15:00 stats
total collected
p1 16742
p2 20248
K/D / army size
p1 2/11/86
p2 11/2/125
While one of your AIs managed to build second tc before 15 min they did worse in all other stats.
Default AI collected more resources, aged faster and had more army. In default game p2 managed to kill 11 villgers from p1 but p1 still collected more than jakcie ai which did not lose any villagers.
I dont really know the reasons for these differences but I think that the default AIs economy has been improved in the last couple of patches
I still think that your AI might be more efficient in the late game since default AI builds 10 extra tcs even when it no longer needs new villagers.
(old and kinda outdated in going into what my mod has) but the outrageous ai only gets a 1.15 resource multiplier with a 200 food 100 gold and 50 wood bonus while the vanilla outrageous gets a whole 1.5 multiplier. With that context I am much more proud with the results actually.
I should actually come clean on what each cheater difficulty does because I did change them. Currently:
1st cheater difficulty gets only a 200 food bonus to age up faster
2nd gets the 1.15 multiplier with 200 food 100 gold and 50 wood
3rd gets 1.5 multiplier with 200 food 100 gold and 100 wood
as a ex diamond that stopped playing for a while, the 3rd cheater is what I consider at my own skill level currently.
I tried 1v1 against the jakcie absurd since I usually play against outrageous.
I think that about 50% cheat percent feels still pretty fair since ai is so inefficient with keeping army alive.
Your AI is clearly harder and feels better to play against. They are pretty active in early and mid game but have the same issue with lategame where they like to sit in their base with bunch of units.
One issue I noticed that they have the same bug as default ai where sometimes if they have more than 3 relics but only 1 monastery they dont put the relic inside but carry it with their army as normal monk
I have done some testing with the absurd AI and it seems to be in good spot. It starts attacking maybe a bit later than default AI but keeps up way better with eco side
The biggest problem the AI has it that it is pretty bad with capturing resources, especially gold in the lategame. So the easiest way to win them is just to wait that they run out of gold and then kill them.
Here are some ideasthat could be changed to help with this:
Make the ai build keeps / towers to make their gold gatherers safe / make sure they have space to grow.
This also applies to situations where AI has like 50 villagers in the same place they should build some defences there
AI should be more aggressive with using market to get gold lategame. I think there should be a rule that if AI is in imprerial and has more than 1-2k of food/wood AND has less than 1k of gold AND market price is more than 50 g then sell extra.
This could be further expanded so that if AI has more than like 5k food/wood then it would always sell even if market price is not good and they have gold.
as long as trade unit brings less than x (maybe like 50) gold per trip then AI should not create trade units
1: I dont think I can influence that as much as you would like to other than just telling the ai to build more outposts when it think it should more frequently, but if prompted they will build more defense so I hope that just raising the numbers could help with that problem.
2: I try my best to limit market trades, unless the vanilla ai is already good at that themselves, I can’t influence what they will trade. I opt to have them just stop or limit trading.
3: I cannot or dont know how to influence that. I also cant tell the ai to build markets at actually good locations as I can only have them build building next to their own buildings Although, the ai would prevent themselves from making a lot of traders if the payoff is bad.
My gameplan with the ai is not really to let them win like how players win. I seriously don’t know how to have them win with actual pushes and destroying buildings. Rather, I have them do the things they do best (harass and micromanage their eco with their army) to outlive the player until they can’t take it anymore. That should involve having the ai play defensively like how the vanilla ai already does that. Hopefully I can improve their defense/harassing the future.
Right now I should prepare myself mentally for the expansion. I already know that it will massively change the ai files as they will include the new civs and thats actually why the last ai mod (zycat) quit back when the malians and ottomans came out.
Biggest problems include early and late game as the ai does not combine their armies in the late game and does not have a good build order early game.
Great points and you have done good job making the ai more fun to play against.
one small thing I would like changed is that the gap between cheating 1 and 2 is really small and really big between 2 and 3. I think that outrageous should cheat 1.25 and keep others as is
When testing outrageous and absurds score difference is double at points where as ridiculous and outrageous seem to be in similiar levels
I wish I can, although I can control what percentage of the population the ai reserves for villagers (55% or 110 pop for 200 pop games), I cannot really tell the ai when to stop besides that.
To sort of visualize what I think I am able to achieve with current ai modding:
Reason to make villagers (hidden in lua files I cant know about) -xxxx–> How important are those reasons (can change in the modding tools) ------> actually making the villager
basically, I am able to tell the ai to keep making villagers, but I cant tell the ai why to make them and thus when to stop besides the hard cap. I do think that I can also just lower the cap and call it which I would.
This applies to everything else the ai makes too. The only thing I can do to influence the ai is basically to overinflate the importance of certain items.
That has been a problem for forever and I dont know what really causes it because it randomly happens every few games. Every update I tweak how villager mining works just to see if the problem goes away but it does not .
Every patch small things get tweaked to try and fix problems with the ai. Im actually kind of thinking about putting up a huge change list as a google doc to go into detail on what the mod actually changes when I get to a point where I can be 100% happy with the mod’s current state and when the new expansion comes out. You can look out for that when the time comes.
I’m not working on it right now as when the new dlc comes out in a couple of weeks, most of the mod would not work with the new civs/potential balancing so there’s not really much of a point to update something that will be out of date so quickly. But i’m ready to start working on it as soon as the dlc comes out
I am here to say that I played the dlc and am working on this again! Im very thankful to the devs for the ai fixes this patch and am making an up to date version of this mod to fit in with the new civs and currently only a few changes (I had to restart parts of the mod) but im overall im satisfied with the dev’s ai changes and I hope it can continue!
.