BTS AI difficulty level improved not effective

I'm not sure about that but I'm pretty sure that AI difficulty level was improved in BTS.

Maybe it was improved... but not as effective as it was mentioned

What I remember about BTS improving AI: remove handicaps (for example: x1.5 tech cost multiplier) and just make AI play smarter.

I'm trying to play on Monarch on large map, 15 AI.
- I start as Pacal
- I set city for commerce, I got 10 beakers per turn
- I research politeizm
- AI got politeizm first

I'm sure while playing on Monarch I have some negative tech cost multiplier.
It means BTS didn't improve AI difficulty levels as it should!

Is there any way to have same tech cost as AIs, just make AIs play smarter? If there is no way in normal BTS maybe some mod? I tried betterAI on warlords but it was not effective

Play on Noble if you want no/minimal AI bonuses... the AI is only so smart... all levels use the smartest AI.

Well I don't think the tech cost part of the handicaps got removed. I think at monarch it's more like 1.2 than 1.5 tho.

But even if it did, at 10 beakers someone with no advantage might still beat you. Pacal is financial but you weren't working a 2 commerce tile. If a financial leader was working an oasis, or they started with fishing and worked a lake tile, they could earn 12 beakers per turn, compared to your 10 beakers per turn. I don't know what speed you were playing so I don't know if it would be enough to learn it first, but they would be researching faster than you even with no handicaps.

I win prince quite easy... Monarch is fine difficulty for me, but very unfair

I know 10 beakers isn't very much but I missed politeizm multiple times trying to get it on 10 beakers, every time AI got 12? no way

Yeah, Monarch from Prince is a steep jump, I've been on Monarch for a while now, but I still sometimes drop down to Prince for more of a "fun" game, since I still lose on Monarch a bit more often than not (unless I feeling like being a cheater and reloading a lot).

How many opponents are you playing with? I play huge maps, which IIRC is 10 AI opponents. I pretty much always research polytheism if I'm a financial+mysticism civ, and I do make it there first sometimes. When you say you missed it multiple times, do you mean you just reloaded the same game? It's all based entirely on whether or not the right civ got rolled in the right place. If they did, they'll always beat you to it on Monarch, but sometimes they didn't and you can get to it first.

Even so, if you miss polytheism, just grab masonry next and head for monotheism. The AI is pretty good at going for that as well, but if you bee-line for it with a mysticism civ you should be able to make it.

MusX said:

I'm trying to play on Monarch on large map, 15 AI.
- I start as Pacal
- I set city for commerce, I got 10 beakers per turn
- I research politeizm
- AI got politeizm first

Click to expand...

MusX,

It is entirely possible for a civ to start with Mysticism and Fishing, and have the city starting on a river area. Between a river'd city, and coastal fish, you could be making more beakers. The chances of this happening are very slim, and I think it was more that the AI got the bonus than anything at all, but it is still possible.

Regardless, in the higher difficulties, you shouldn't try for any religions. Either let them naturally spread (open boarders and roads help, though they are not necessary) or conquer cities that had the religions and spread them yourself. Trying to out-tech the computer for religions is especially rough in the early game of higher difficulties.

Kesshi,

or Fin leaders just need to have oasis or gold near river.
I'm able to get later religions but it's unfair

MusX said:

Kesshi,

or Fin leaders just need to have oasis or gold near river.
I'm able to get later religions but it's unfair

Click to expand...

MusX,

I agree. It is very unfair. I wish they could develop an AI that could stand up to a human on equal footing, but sadly that is a reality that isn't going to happen.

The computer can't think 1000, 2000, or 4000 years ahead. Where as you and I can. We can see a spot with a few floodplains and a lot of forests and think to ourselves, "Hey, that'd be a great spot for a national park city in a few thousand years. Let me make sure I don't chop those forests." Where as the computer would think, "Ooooh! Flood Plains, lets build a city and start chopping!"

Humans can think "Oh, I have X floodplains, and with a city size of Y and a globe theatre, I can draft one Rifleman per turn, and grow back on the next turn, and I had better do this before I research this next tech, because Rifelmen will be obsolete and get replaced by Infantry, and those take 2 population." Computers.... can't.

Humans can program subalgorithms for every situation, where as computers can only think of that which it is trained to. To cover every situation one might run into, in Civ, the AI would need to be larger than the current size of game itself...let alone the CPU cycles necessary to execute such a rigorous AI. What you are asking for is not feasible with our currently technology, and even if it was, you wouldn't be able to afford it just due to man-hours writing all the AI algorithms.