AI seems to not be doing a good job of prehistoric research.

[Johnny Canuck]

I started a new game with revision 5215 (i.e. with Koshling's latest AI changes), using the same settings as before (Alexander, Large/Epic, Immortal to Diety). The result? An unmet AI finished Sedentary Lifestyle on Turn 157, one turn before I would finish it. Well played, AI - and well done, Koshling

Koshling, here is the Turn 157 savegame file & AI logs, in case there is anything else in them that might be useful.

 

[Koshling]

I was disappointed to find that I STILL reached sedentary lifestyle first (at immortal) in a test game yesterday (large map, fairly slow speed [marathon I think it was]). Anyway, having analysed where the AI was lagging I have made some more changes, which I am currently testing, so watch out for another SVN update to improve early AI competitiveness later today or tomorrow.

 

[strategyonly]

I was disappointed to find that I STILL reached sedentary lifestyle first (at immortal) in a test game yesterday (large map, fairly slow speed [marathon I think it was]). Anyway, having analyzed where the AI was lagging I have made some more changes, which I am currently testing, so watch out for another SVN update to improve early AI competitiveness later today or tomorrow.

Great, i was wondering about this also, cause i have only made it to SL twice, now i make it there all the time first
Glad you finally had a chance to play you should do that about twice a week (when possible).
Another question about Raging Barbarians??

I click that on also, and it doesnt do anything different from what i am seeing except for adding in more Neanderthals?? So i guess the question is, can "WE" make the Raging Barbs more into C2C normal (i think its done by dll??) and make Raging EVEN HARDER LIKE IT USED TO BE, plus THERE IS no MORE barbarian SHIPS AROUND EITHER??? with landing parties??

 

[Dancing Hoskuld]

Raging Barbarians is supposed to triple the spawn rate of barbarians, animals and neanderthals. I put the code in my old Subdue Animals code and assumed it was transferred to the Spawn code.

However, if I get to the options screen, we can just make Raging Barbarians triple the modifier rate for each type. Barbarians don't use any Spawn XML so I assume they are still using the old method. has something happened to that old method? I have not seen any barbarian ships either.

I have been looking for a good graphic for a Pirate Cove then I could spawn ships from it until someone comes along and destroys their base. It would be part of my Bandits mod.

 

[JosEPh_II]

Religion founding is still lagging badly in SVN 5247. The adjustment to the AI about researching Religion techs already researched needs to be reversed.

Don't care that a new player complained the AI was researching a Rel Founding tech After another AI or Player had already researched it. That was the proper way.

JosEPh

 

[strategyonly]

I have been looking for a good graphic for a Pirate Cove then I could spawn ships from it until someone comes along and destroys their base. It would be part of my Bandits mod.

I have these ??

 

[Dancing Hoskuld]

Yes they may work very well. I have also been thinking that the events that spawn pirates and some other units should probably have a base put on the map and have more waves of units spawn until someone does something about them. This extension definitely goes in the Bad_Stuff_Happens mod.

 

[strategyonly]

Yes they may work very well. I have also been thinking that the events that spawn pirates and some other units should probably have a base put on the map and have more waves of units spawn until someone does something about them. This extension definitely goes in the Bad_Stuff_Happens mod.

HUH, the "spawn a new civ" i want a year ago could work there, and make an event for that also. See that is what i wanted to do with that Barbarian thread i set up months ago, but still waiting for the Mongol Camp also

HUH?

 

[Dancing Hoskuld]

I am totally against "Spawn a Civ" in random map games. In scenarios it is OK. I will have a look at it again and the camp.

But I am trying to get the screens stuff working this version first. I have done most of the pedia stuff, next is the city screen then the C2C game options screen. I am talking about and gathering bits for other stuff but except for bug fixes the screens are highest priority. OK I did the tar pit improvement as a break but they are fairly straight forward to do. The geoglyph one is still only partially done and I could not get the Moai one to work, so they are back on hold until I get some screens done.

 

[ls612]

I am totally against "Spawn a Civ" in random map games. In scenarios it is OK. I will have a look at it again and the camp.

But I am trying to get the screens stuff working this version first. I have done most of the pedia stuff, next is the city screen then the C2C game options screen. I am talking about and gathering bits for other stuff but except for bug fixes the screens are highest priority. OK I did the tar pit improvement as a break but they are fairly straight forward to do. The geoglyph one is still only partially done and I could not get the Moai one to work, so they are back on hold until I get some screens done.

Me too. Spawn a Civ was designed specifically for scripted scenarios, and is something I would hate to see in random maps.

 

[Hydromancerx]

I have these ??

I like the 2nd one. The 1st one also looks like it could be used for something.

 

[strategyonly]

Yes they may work very well. I have also been thinking that the events that spawn pirates and some other units should probably have a base put on the map and have more waves of units spawn until someone does something about them. This extension definitely goes in the Bad_Stuff_Happens mod.

Here is both. Added bonus, sound for it also.

 

[Koshling]

Here is both. Added bonus, sound for it also.

Quit hijacking this thread for something totally unrelated please!

 

[Koshling]

Religion founding is still lagging badly in SVN 5247. The adjustment to the AI about researching Religion techs already researched needs to be reversed.

Don't care that a new player complained the AI was researching a Rel Founding tech After another AI or Player had already researched it. That was the proper way.

JosEPh

Since I never modified religion evaluation there is nothing to reverse. Now, it is possible (probable even) that the things I **did** change were actin as a crutch for religion techs because they happen to have wonders keyed off them, but there was no actual change to the religion evaluation code itself, and 'fixing' I by re-breaking other code that happens to help out religions as a side-effect is not the right reaction here. It may be that the religion evaluation itself need an increased modifier now, and if after making other changes I find hat to be he case I will do so.

However, what I am seeing in my test games is a general lagging in research, rather than something specific to religion techs. Do you have evidence that the current problem is NOT just general poor research, but rather specifically not choosing religion techs within the generality of research?

 

[JosEPh_II]

I have only been taking some screen shots of the Religion section. But if you the player are able researching Judaism which puts you into the Classical Era and only 1 out of 10 to 11 AI have even researched Mesopotamism which is not normal then I took it to be a Religion problem.

But then I also began to notice during Diplomacy talks that ALL the AI had none or only 1 or 2 techs to trade by this time. And the earliest religions (Druid/Shaman/Ngaii) were not showing up either. Okay I thought DH added in Canaanism and then 2 new Techs were madeto be preqs for it, maybe this had something to do with it. But I soon realized my list of techs to trade was anywhere from 10 to 20 techs long vs the AI having 1, 2 or none. But the AI generally by Sed Life has Druid and Shaman up for trade, not so in the previous game. And the current game they did not show either until I met thru contact trade the 1 AI that did have Druid. And that AI was closet to me in size and had founded and converted to that early Religion. I still, in the current game, have not met the AI that founded Shaman. The Sumerians who are neighbors just founded Mesopotamism and they have 10 cities to my 15. But are 7-8 techs behind me (tech diffusion helping them I suspect being a neighbor with adjacent cities now).

Dumb question but where are the BBAI logs located...again? Is it in the Autolog folder in the Mod?

EDIT: Autolog file attached for current game SVN 5234.

JosEPh

 

[Dancing Hoskuld]

BBAI logs are where the other logs are in your Logs folder near your saves folder - "My Games/.../Beyond the Sword/Logs".

 

[JosEPh_II]

Thanks, just didn't see it for some reason last time. It's 26MB though will have to cut it down to size and then compress.

Did start a new game with SVN 5261. And Druid and Shaman were both founded (on Epic) at about the timeframe expected by the AI.

JosEPh

 

[JosEPh_II]

@Koshling,
Here is the BBAI log. It covers the current game ( named and started with SVN5261 and Updated to 5267) and Previous game (named and started with SVN5234).

Probably too late, but I had a "duh" moment for the last 2 days.

JosEPh

 

[Koshling]

@Koshling,
Here is the BBAI log. It covers the current game ( named and started with SVN5267) and Previous game (named and started with SVN5234).

Probably too late, but I had a "duh" moment for the last 2 days.

JosEPh

Ok thanks. I think I'm making progress, so hopefully you're seeing improvement each revision now, but it's going to be an iterative process.

 

[JosEPh_II]

It's "better" in 5261 game beginning than it was in 5234 game. But 5267 needs further play for better comparison. (I posted 5267 but it was started with 5261 and then updated to 5267, I'm slippin' up on my details. )

Edit: Got beat out to Tengrii! 1st time in a long, long time this has happened on Prince Difficulty level. Now I'm bee-lining Shinto (which is usually my 2nd Religion and normally easily gotten).

JosEPh