Balance Factors

[raxo2222]

Well... I'm ok with that. Just saying that's the most forgiving setting so if you're on settler, and things are gauged for noble, then you're going to be hitting the mark way early and perhaps it's better to use the EASIEST case to reach 12k scenario possible. Going off Noble with human ingenuity hitting Sed Life at 12k means that most AI players will probably reach it a bit late and that's an interesting way to get a measure on true AI effectiveness vs Human.

Player has 60%/40%/20% cheaper techs on Settler/Chieftain/Warlord handicaps.
AI has 12%/8%/4% more expensive techs on those handicaps.

 

[T-6000]

Personally I feel like human players getting disgustingly far ahead of the calendar on low difficulty settings is kind of a civ tradition, as well as a solid indicator that it's probably time to move up a difficulty level or two, so it doesn't really seem like a bad thing assuming the moderate difficulty levels are more balanced in that respect.

Edit: Also it seems like it'd be a tricky thing to prevent without making the difficulty settings shallower or nudging the calendar progression based on difficulty, neither of which feel like good outcomes.

 

[raxo2222]

@Thunderbrd also I don't beeline in my games (and didn't do that when I calibrated tech cost global), so others would be better at maximizing of research

That is if tech leader is meant to research Sedentary Lifestyle at 9.77% +-0.977% of game, and then I calibrate that, then optimizing player will reach next era faster.

So I guess I could compensate that by playing on one handicap below Noble

 

[raxo2222]

Personally I feel like human players getting disgustingly far ahead of the calendar on low difficulty settings is kind of a civ tradition, as well as a solid indicator that it's probably time to move up a difficulty level or two, so it doesn't really seem like a bad thing assuming the moderate difficulty levels are more balanced in that respect.

Edit: Also it seems like it'd be a tricky thing to prevent without making the difficulty settings shallower or nudging the calendar progression based on difficulty, neither of which feel like good outcomes.

Handicaps above Noble only cheapen techs for AI.

 

[SirJohnEh]

Test game measurements and observations to my completion of learning Sedentary Lifestyle (tech learning sequence, turn counts, calendar dates, science and gold rates, etc.) to provide feedback on the recent Civics and other adjustments including Complex Traits. I ran 3 tests at different speeds (Long at SVN10517, Epic at 10534, and Marathon at SVN10561), all at Emperor difficulty on Standard C2C_World maps with 7 AI opponents.

Summarized comparison results for the 3 tests, plus the correlation I calculated between turn counts and the Technology beaker requirement uplifts at these 3 game speeds:

(1) At Long speed (on SVN10517), I completed Sedentary Lifestyle at turn 472 in 9449 BC, learning 88 technologies).
--- I believe this matches general turn count and game date expectations. At this point, I was close to to the leading AIs in score and technology progress.

(2) At Epic speed (on SVN10534), I completed Sedentary Lifestyle at turn 713 in 9357 BC, learning 86 technologies).
--- I believe this matches general turn count and game date expectations. At this point, I was close to to the leading AIs in score and technology progress.
--- Correlation note: This turn count is 1.51 times the 472 turns at Long (within 1% of the 1.5 times beaker uplift).

(3) At Marathon speed (on SVN10561), I completed Sedentary Lifestyle at turn 965 in 9143 BC, learning 85 technologies).
--- I believe this matches general turn count and game date expectations. At this point, I was far behind the leading AIs in score and technology progress.
--- Correlation note: This turn count is 2.04 times the 472 turns at Long (within 2% of the 2.0 times beaker uplift).

Other factors impacting these game tests at different speeds:
(1) In all 3 games, I followed a fairly consistent playing style. However, actual game play at each speed on randomly-generated maps depends on many factors, that may act to balance each other out.
--- (Confession time) I mitigate the risk of significant turn count differences (at least in the early eras) by checking each newly-generated map using WorldBuilder, and regenerating to avoid maps that are not reasonably playable or fun (isolated on a small continent, excessive ice or desert or mountains, poor starting location or sites for the next few cities (lack of ocean access, river access, productive terrain, or near-by key resources).
--- Other game variability factors include resource-dependent and culture-dependent buildings and units, random event bonuses and penalties, learning resource-dependent optional technologies or not, AI interactions (particularly war, trading, TD and WFL), and running at different SVN levels (see example below). I ran all tests on with TD on but WFL off. For the Marathon game I added spreadsheet calculations splitting the overall average of 55 beakers per turn as 42 "earned" beakers (76%) from the slider, plus 13 (23%) TD bonus beakers plus 1 (2%) from random event netted bonus minus penalty beakers.

(2) Test #3 at Marathon speed was the only one that included the SVN 10543 gold reductions. I noticed a definite impact on the gold usage slider - I normally run at 90% to 100% science all game, but in this game was down to an average 85% for science in the first 3 quarters of the game, and to 55% in the last quarter. In addition, I missed getting my normal Golden Age boost from building the Captured Fire National Wonder right after founding my third city (probably due to my lack of attention, but I'm blaming the slow game progress).

Complex Traits Selection and Impact Notes:
- Reached Leadership Level 1 at turn 136 at Long speed, versus turn 214 at Epic (1.4 times slower) and turn 514 at Marathon (3.8 times slower). Took positive Preeminent I trait in all cases, for faster growth and impact on the capital city.
- Reached Leadership Level 2 at turn 250 at Long speed, versus turn 396 at Epic (1.6 times slower) and turn 901 at Marathon (3.8 times slower). Took positive Financial I and negative Minimalist I traits in all cases. Took Financial I trait because in my testing last year I found the original Financial trait was better for technology attainment speed than Scientific or Industrial. Took Minimalist I trait as the "least-worst" for my playing style (especially the small science boost). (Confession Time: Another reason is that I felt Minimalist was the least objectionable Negative trait name to link myself to.)
- Reached Leadership Level 3 at but forgot to log the turn number (probably about turn 625) at Long speed. Took Scientific I trait at Long speed for the science boost to speed technology progress, but did not reach this level when testing at Epic or Marathon speeds. `

- For further details and observations, see file "Test 1904A+B+C Measurement Results Summary.txt" in the attached zipped file.
- For an excruciating level of detail, see file "Test1904A+B+C C2C Game Target and Detailed Set-up.txt" in same zipped file.
- For the truly brave with LibreOffice installed, see my event recording logs and calculations for each of the 3 games in the zipped attachment, named "Test1904A TechTree Learning Log" at Long speed, "Test1904B TechTree Learning Log" at Epic speed, and "Test1904C TechTree Learning Log" at Marathon speed (probably the best of the 3, but still has flaws and unresolved calculations).
Start-up and completion saves also included in the zipped attachment.

UPDATE: I started poking at the split between beakers that are due to the current slider science percentage setting, versus "base" beakers at 0% science allocation on the slider. I think the "base" beakers at 0% allocation represent beakers from city buildings, special units, civics, traits and so forth. I plan to run another test game to capture and compare the results. At first glance, the "base" beakers seem more important that the slider science rate.

 

[Thunderbrd]

Thanks for the feedback!

 

[raxo2222]

Looks like <iStartPercent> in era infos is responsible for turns remaining and this means displayed date too when starting in Ancient and later eras.

But currently it is too rough to commit its corrections in era infos - 10x (or even 100x) higher precision would be needed - so 1000 (or 10000) would be end game instead of 100.

 

[Thunderbrd]

Looks like <iStartPercent> in era infos is responsible for turns remaining and this means displayed date too when starting in Ancient and later eras.

But currently it is too rough to commit its corrections in era infos - 10x (or even 100x) higher precision would be needed - so 1000 (or 10000) would be end game instead of 100.

Sorry, but you lost me on most of this. What is too rough? What problem(s) are you trying to solve?

 

[raxo2222]

Sorry, but you lost me on most of this. What is too rough? What problem(s) are you trying to solve?

iStartPercent is used if you want to start game in Ancient and later eras.

Its values can be inbetween 0 and 100 (integers).

Lets say you want to start in Ancient era on Normal or Eternity.
You can't set iStartPercent to 9.77 - this is when calendar should be on 12000 BC (but in reality its very close to that date).
9.77% of 2000/20 000 turns is 195th/1954th turn in game.

Now closest iStartPercent for Ancient era would be 10 - that is 200th/2000th turn of game.
So this means it can be up to 0.5% off mark - up to 10 turns off target in Normal speed or 100 turns off target in Eternity speed.

 

[Thunderbrd]

I see. You're working on trying to add better support for after-prehistoric starts and focusing on the calendar having problems with those. Got it.

 

[Noriad2]

After a year of absence I decided to update to latest SVN and check out the game as it is now.
Trying out regular deity/large map/snail. I noticed the game was a lot slower than I was used to so I eventually changed game speed to Epic. Not a complaint, just an observation and an adjustment.

So I played 2 games on regular deity/large map/epic, to somewhere in ancient.

Both cases, despite having tech diffusion on, the tech development in early prehistoric was extremely slow. After building 2 wonders, some wanderers for exploration and some buildings, I ended up several times with nothing worthwhile to do with my hammers, except wasting them on Meager Wealth, or spamming more wanderers to chase birds (at 8% subdue chance).
I noticed that the slow tech was because of stacking several tech penalties from your starting civics. Eventually I went through this period, and as I collected myths and better civics, the ratio of tech speed (getting new stuff to build) and hammer cost of stuff became satisfactory. It was just the beginning part that had too slow tech development in my opinion.

While I'm all for putting penalties in the game, I think that putting too many tech penalties right at the start, when there is little to build and nothing to do is a bad idea. Just entering through your turns waiting for tech to catch up is boring. Better speed up tech in the early game and instead put more tech penalties in the mid and late prehistoric if you want to slow down the player.

Furthermore, there is way too much money in the prehistoric. First game the only time I went negative income is when I built up an army to take another city in a spiked clubman rush. But I already had a sizeable amount of cash from goody huts, wonders of nature etc. When I finally took the city (14 tiles from my capital) I braced for financial impact, but only got 1.69 cost distance from palace and 2.22 number of cities. Which is ultra cheap. I finished the prehistoric with thousands of money and still rising. I went wild conquering stuff (elephant riders are still overpowered) and bumped into the happiness penalty from number of cities before I reached copper working. Which meant no more expansion until bronze working or monarchy.

2nd game was the same. Tediously slow tech in the beginning, often building meager wealth to sit out the turns, it became better later in the prehistoric. I did a military rush when stone macemen became available. Ended prehistoric with thousands of cash, kept expanding until I bumped into city happiness limit during early ancient.

So my suggestions are:
1) speed up tech in the early prehistoric and slow it down in the late prehistoric. E.g. by removing some tech penalties from starting civics but keep those that remain for longer.
2) increase city upkeep cost from distance to palace. Possibly also from number of cities.
3) Once elephant riders are available, spamming those is a no-brainer. I'd think about increasing their hammer cost and upkeep cost given how expensive it is to tame, keep, train and feed elephants in real life.

 

[JosEPh_II]

After a year of absence I decided to update to latest SVN and check out the game as it is now.
Trying out regular deity/large map/snail. I noticed the game was a lot slower than I was used to so I eventually changed game speed to Epic. Not a complaint, just an observation and an adjustment.

So I played 2 games on regular deity/large map/epic, to somewhere in ancient.

Both cases, despite having tech diffusion on, the tech development in early prehistoric was extremely slow. After building 2 wonders, some wanderers for exploration and some buildings, I ended up several times with nothing worthwhile to do with my hammers, except wasting them on Meager Wealth, or spamming more wanderers to chase birds (at 8% subdue chance).
I noticed that the slow tech was because of stacking several tech penalties from your starting civics. Eventually I went through this period, and as I collected myths and better civics, the ratio of tech speed (getting new stuff to build) and hammer cost of stuff became satisfactory. It was just the beginning part that had too slow tech development in my opinion.

While I'm all for putting penalties in the game, I think that putting too many tech penalties right at the start, when there is little to build and nothing to do is a bad idea. Just entering through your turns waiting for tech to catch up is boring. Better speed up tech in the early game and instead put more tech penalties in the mid and late prehistoric if you want to slow down the player.

Furthermore, there is way too much money in the prehistoric. First game the only time I went negative income is when I built up an army to take another city in a spiked clubman rush. But I already had a sizeable amount of cash from goody huts, wonders of nature etc. When I finally took the city (14 tiles from my capital) I braced for financial impact, but only got 1.69 cost distance from palace and 2.22 number of cities. Which is ultra cheap. I finished the prehistoric with thousands of money and still rising. I went wild conquering stuff (elephant riders are still overpowered) and bumped into the happiness penalty from number of cities before I reached copper working. Which meant no more expansion until bronze working or monarchy.

2nd game was the same. Tediously slow tech in the beginning, often building meager wealth to sit out the turns, it became better later in the prehistoric. I did a military rush when stone macemen became available. Ended prehistoric with thousands of cash, kept expanding until I bumped into city happiness limit during early ancient.

So my suggestions are:
1) speed up tech in the early prehistoric and slow it down in the late prehistoric. E.g. by removing some tech penalties from starting civics but keep those that remain for longer.
2) increase city upkeep cost from distance to palace. Possibly also from number of cities.
3) Once elephant riders are available, spamming those is a no-brainer. I'd think about increasing their hammer cost and upkeep cost given how expensive it is to tame, keep, train and feed elephants in real life.

Since I'm doing the Civics Your criticisms are Noted because you are an experienced Deity player and I respect them. But I have to say I feel you are premature in your analysis. And that your view of which Maint Malus to use is a bit outdated.

The Civics are far from finished work. I have had base values in Research, Gold, and Production changed by fellow modder(s) ,during the course of even adjusting just the starting 3-5 Civic in each Category. That have forcibly changed what and how the early set of Civics work and interact. The Research rate Must be slow in early Prehistoric. Because once you hit late Ancient and then Early Classical it skyrockets, especially for the AI who is given bonuses that increase as the Difficulty level increase. And also by Era reached. So that by the time you the player hit Early to Mid Classical era the AI was consistently researching the Ren and Ind Eras Techs. Some Leaders even had researched tech in the Atomic Era! Of course all on the upper 4 Difficulty levels in particular.

Gold is still a work in progress. Because by the time you get to late Ancient and early Classical you can have a Treasury (on Immortal Diff) in the 30-70K Gold range. With income per turn in the hundreds and even low thousands without beelining every Wonder available. Or capturing AI cities as well. You start Taking over a neighbor and by accumulating and keeping the larger Pop Cities they had it only stand to reason that your Gold will expand proportionally.

I have maint Modifiers besides Dist to Palace and Num Cities that are much more effective in throttling the Gold rate. I have turned many of those down or off in the Early Preh Era. But they begin to come into play later. BUT...If they come on too strong too early the player And the AI will stick longer on the earlier Civic in a category. There Must be incentive to eventually move away from those Early Civics. The balance is finding when and how to trigger the need to move on to the next Civic.

Question: because you had TD On how much did it benefit you the Player (even being on Deity) vs the AI? Did you receive extra research at Any time from it? I would venture that you probably did being On Deity.

All this said, balancing the interactions of the Civics to game play And to each other is not a matter of just tweaking a few variables and then running Auto play for hundreds of turns. The early Civics in each Category right now are being worked to be an interweaving effort. That each new civic unlocked must be weighed and considered to what you already have active as to whether you should activate immediately Or wait till another Civic is unlocked Or a particular profound game play Tech is unlocked.

I am also under the constraint that all Civics Must be "complicated". Moreso as the Eras progress. This is a direct edict from the Top of the Mod food chain.

I would have been further along with them if it were not for the 3 sweeping changes I alluded to early in this reply. Each time these sweeping changes were made I basically had to restart Test games all over again.

And this is why they are not finished nor have changed much once you hit the Civics that come into play by Early Med Era. Any Civic past Med Era is still basically the same that they were Before I started this task. And Pepper2000's Late Game Civics I have basically not touched. Although iirc raxo has.

I could have addressed you particular points (1. and 2. ) but felt that a general overview of the overall Civic status was due 1st.

 

[Noriad2]

Hi JosEPH_II,

Regarding your question on Tech Diffusion (TD), I loaded a few old saves. I have contact with several other civs.

Oi, I just noticed: I accidentally still have nightmare mode on haha! Wasn't my intention.

Gathering: +4, TD: +4
weaving: +8, TD: +11
Theft: +18, TD: +24
composite tools: +56, TD: +30
Atlatl making: +74, TD: +15
mining: +178, TD: +20
Orchards: +336, TD: +41
Exploration: +413: TD: +0

As you can see, TD is quite a bit of help speeding up in the beginning, but later it becomes more and more trivial and vanishes as I progress in Ancient Era. I'm also number 1 score worldwide now that I conquered my 11th city (removing fog-of-war with CTRL-Z) but the number 2 is researching writing (still 11 techs away from me) and the number 3 is researching ornamentation (which is still 7 techs away from me). So Tech diffusion stopped working even though I'm still behind on the AI. It almost looks like the display shows Win For Losing instead of Tech diffusion, although I have TD on and WFL off. In fact, no tech I can select gives any TD. Not even on archery (which I'm sure the AI has because of archer units existing on the map) or on prehistoric techs I skipped.

Soo...display bugtest: I research archery. Numbers seem correct; the research bar gets upped by exactly the amount of research I produce. But no TD bonus. Even though the AI has archery. And has several other techs ahead of me.

Perhaps I should make a bug report? TD seems wonky.

 

[JosEPh_II]

The Display will give a combination for TD and WFL even if only one of them is actively On.

In my test games for Civics I do Not have either ON. Mainly because the AI is much better than it was a year ago at researching. For awhile (back in the late winter) the AI on Immortal would be 2 full Eras ahead of me in Research (an Era to an Era and a half on Emperor). This was turned down a notch or so in late spring by raxo. I feel the Need for TD being On (or WFL either or both) is not what it used to be for the mod's play.

 

[Noriad2]

I made a bug report on the wonkiness of TD as I observed:

https://forums.civfanatics.com/thre...f-february-2018.629241/page-192#post-15486993

Regarding money excesses, perhaps make all mounted units need +1 money upkeep? It takes a lot of time and effort to train elephant units, that's why they were only used occasionally in history, and only by the richest empires. Same for later units that are based on expensive equipment like helicopters. The AI doesn't make all that many mounted units (especially not in the beginning) and is swimming in money anyway. Ships may be exempt as the AI tends to build a lot of ships.

 

[JosEPh_II]

The AI doesn't make all that many mounted units (especially not in the beginning)

This is really a game specific observation and is dependent upon Mounted animal resources available. And in the beginning neither do most players for the same reason. Resource availability and tech to use it.

 

[Noriad2]

I have maint Modifiers besides Dist to Palace and Num Cities that are much more effective in throttling the Gold rate. I have turned many of those down or off in the Early Preh Era. But they begin to come into play later. BUT...If they come on too strong too early the player And the AI will stick longer on the earlier Civic in a category. There Must be incentive to eventually move away from those Early Civics. The balance is finding when and how to trigger the need to move on to the next Civic.

Don't dist to palace and num cities depend on difficulty level? So you can increase these and perhaps other throttles on gold for higher difficulty level without penalizing the AI.

 

[Thunderbrd]

The Display will give a combination for TD and WFL even if only one of them is actively On.

Yes, but only that which applies. It's not giving WFL bonuses if WFL is not on, for example. The wording is delivered through python so I didn't have as much programming skill to vary the message so just said the difference between normal and adjusted accounts for both, which basically it does, regardless of which it is or both. Basically it just determines (Actual amount you are receiving in total - amount it would be without TD and/or WFL) and delivers that as TD&WFL amount.

 

[Noriad2]

Yes, but only that which applies. It's not giving WFL bonuses if WFL is not on, for example. The wording is delivered through python so I didn't have as much programming skill to vary the message so just said the difference between normal and adjusted accounts for both, which basically it does, regardless of which it is or both. Basically it just determines (Actual amount you are receiving in total - amount it would be without TD and/or WFL) and delivers that as TD&WFL amount.

Toffer90 gave a good explanation in the bug thread. If less than a third of civs have a tech, it's easy for TD to get rounded down to zero. If more than two-thirds of civs have a certain tech, the TD bonus is magnified. You don't have to have met all of those civs to get this effect.

 

[Thunderbrd]

Toffer90 gave a good explanation in the bug thread. If less than a third of civs have a tech, it's easy for TD to get rounded down to zero. If more than two-thirds of civs have a certain tech, the TD bonus is magnified. You don't have to have met all of those civs to get this effect.

I read it later and he did explain it well - I agree. I didn't edit my response here because it was addressing the reason that the display itself will say 'from TD &/or WFL' on the added amount, regardless of which you are getting or both. I'm just hoping to enhance awareness of the reason it doesn't get more specific.