JosEPh_II
TBS WarLord
What SVN version is this played on?
Not if its being played on v37 and Not the current svn. Iirc MagnusIlluminus does not use the svn to play his games on. He can correct me if I'm wrong.
Caveman 2 Cosmos
- Thread starter strategyonly
- Start date
Not if its being played on v37 and Not the current svn. Iirc MagnusIlluminus does not use the svn to play his games on. He can correct me if I'm wrong.
I wonder what would happen if Map Size, Handicap and Game Speed modifiers were flattened.
For now handicap with map size can change cost by 25x, ResearchPercent in EraInfo xml always is 1, and beeline stings can change that a bit.
Game speed can change cost of tech by 20x.
This means there is 500x leeway (between lowest and highest cost) for techs and only 8x leeway for buildings/units/improvements.
If you wanted tech cost in game equal to predefined tech cost, then you would have to pick Standard/Noble/Normal.
Red and green numbers would be 0, meaning cumulative tech modifier would be equal to 1.
And game speed modifier changes cost of tech in straightforward manner.
Would be flattening for small maps and easy difficulties be good?
That is changing this (number used in calculations):
Sizes:
Duel - 0.5x (-0.5)
Tiny - 0.7x (-0.3)
Small - 0.85x (-0.15)
Standard - 1x (0)
Difficulty (only player research cost is changed):
Settler - 0.6x (-0.4)
Chieftain - 0.75x (-0.25)
Warlord - 0.95x (-0.05)
Noble - 1x (0)
To this:
That is changing this:
Sizes:
Duel - 0.75x (-0.25)
Tiny - 0.80x (-0.2)
Small - 0.9x (-0.1)
Standard - 1x (0)
Difficulty (only player research cost is changed):
Settler - 0.75x (-0.25)
Chieftain - 0.8x (-0.2)
Warlord - 0.9x (-0.1)
Noble - 1x (0)
This way smallest map sizes and easiest difficulty would cut tech cost by 2x instead of 10x
Defined cost of last tech is 960359
, it seems like tooltip already includes TechCostModifier, as tech costs 1630 000Normal/Standard/Noble settings.
Spoiler :
Now there is much more tech cost difference between Duel/Settler and Noble/Standard (10x) than between Noble/Standard and Deity/Gigantic (2.5x).
Last edited:
All the modifiers are included in the tooltip as the tooltip shows the final [Tech Cost].Defined cost of last tech is 960359
Now there is much more tech cost difference between Duel/Settler and Noble/Standard (10x) than between Noble/Standard and Deity/Gigantic (2.5x).
I agree that the research values for map size and difficulty could need some adjustments so they have less impact on the lower than 1x values (reduction modifiers).
This is Joe's modding domain.
Last edited:
Why <iResearchPercent> in erainfos xml is used at all, if its set at 100 for all eras?
Is that because its effect is extremely dependent on map size and difficulty?
Effectively <iTechCostModifier> is used to change tech cost for each era, as it seems to be clean multiplier.
Now that building costs were reduced for Classical and later era now its easier to keep up with upgrades.
But still info of what resources you are missing would be nice to have.
Also list of buildable property buildings would be nice (buildings, that got unlocked by tech).
Example: You have +50 education/turn worth buildings.
You can build X, that would add +5 Education, Y, that would add +10 Education and Z, that would add +2 Education.
Last edited:
Because no modder on the team have invested effort into making later than prehistoric game start balanced. It was probably adjusted to 100% for all eras to really indicate that work needs to be done on the later than prehisoric era start.
Well if Mapsize+Difficulty would reduce modifier to 0.1, then suddenly techs could cost nothing or be several times more expensive, just because starting era modifier strays little too far.:
(Settler/Starting era modifier/Duel/Beeline stings, in that case beeline stings hit really hard too)
1 + [Cumulative Tech Modifier] = 1 + [ -0.5 + 0 + -0.4 + 0 ] = 0.1
So now we know, that sizes and difficulties below Standard/Noble go too low with their tech cost modifiers.
There is 2.5x difference in costs between Noble/Standard combination and Deity/Gigantic combination.
Eternity tech costs are 20x higher than Normal speed tech costs, but everything else is created 8x times slower giving effective slowdown of 2.5x (building to tech ratio).
In theory 8 turns of Eternity/Standard/Noble would see same amount of stuff happening in certain region as turn of Normal/Gigantic/Deity.
Last edited:
There would be little similarity for 8 game turns between those two setups.
Eternity/Standard/Noble
[Tech Cost] = [Base tech cost] * 1 * 20 * 1 = 20x
Normal/Gigantic/Deity
[Tech Cost] = [Base tech cost] * 2.5 * 1 * 1 = 2.5x
AI would advance far faster on Normal/Gigantic/Deity when compared to the human player than on the Eternity/Standard/Noble
Well I meant before contact with AI (or without AI altogether), that is for human players.
That is similar amount of buildings/units/improvements created/upgraded and techs researched within 8 turns by player on Eternity/Standard/Noble as in single turn of Normal/Gigantic/Deity.
By the way it seems like smaller maps make difficulty options more pronounced when it comes to tech cost:
Human/AI tech cot modifiers.
Duel + Settler -> 0.1x / 0.5x (human has techs 5x cheaper than AI!)
Gigantic + Settler -> 1.6x / 2x (human has techs 1.25x cheaper than AI)
Duel + Deity -> 1x / 0.5x (human has 2x more expensive techs)
Gigantic + Deity -> 2.5x / 2x (human has 1.25x more expensive techs)
Standard + Deity -> 1.5x / 1x (human has 1.5x more expensive techs)
Gigantic + Noble -> 2x / 2x (Everyone has doubled price for researching tech)
Last edited:
MagnusIlluminus
Emperor
While I do not normally use the SVN, I did download it early last week (on 27 Jan 2018) so as to be able to play on the VerticalSolar (then called WerticalSolar) scenario. This game is using that SVN version.
Thunderbrd
C2C War Dog
Clearly the variation between a number of factors and the way they influence research speeds are not being anywhere near being matched by production modifiers.
I know the reason for map size modifier really cannot be, which is kinda an interesting head trip. It adjusts research rates because you won't have as many cities on a smaller map, but then that really messes with the balance vs production. Perhaps those modifiers should be toned down a lot (brought much closer to 0) to help smooth that out a bit and just let the natural fact that your opponent(s) are in the same 'less or more cities' status as you are.
I think, just looking at it, that this is probably a fairly good suggestion after analysis.
But we also need to look at the production variations. Do they match up to anywhere near the same amount on difficulty variations? On game speed variations? I can look at all this hands on but I'm in the middle of some other projects. Talk it out guys. Find where the production variation is so dramatically different here. I think it has to do with difficulty variations between research and production more than anything. Maybe the formula changes are in part to blame, where difficulty variation doesn't have the same impact on construction rates as it does on tech achievement rates now. Toffer, maybe you can look at the code with an eye for that?
I would actually be all for removing the research modifier on map size altogether.
A long-term stretch goal should be to make it easier for small nations to keep up with large nations in tech, if we manage that one day, the map size research modifier should not be necessary.
A long-term stretch goal should be to make it easier for small nations to keep up with large nations in tech, if we manage that one day, the map size research modifier should not be necessary.
Several short questions:
1. How much gold does a chest equal to?
2. Does population intervene in criminal spawn rates of buildings and pseudo-buildings?
3. Will we be able to play the big maps in V38 without a bug quitting our game very often?
4. Now gold is fixed with chest, how much is the maximum of research, culture, etc, 21 million? 210 million? How many decimals does it takes?
1. How much gold does a chest equal to?
2. Does population intervene in criminal spawn rates of buildings and pseudo-buildings?
3. Will we be able to play the big maps in V38 without a bug quitting our game very often?
4. Now gold is fixed with chest, how much is the maximum of research, culture, etc, 21 million? 210 million? How many decimals does it takes?
Toffer, I made your tech calculator in excel file.
Units are cheaper in easier difficulties.
Settler - 20%
Chieftain - 30%
Warlord - 90%
Noble - 100%
Prince - 105%
Monarch - 110%
Emperor - 115%
Immortal - 120%
Deity - 150%
Building cost for player doesn't depend on difficulty at all.
But AI has cheaper buildings on higher difficulties.
Settler - 160%
Chieftain - 130%
Warlord - 110%
Noble - 100%
Prince - 80%
Monarch - 90% - FOUND BUG HERE (From Monarch iAIConstructPercent should br 75%/70%/65%/60% or something like that)
Emperor - 85%
Immortal - 80%
Deity - 60%
Game speed iConstructPercent modifier (upscaled costs option)
Normal - 100% (135%)
Epic - 200% (270%)
Marathon - 335% (405%)
Snail - 500% (675%)
Eons - 650% (875%)
Eternity - 825% (1080%)
Units are cheaper in easier difficulties.
Settler - 20%
Chieftain - 30%
Warlord - 90%
Noble - 100%
Prince - 105%
Monarch - 110%
Emperor - 115%
Immortal - 120%
Deity - 150%
Building cost for player doesn't depend on difficulty at all.
But AI has cheaper buildings on higher difficulties.
Settler - 160%
Chieftain - 130%
Warlord - 110%
Noble - 100%
Prince - 80%
Monarch - 90% - FOUND BUG HERE (From Monarch iAIConstructPercent should br 75%/70%/65%/60% or something like that)
Emperor - 85%
Immortal - 80%
Deity - 60%
Game speed iConstructPercent modifier (upscaled costs option)
Normal - 100% (135%)
Epic - 200% (270%)
Marathon - 335% (405%)
Snail - 500% (675%)
Eons - 650% (875%)
Eternity - 825% (1080%)
Last edited:
1. Million gold.
2. Higher population in cities means more crime is generated. Depending on crime level criminals may spawn.
3. That bug is most likely MAF - memory limitation for game, that can't use more than 2.9 GB of RAM - how much RAM do you have? With 8 GB you should have enough space for game and other things.
Ok, thanks, I got enough RAM to spare.
Last time I played Gigantic was 2 years ago in a very buggy build, it was especially buggy in the biggest map.
Just wanted to be sure.
Game is much more RAM efficient now, you can try to play on Gigantic (or Giant)/Noble/Normal to see how it plays.
But most likely Huge is biggest map, where you can play safely (like no more MAFs than once per day)
Thunderbrd
C2C War Dog
The more I think of it, the more I think you're right. The differences is naturally adjusted in reference to all other players anyhow. It does throw off the comparison to the dating some to remove it entirely though.
Exactly what 'win for losing' accomplishes.
This must be in Toffer's modmod but I suspect it means 1 million gold.
I can't recall. Population might form a modifier to spawn rates or maximum rates but I don't remember if I included any consideration for it. What DOES influence spawn rates dramatically is how many criminals are already there. Each one present reduces the final chance by a tenth of the full chance based on crime level so effectively there is a limit of 10 criminals in a city before spawning stops entirely. Strategy hint: those CAN be your own criminals.
Seems we've had some serious improvements to that.
A bit too generic a question with too wide a parameters to answer. I'm not sure without digging through the code and working it out.
This makes sense and seems to fit. However:
This could be a problem and part of why someone on Noble would complain they are not getting their buildings built fast enough while on Immortal, buildings are, in comparison to tech achievement rates, much cheaper (since they haven't been adjusted.) By playing on Immortal for our tests, Joe and I have seen comparatively cheaper feeling buildings by a long ways compared to the tech achievement rates, which are stretched out. These need to be given the same arc as the units! (which incidentally matches the same arc as difficulty has for tech acheivement rates.)
Only the AI. And the bug is at prince, not monarch. Should be 95% if we're going off the other charts and patterns of progression as a guide. I'm not convinced this doubling of benefit for the AI is necessary to be honest. Maybe all these should be just a flat 100 to keep the AI set at what noble is set at without being adjusted further by game difficulty of the player. I think Joe would agree with me on that. But if we want the harder settings to be even harder, I can see why this is setup this way.
Thunderbrd
C2C War Dog
@JosEPh_II So the lack of use of the construction cost variable on difficulty seems to be our primary bugbear here. See above. With our tests so far seeming to see the balance correct at Immortal level, I suspect we'll need to also shift the overall global on all costs down by 20% once we bring UP the costs by difficulty according to the normal chart progression.
Do you concur?
Do you concur?
Thanks, I also noticed that the mod is now much better optimised in SVN 9634, the SVN that I'm now playing.
Edit: typos
JosEPh_II
TBS WarLord
The bug is at Prince should be 95%. The more difficult levels do not scale the same as the easier levels. So don't assume they Prince thru Diety should be 10/20/30% increments like Noble down to Settler.
I have posted before that the Handicaps need to be cleaned up but was waiting for v38 to be released before starting that project.
EDIT: During the time it took me to type and post this you all added 5 posts inbetween! Yea Gads I'm slow.....and
........and tired.......
Last edited:
Similar threads
