patterned research

[Thrar]

every now and then you read about using the AI research patterns for your own benefits, especially on Emperor/Deity. However, what i could not find, is how exactly they do look. Here's my impression:

ancient age:
Literature and Polytheism seem to be quite low on the AI's priority list - and if you can hold on Polytheism for a while to get a head start on monarchy, you may be first to get it too.
As for the first tech to research, i'd go for Pottery at 80-100% (depends on lux), if i don't need it, then Alphabet at 40 turns or for religious Myst in 40 turns. I'm not sure which is better here, or if it would make sense to research faster than 40 turns here.

medieval age:
along the top branch, there's printing press, if you are not too far behind the AI in researching Theology. After that, Democracy and Free Artistry are possibilities, but chances are the AI may go after them too. Economics seems to be researched quite early usually, and thus is no good choice imo. The bottom branch consists only of necessary techs and MT (which the AI will always get asap), however it seems to me as mostly, they'll go along the top, at least Monotheism-Feudalism-Chivalry first, so you may go for the gamble to research Engineering. If there's a scientific civ around who gets it, or a lot of tech trades in the early medieval, this can go wrong.

industrial age:
medicine->sanitation is a good idea, imo. Unless there's a scientific civ who gets medicine for free, you can expect them to research at least nationalism, steam engine and industrializaton before medicine, probably more. Once they have medicine, they may go for sanitation, but usually do not. To be on the safe side, you may want to hold back medicine for a while, unless there's a scientist who will soon reach industrial, or the other civs make progress fast.
On the other hand, Nationalism and the followups Communism and Espionage, although not necessary for modern age, are typically researched very early, neither of those are a good idea to research yourself if you want to get a good trade.
By the end of industrial, mostly the player will have a tech lead anyway, if not, i think Flight isn't valued too high by the AI, they tend to get tanks and radio first.

modern age:
no idea. not the slightest. Maybe someone who knows this part a bit better can comment?


Thoughts? Additions? This is just my impression, not backed up by any tests or data, so please add your ideas.

[badman]

Modern Age: Ecology. AFAIK it's low on the AI's priority list and you want the other Civ's to have it ASAP (ate least if pollution affects the whole planet, I'm not quite sure about that).

[Nor Me]

Have you seen the expression alexman came up with after testing for the AI research priorities. It's here.
There's a more recent thread on CFC but I don't think anything more has been discovered.
It's believed that there is a random factor as well but nobody knows what it is. I've seen confirmed results that can't be explained by this formula.

Personally, I usually research at more than the minimum in the ancient age except for starting Writing in 4000BC or a later Currency or Polytheism after researching Mathematics or Mysticism at 100%.

I find that 40 turn research is only useful if you get there first and if you can, you could probably get more useful techs faster at 100%.

Remember that if you research techs the AI doesn't for trade, you accelerate the tech pace for the AI. Whether or not this is a good idea depends on your overall startegy.

I tend to find that you only end up doing this for one era. The rest of the time you'll either be too far behind the AI or too far ahead.

I've most experience on Emperor which is why I know most about the ancient age. Otherwise I'd guess on what the techs allow rather than on experience. It's just easier because you're rarely in similar situations.