Caveman 2 Cosmos
- Thread starter strategyonly
- Start date
JosEPh_II
TBS WarLord
Yes they priorities the human player above all else.
Thunderbrd
C2C War Dog
Sorta. There's some, yes, but to do it for 'reals' - that's a LOT of things to process to come to any kind of measurable conclusion. Very difficult. What's there is weak but improved from what it used to be a while back when they wouldn't ever attack a barb city.
Thunderbrd
C2C War Dog
Settling sites won't consider anything within existing borders anywhere. War and conquest are completely handled in another set of considerations. These two do not attempt to communicate a desire to settle where cultural boundaries exist.
If the logic that evaluates the value of a location and the logic that rejects sites blocked by enemies were separated, the blocked sites that are otherwise good could be written to a list that's used when weighting potential attack targets? I tried looking at the code and if I understand it right, the biggest issue is that other reasons settling might get blocked would need to be separately added.
Based on what I've seen in my last game, it might give the AIs a lot of extra room if they understood that part of strategy.
Based on what I've seen in my last game, it might give the AIs a lot of extra room if they understood that part of strategy.
Nightmare AI handicap for units is 44.
In future era all AI handicaps will be lowered by 5*13 = 65
44 - 65 = -21.
Same thing would happen for buildings (50) and projects will be free (65).
Only techs just will be much cheaper: 79-65 = 14
Some handicap modifiers would be set to negative values on certain eras.
Oops?
Last edited:
The "per era modifier" handicap tag I recently tweaked was not used before because it is quite volatile, especially since we have so many eras now.
There is an error in your assumption though.
It would be like this for the AI if a unit cost 1000
and it's the future era on nightmare difficulty:1000 * 0.44 = 440
▬ <iAITrainPercent>440 * 0.35 = 154
▬ <iAIPerEraModifier>
Last edited:
Ah so its multiplication, not subtraction between handicap modifier and (100-PerEraModifier).The "per era modifier" handicap tag I recently tweaked was not used before because it is quite volatile, especially since we have so many eras now.
There is an error in your assumption though.
It would be like this for the AI if a unit cost 1000
1000 * 0.44 = 440
440 * 0.35 = 154
So tech cost would be like this:
It would be like this for the AI if a tech cost 1000
and it's the future era on nightmare difficulty:1000 * 0.79 = 790
▬ <iAITrainPercent>790 * 0.35 = 267
▬ <iAIPerEraModifier>Or research costs aren't scaled by PerEraModifier?
No, it doesn't affect tech cost, I listed what the PerEraModifier affects in the SVN changelog.Ah so its multiplication, not subtraction between handicap modifier and (100-PerEraModifier).
So tech cost would be like this:
It would be like this for the AI if a tech cost 1000
1000 * 0.79 = 790
790 * 0.35 = 267
Or research costs aren't scaled by PerEraModifier?
Ah so its just easier for AI to expand.
JosEPh_II
TBS WarLord
Which in turn allows the AI to keep their Research Slider at max 100%. So it has an Indirect effect on the AI's tech research ability. Or to even invest in Espionage Slider or Culture slider.
So as the eras progress, If the player is behind, it will take much more effort and time game play wise to catch up.
Now if we could only get players to play the more current versions we "might" get less complaints about AI "too easy". BVut many will play a game started 30 SVN versions ago and never update during their play thru. But they will come and post "bugs"/"problems" that might not exist under current SVn assets. That is frustrating and makes the demand for details from the player that much more imperative, imho.
I guess I should be more consious about the aspects you point out there when reading user feedback.
I did make tech cost for the AI far higher than what it was before I rebalanced the handicap levels when compared to the tech cost of the human player:
Before rev. 9976: The human player tech cost on nightmare was 2.25 times higher than the tech cost the AI had.
Currently: The human player tech cost on nightmare is 1.266 times higher than the tech cost the AI have.
So I assumed the feedback about AI being too easy now after I made that change was warranted; and that is why I decided to try out the iAIPerEraModifier.Currently: The human player tech cost on nightmare is 1.266 times higher than the tech cost the AI have.
Last edited:
Exactly how fair is it to expect us to start a new expeditive game every time the svn is updated? What the heck? That's bloody absurd. Theres been times when there's 10 updates in a day. And it takes a day sometimes to even GET to tribalism. Are we meant to not have jobs when we play this? Families? What about other hobbies? Or even the fact that updating mid save can change things enough to a point where we feel the rules habe changed far too much to get a good clean game going?
I mean really. Is that now the expectation For us players???
You don't need to start a new game when updating, unless they explicitly say so.
Thunderbrd
C2C War Dog
Might also lead to a lot more turn time processing to make those determinations every round. These kinds of deep concept AI improvements are great projects to do someday but can't take much priority at the moment because the mod isn't even really at that stage of development where all the elements intended are in the game yet. This isn't as visible an improvement but is much more complex and difficult to implement. If you're a good coder, I'd welcome your efforts directly on that however. Team members all have their own pet projects usually.
Most updates, at least during a repair cycle like we're in, where nothing but fixes are being done, are good to update to as quickly as you can. Right now, but not after the next release or re-release, I advise updating before each time you play.
While I understand your perspective as a project leader, but you should have given certain other team members the idea that they should have HELD OFF on major changes for a week or two after the first v38 release. I've been angry over the way the post-release was handled for some time now, and it boiled over in my previous post. And to the wrong person no less.
My apologies. I'm not deleting or changing that post however, as it still has a valid statement behind it. Neither side of this project, whether it be players, or the team, should be so entitled as to demand immediate responses or actions from either side. Balancing and testing takes time and a lot of effort. So does trying to play and enjoy the game. I have a full time job, and I play other games and do other things than sit at the keyboard. Is it really such a surprise that a game I started in the middle of next week may still be in late prehistoric? Play a couple hours here and there? I'm lucky I'm not like my brother who has two kids. He has barely any time at all to enjoy video games these days. He specifically avoids early access games or games that are broken upon release because its not worth his time, money or effort to deal with it. He's also heavily moved away from really detailed and overly complex games, despite enjoying them in the past.
Lastly..
These aren't just fixes, there's been some fairly massive fundamental changes going on in progression both in how the player plays and in the way the AI plays.
Thunderbrd
C2C War Dog
If those major changes weren't major fixes than I'd agree. A fix is different than a change, it's addressing a problem that needs to be addressed. Just ignoring a problem just so it can be baked into a release is not a great idea imo.
Also consider the difference in roles between a team lead where everyone is entirely volunteer, and an employer.
JosEPh_II
TBS WarLord
I don't/we don't. You just update and play on. The Re-Calc resets the base values so the new modifiers can then start with the base values and build from there.
I started a game on 9979. I have updated many times. The last was just done today to get to the latest SVN of 100005. The game is playing just fine. I'm in late medieval Era with several AI almost 2 eras ahead of me in Tech research.
Generally if an SVN Update is going to Break a save the Modder posting will say so. And that is only done now After consulting with the other Modders on the Team 1st.
I'd agree with that on some of the stuff prior to v38 being first released way back when, but the issue I have is just how widely the tech growth has swung on multiple instances, not to mention the production issues we ran into a short time ago, and a few other things. All of that heavily damages what I consider to be a "clean game". A "Clean game" in my view is one that doesn't have underlying mechanic changes that, on the same map, would vastly change how going from turn 1 to turn 200 would play out. Midway through would be the AI or the player having significant benefits in one way or another. Considering all this occurred in the last couple of months, that's how I've been feeling regarding the SVN updates in general of late, which require ( in my view ) starting utterly new games upon update.
Similar threads
