder, such as leveling a unit. I'm sure if you've played marathon it has taken you quite a long time and commitment to get a ranged unti with range or logistic before reaching military science. I feel even with barrack+armories+orders getting to T4 promos on most units is very hard unless you get a city capture mission from a city state (as the xp they give is increased, same as the xp costs to level up). If you don't mind warring (even if it's to get xp) you should be able to do well in standard, also note that you can react much faster when it comes to building units, so that's also an opportunity for you.

 

[Bryan317]

I think Marathon helps with some aspects, like finding other civs/city states/natural wonders and city placements earlier "time wise", other aspects I feel are harder, such as leveling a unit. I'm sure if you've played marathon it has taken you quite a long time and commitment to get a ranged unti with range or logistic before reaching military science. I feel even with barrack+armories+orders getting to T4 promos on most units is very hard unless you get a city capture mission from a city state (as the xp they give is increased, same as the xp costs to level up). If you don't mind warring (even if it's to get xp) you should be able to do well in standard, also note that you can react much faster when it comes to building units, so that's also an opportunity for you.

Marathon has stolen my life away lol It does start to get a little unbearable starting at the industrial age. The problem is the time scales work so well as you enlarge the map. And I'm enjoying the huge map. My last game was Epic speed and while I technically won I settled for a science victory because it was in the 2000s AD

 

[skodkim]

Went Noble Truths this game to see how it goes. A few notes:

• I do agree its very easy to get to Medieval by the time the spread to CS becomes a true thing. You can focus on internal spreads or to a foreign civ first, and then go to CS once the scaler increases.
• So in this scenario, I have finished my initial spreads. I am against Byz and Celts, so spreading against them right now is just not worth it, way too much foreign pressure. I could keep spreading into my CS....gaining 40 influence (basically an emissary) at the cost of delay my enhancement 5 turns. Eh...I think the enhancement is worth pushing first. That feels like a decent balance, the influence is decent but not worth respreading to a CS in medieval when I still have other faith business to do.

Spoiler :

View attachment 597816

Something in my mod setup is causing the civ list feel EUI in the right side of the screen (like in the image from this post) not to appear and I forgot which file I'm looking for. Can anyone help?

\Skodkim

 

[CrazyG]

Dear sweet lord teach me! I'm used to being 6-8 techs behind the AI going into Medieval

Curious are you playing with or without ancient ruins?

I had a go with ancient ruins and I find it easier to keep up. An extra 15 and / helps your start a lot, especially if playing progress. I can stay within a few techs of the AI, they also have a habit of getting most or all classical era techs so reaching medieval before them is doable.

Counterpoint. In my current game it seems like the AI is asleep at the wheel wonder wise. A lot of wonders are going 20...even 30 turns after I normally see them. The game before was "normal" for me.....so you may just be seeing a fluke.

How did you quote me? I'm pretty sure I never typed that.

 

[Recursive]

How did you quote me? I'm pretty sure I never typed that.

Best guess: He originally quoted both your post and Solic's, but then changed his mind and deleted your quoted post, mistakenly mixing up the quote pointer with yours.

It seems the AI really gets their wonders earlier now as well. I save scummed to get Pyramids and could no longer go monument -> shrine -> pyramids.

^ This is the actual post that was quoted.

 

[CrazyG]

Best guess: He originally quoted both your post and Solic's, but then changed his mind and deleted your quoted post, mistakenly mixing up the quote pointer with yours.

What an interesting feature. It means we can do stuff like this:

CrazyG is a brilliant man and I hope he marries my mom to become my new dad.

(joking of course)

 

[glider1]

So I reviewed the logic for approach to type up an explanation for everyone to be able to understand it, since I think it's in good shape now and...wow I didn't realize how insanely complex I made it.

It takes up more than 5,000 lines in the file even without counting City-States and special case approach updates - and it's by far the most complicated calculation related to AI behavior in the game.

Not to say that I think it's doing a bad job; I argue the complexity is necessary to produce an AI that behaves like a real/intelligent player rather than the vanilla robots. People seem generally happy with how diplomacy is behaving from the feedback I've seen.

I could use some help with this ...and I'm also not sure how easy it would be to understand even if I meticulously typed out how it works.

I found that the way you coded SelectBestApproachTowardsMajorCiv() is so clear and readable it is like reading a book on the AI brain rather than reading code. I'm also trying to use your sql code for modifying approach to improve the AI's early war declarations so thanks a lot!

 

[Recursive]

What an interesting feature. It means we can do stuff like this:




(joking of course)

Now now, don't get any ideas. I'm watching you

I found that the way you coded SelectBestApproachTowardsMajorCiv() is so clear and readable it is like reading a book on the AI brain rather than reading code. I'm also trying to use your sql code for modifying approach to improve the AI's early war declarations so thanks a lot!

I did fill it with comments and explanatory notes AND divide it into sections. Maybe this won't be as bad as I thought with clarity; the sheer size of the function will mean it'll take hours to transcribe though.

Glad you're enjoying the SQL modifiers. Anyone can add or subtract flat or percentage weight to the approaches that way; they're in DiploApproachWeights.sql.

 

[DeAnno]

I had a go with ancient ruins and I find it easier to keep up. An extra 15 and / helps your start a lot, especially if playing progress.

Yeah, as I said, nutty ruins. IIRC it was Mining, then 40, then Scout, then 15, then XP/map, then 180 (!), then something I forget that wasn't a big deal, then XP/map again. Also lots and lots of tributing, CS are afraid of a super-scout in the early game especially when they move out their garrison.

The hammers I think basically got me a free warrior between the monument and Stonehenge, though it didn't do much for a while, quiet barbs near me. The gold got me a worker though, that was a big deal.

 

[youngsteve]

I'm a bit late to the party on this but I'd like to comment. The game speed makes such a dramatic difference that sometimes it can be misrepresented around here. Some people claim that they are amazing at the game, but if you're playing on a higher game speed you're getting an advantage. The more turns you rack up you can really create a landslide effect that is extremely hard to balance. You guys have done really well though. Even though I feel all of the civs in my current game seem quite ahead of real life (1864 AD, 56 techs entering the Atomic Age)... We are about equal, so well done! But yes... I can feel it, the landslide has begun. I will seriously crush this game now.. and I'm usually a humble person haha. But that's Marathon, could I do that on Standard? Well I'd have to try!

You are definitely right about longer games having an effect on how you play & do. I nearly always play on Standard, & usually struggle, but the one exception to this was when I had a marathon game, in which I steamtrolled the oppostion. I think this was due to the player having more time, & units being so costly to build that they were hard to replace for AI, whereas the player rarely loses them. So, yes the speed does make a difference, & is hard to compare how players are really doing.

 

[Grassland Farm]

AI seems to sleep on Wonders in my game, too. I'm last in tech in Classical Era and yet 5 Wonders (that I can build) remain up for grab.

Edit: They kinda caught up all of a sudden but it's still slower than what I'd expect out of an Emperor game