AND SVN Build Thread

[45°38'N-13°47'E]

Afforess, I've seen you've modified erainfos and now first eras are very short. Is this intentional? I mean, when I first balanced the game I've made it so every era has roughly the same number of turns. For Blitz speed (600 turns), this was about 80 turns. Industrial/Modern/Transhuman has some more turns because they have more techs. But now Ancient/Classical pass both in about 50 turns each, possibly less if you beeline for a medieval tech. It's ok to me if we want to do it that way because there's less to do on ancient/classical eras but then I'll rework years per turn lenght. I don't want years to fit perfectly real human history, but I don't like having tanks in year 1300 AD too, beside the fact that if you play mastery you won't have anything to do for half of the game having already researched every tech.

Edit: I know you're trying to nerf tech diffusion again but again we do have some civs researching Screw Propeller (Industrial Era), while some others don't even hava Papacy (Medieval). More than 2 Eras of distance is way too much IMO. That makes it way too easy to overcome those small civs.

 

[Afforess]

Afforess, I've seen you've modified erainfos and now first eras are very short. Is this intentional? I mean, when I first balanced the game I've made it so every era has roughly the same number of turns. For Blitz speed (600 turns), this was about 80 turns. Industrial/Modern/Transhuman has some more turns because they have more techs. But now Ancient/Classical pass both in about 50 turns each, possibly less if you beeline for a medieval tech. It's ok to me if we want to do it that way because there's less to do on ancient/classical eras but then I'll rework years per turn lenght. I don't want years to fit perfectly real human history, but I don't like having tanks in year 1300 AD too, beside the fact that if you play mastery you won't have anything to do for half of the game having already researched every tech.

Edit: I know you're trying to nerf tech diffusion again but again we do have some civs researching Screw Propeller (Industrial Era), while some others don't even hava Papacy (Medieval). More than 2 Eras of distance is way too much IMO. That makes it way too easy to overcome those small civs.

What happened is that with the new government civics, expansion is a bit harder and slower, so I had to decrease tech costs across the board. Decreasing tech costs also meant decreasing tech diffusion too.

It seemed to be about right when I played with it over the weekend, but I am sure more balancing will be needed.

I tested on Epic not Blitz. Gamespeeds may need individual adjustments too.

 

[45°38'N-13°47'E]

What happened is that with the new government civics, expansion is a bit harder and slower, so I had to decrease tech costs across the board. Decreasing tech costs also meant decreasing tech diffusion too.

It seemed to be about right when I played with it over the weekend, but I am sure more balancing will be needed.

I tested on Epic not Blitz. Gamespeeds may need individual adjustments too.

I've done some test on Epic and it's way too fast too. Ancient/Classical might be ok but Medieval Era is really too fast. On blitz speed, each era should last about 80 turns so on Epic it should take about 240-250 turns (250*7 eras = 1750 turns, Epic has 1800 turns). It's ok to have Ancient/Classical in about 450-500 turns but then Medieval Era takes about 120 turns which is half than what it should be. Civs enter Renaissance between turns 620 and 650 while they should not before turn 710. Same applies for following eras. Civs are researching slower because there's a drop in research rate due to lack of gold, possibly because of civics, but it's definetely too much reducing both tech diffusion AND tech costs. I would suggest increasing again tech costs from medieval era onwards; maybe smaller values than it was before but definetely higher than they are now.

Edit: also industrial era is starting about turn 840, well before reaching half of total turns, while it should start not before turn 970-1000

 

[os79]

Noble difficulty. Epic speed. Gigantic Earth3 mapscript. I'm just volunteering this save in case you want to check the balance, etc.
Forgot to add: Add in Custom LeaderHeads from updated MegaPack. I even have Alternate LHs but I think it's not necessary to add that.

 

[Afforess]

I've done some test on Epic and it's way too fast too. Ancient/Classical might be ok but Medieval Era is really too fast. On blitz speed, each era should last about 80 turns so on Epic it should take about 240-250 turns (250*7 eras = 1750 turns, Epic has 1800 turns). It's ok to have Ancient/Classical in about 450-500 turns but then Medieval Era takes about 120 turns which is half than what it should be. Civs enter Renaissance between turns 620 and 650 while they should not before turn 710. Same applies for following eras. Civs are researching slower because there's a drop in research rate due to lack of gold, possibly because of civics, but it's definetely too much reducing both tech diffusion AND tech costs. I would suggest increasing again tech costs from medieval era onwards; maybe smaller values than it was before but definetely higher than they are now.

Edit: also industrial era is starting about turn 840, well before reaching half of total turns, while it should start not before turn 970-1000

Okay, thanks.

 

[45°38'N-13°47'E]

Update, halfway in the industrial era, 30 civs out of 32 are using Monarchy, no matter the size of their empire or number of their cities. Just to let you know that probably the idea that smaller empires should prefer Republic isn't working very well.

 

[Afforess]

Update, halfway in the industrial era, 30 civs out of 32 are using Monarchy, no matter the size of their empire or number of their cities. Just to let you know that probably the idea that smaller empires should prefer Republic isn't working very well.

Upload a save

If I had to guess, I would guess that the AI is too pessimistic about the maintenance costs. I know earlier I had to fix the AI not understanding civic upkeep costs.

 

[45°38'N-13°47'E]

Upload a save

If I had to guess, I would guess that the AI is too pessimistic about the maintenance costs. I know earlier I had to fix the AI not understanding civic upkeep costs.

Here you are, 31 civs here, 4 using Republic, 1 using Federal (LOL, Wan Kong has just 1 city, I wonder what he's federating), everyone else using Monarchy. Around turn 1000/1800 and civs are almost reaching modern era.

 

[Afforess]

Here you are, 31 civs here, 4 using Republic, 1 using Federal (LOL, Wan Kong has just 1 city, I wonder what he's federating), everyone else using Monarchy. Around turn 1000/1800 and civs are almost reaching modern era.

Did you forget to attach the file?

 

[45°38'N-13°47'E]

Did you forget to attach the file?

I've probably closed the upload window before finishing the upload.
Uploaded now.

 

[Afforess]

Revision 768

• Improve AI understanding of civic maintenance costs
• Improve AI understanding of civic yield & commerce modifiers
• Tweak tech speed balance

45*, before Rev 768, the AI was just sort of fudging numbers and not really even understanding the commerce, yield, and actual maintenance costs. They just looked at whether the modifier was "big" or not. Now they sum up the total effect in terms of yields and commerces and maintenance on all their cities, and consider the total effect instead. It leads to much more sensible civic selection where those modifiers are concerned.

I also increased tech research costs a little bit for all eras. Let me know if it was too much or too little.

Edit: It seems better, but not great. There is still too easy a supply of food, which leads to tons of specialists, which leads to abundant gold, which allows fast technology research, which leads to better access to food and improvement upgrades, which perpetuates the problem.

It seems easiest to fix the over-abundant food problem by making a few more buildings require a citizen to work...

 

[45°38'N-13°47'E]

I also increased tech research costs a little bit for all eras. Let me know if it was too much or too little.

Medieval and Renaissance eras still seem too fast, but I've only tested on Blitz; I'll upload a save on Epic as soon as I can. I've noticed anyway than now most civs choose Republic, more or less in the same way they were choosing Monarchy on the previous revision with almost no difference between huge empires and small nations.

Edit: by the way, is there a reason why you exchanged future and transhuman eras?

 

[os79]

Medieval and Renaissance eras still seem too fast, but I've only tested on Blitz; I'll upload a save on Epic as soon as I can. I've noticed anyway than now most civs choose Republic, more or less in the same way they were choosing Monarchy on the previous revision with almost no difference between huge empires and small nations.

Edit: by the way, is there a reason why you exchanged future and transhuman eras?

I don't have a job and have a lot of free time on my hand now. Therefore, I'm willing to balance eras.

How do you balance eras? Auto-play? Actually play?

If Auto-play, then do you divide the turns by 7, then auto-play until first division. Then look for AIs that got to next era. If any got there, then it is fine. Auto-play to next division. Look for any AI that got to next era. Repeat until Transhuman/Future era division. However, if any time no AI got to next era by that division (whichever it is), raise alarm.
Is that how you balance eras?
And which XML integer do you tweak? iResearch? iTechCost?

That way, I can use my abundant free time to help out. Interesting note: Afforess have a point about food so what about tweaking few buildings? For your info, my latest game, I'm in late Medieval era (passing too fast for Epic, it is from the last revision) and almost half of my cities are approaching 20.

 

[val3]

Revision 768

• Improve AI understanding of civic maintenance costs
• Improve AI understanding of civic yield & commerce modifiers
• Tweak tech speed balance

45*, before Rev 768, the AI was just sort of fudging numbers and not really even understanding the commerce, yield, and actual maintenance costs. They just looked at whether the modifier was "big" or not. Now they sum up the total effect in terms of yields and commerces and maintenance on all their cities, and consider the total effect instead. It leads to much more sensible civic selection where those modifiers are concerned.

I also increased tech research costs a little bit for all eras. Let me know if it was too much or too little.

Edit: It seems better, but not great. There is still too easy a supply of food, which leads to tons of specialists, which leads to abundant gold, which allows fast technology research, which leads to better access to food and improvement upgrades, which perpetuates the problem.

It seems easiest to fix the over-abundant food problem by making a few more buildings require a citizen to work...

Great! I often give AI Autoplay some turns to taste, and what he is doing first?
Make a revolution and switch to a civics that decrease science outcome about 30-40%. From 1500 to 1000, for example. It seems stupid, because science - it is the first important thing in this game for AI. I hope that now he will not doing it.

Also no wonder why now AI prefer Respublic - it is the best government choise at current values, because of +20% coins.

 

[45°38'N-13°47'E]

@Afforess, medieval still seems to fast. On Epic it takes of course a bit longer than before but Renaissance is starting too early anyway. I'd say Medieval Era now lasts about 150-170 turns (it was 120-150 before) but it should be around 230-240 turns, more or less. I've not reached industrial, so I'll let you know how Renaissance is going probably in the next days (hopefully tomorrow).

 

[45°38'N-13°47'E]

I don't have a job and have a lot of free time on my hand now. Therefore, I'm willing to balance eras.

How do you balance eras? Auto-play? Actually play?

If Auto-play, then do you divide the turns by 7, then auto-play until first division. Then look for AIs that got to next era. If any got there, then it is fine. Auto-play to next division. Look for any AI that got to next era. Repeat until Transhuman/Future era division. However, if any time no AI got to next era by that division (whichever it is), raise alarm.
Is that how you balance eras?
And which XML integer do you tweak? iResearch? iTechCost?

That way, I can use my abundant free time to help out. Interesting note: Afforess have a point about food so what about tweaking few buildings? For your info, my latest game, I'm in late Medieval era (passing too fast for Epic, it is from the last revision) and almost half of my cities are approaching 20.

Well, I mostly test with autoplay to balance roughly and then play a full game for fine tuning. We're working now on iTechCostModifier in Erainfos; Afforess reduced some cost greatly for some eras a couple of revisions ago, but I think now some eras are passing by too quickly. To tell the truth, I'd like that very few people tweak those parameters; otherwise we just end up modifying them up and down everyone on his own preferences (which is something I would really like to avoid). But there is something that you and anyone else could do to help balance things; the only problem is that it might be quite boring. To test the game properly, especially on slower gamespeeds, I need a lot of process time. I could upload some starting savegames and ask you and anyone else willing to contribute to autoplay the game and post here some saves at specific turns, so that I can compare different autoplayed games. This is the reason why I've started balancing Blitz first, so that I can do it quickly: most of the times, in order to balance slower gamespeed you only need to multiply xml values a proper factor (and then you can fine tune). So, is anyone willing to help in this boring process?

 

[Rezca]

I could upload some starting savegames and ask you and anyone else willing to contribute to autoplay the game and post here some saves at specific turns, so that I can compare different autoplayed games. This is the reason why I've started balancing Blitz first, so that I can do it quickly: most of the times, in order to balance slower gamespeed you only need to multiply xml values a proper factor (and then you can fine tune). So, is anyone willing to help in this boring process?

More testing games?

 

[Afforess]

@Afforess, medieval still seems to fast. On Epic it takes of course a bit longer than before but Renaissance is starting too early anyway. I'd say Medieval Era now lasts about 150-170 turns (it was 120-150 before) but it should be around 230-240 turns, more or less. I've not reached industrial, so I'll let you know how Renaissance is going probably in the next days (hopefully tomorrow).

Ok. I just want to make something clear. We can not use the tech research cost multiplier in eras or handicaps to solve this. It doesn't fix the problem, just hides it.

<rant>

Before I started rebalancing things with trade routes, I was amazed that no one was bothered by the fact that the tech research costs were increased as much as 1000% percent and that tiny civilizations were given literally hundreds or thousands of free beakers a turn to keep up.

No such imbalance exists in vanilla BTS. The reason for the fast research is the super-inflated growth of cities, which allows for tons of extra commerce and beakers for the biggest civilizations. Not helped any by buildings which gave massive sums of commerce or food or science. These need to be brought into line. While the early eras are improving, the later ones are still terribly unbalanced.

Scaling up tech costs and tech diffusion is just a way to hide the real problem under the rug.

From what I have seen, another large part of the problem is happiness. Happiness is the natural limiter of growth in BTS, but in RAND it serves no such purpose. I have yet to play a game in recent memory where happiness was even a mild concern. From my analysis, it is the dozens of new resources added in RAND that give +1 happy that have caused the imbalance. BTS has a total of 12 resources that provide happiness, but RAND has 22!

I, for instance, can not fathom why rubber gives +1 happy. Or Semiconductors! Surely, Glass should not give +1 happy!? (On that note, why is glass even a resource? It's manufactured and only gives production speed improvement, it's not even required for any building or unit!!)

I shouldn't even have to mention that this mod preaches against the negative health effects of alcohol by giving them it a -1 health. Lovely. (But at the same time, it gives wine a +1 health and +1 happiness!?!)

Ok, I am far afield. This long-winded post is meant to express that technology research speed, while a good indicator of whether the game is balanced or not, does not mean that imbalance should be fixed by tweaking tech costs. That is like solving a shortage of money by printing more money instead of cutting back on spending.

</rant>

On to my balancing changes...

Revision 769

• Republic Maintenance costs increased from 200% to 250%
• Republic no longer gives 25% in all cities
• Republic gives 25% in the capital
• Removed +1 health from Wine
• Removed +1 happy from Rubber
• Removed +1 happy from Tea
• Removed +1 happy from Semiconductors
• Removed +1 happy from Smart Medicine
• Removed +1 from Hamlet, Cottage, Village, Town

 

[Rezca]

I've had plenty of cities come under Unahppy issues, but their growth has almost *never* been slowed or crippled by it - unless of course it was a city I took from the AI swallowed up by their culture, but in those cases I usually just raze it anyway.

Cutting back a little bit on what resources and etc provide Happiness would be a good idea I think, especially now that you pointed out how many AND has! (I never noticed myself )

 

[gay_Aleks]

Let us all have a moment of silence for the loss of the 1 from cottage improvement.