• We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
The AI researches commerce (I need to change a bit what it researchs, tho) and throughput if it actually has an industry - which Russia sometimes doesn't get due to the serfdom nerf it sometimes can't get rid off. Russia did research clean coal. Artillery and military science are by far of secondary importance.

The AI needs to research railroads more thoroughly, though this has been a problem since 1.0. Romanticism and so on are quite useless unless you are among the first ones to research it, and buying order matters only for GPs, or the top 16 at most, so I don't see a problem with the lack of prestige research. I agree on the machine guns thing.

When I will get some time i'll run a testrun and look at the data. from that I should be able to make the AI always research the critical techs, which exist undeniably, and leave the rest alone.
 
The AI researches commerce (I need to change a bit what it researchs, tho) and throughput if it actually has an industry - which Russia sometimes doesn't get due to the serfdom nerf it sometimes can't get rid off. Russia did research clean coal. Artillery and military science are by far of secondary importance.

In your opinion :)

The AI needs to research railroads more thoroughly, though this has been a problem since 1.0.

But several of the countries were researching it thoroughly; it was at 4 for half the sample.

Romanticism and so on are quite useless unless you are among the first ones to research it, and buying order matters only for GPs, or the top 16 at most, so I don't see a problem with the lack of prestige research.

Again, this is subjective and entirely your own opinion. Buying order is all-important in AHD, since if you're not in the top 8-16 prestige nations or a GP you won't get any inputs.

When I will get some time i'll run a testrun and look at the data. from that I should be able to make the AI always research the critical techs, which exist undeniably, and leave the rest alone.

It still distorts what's already there :) And what order will it research them in? You'd need to determine which critical tech is the most critical, which is second-most...

I think the main problem is more the existence of supertechs, tbh; since there's a random element in the AI's choices, the techs themselves should all be reasonably equal rather than certain ones being vastly more powerful than the others. You can tinker endlessly with the weighting on the techs etc, and someone will always find the order or choices potentially flawed because what amounts to a good tech or a bad tech is down to the individual player rather than some objective truth. However, if all the techs were roughly equal, then the AI's semi-randomness would be less of a disadvantage compared to the player's guided actions.
 
Nah, I'd just need to decide which 12 or so techs are the most critical. After that, the AI handles itself. And yeh it's a shame about supertechs, but you can offset this by making the crap techs cost less RP.

It isn't my opinion that military science (except the gas tech), tax efficiency (if the AI can't handle taxing properly then it needs to be fixed) and all the naval techs except ship design are utterly useless. It's a fact. Buying order doesn't matter if you're below rank 8, and if you are below rank 16 you have no chance of ever becoming a GP anyway => prestige techs are useless for those below rank 16...


I made a few modifications to my mod, and now Russia behaves realistically. Backwards in 19th century but becomes a giant by 1900 (3rd place in my game. The research AI is simply much better now than before, despite a few quirks (some AI nations don't research the prestige tech etc.).

How can I post what the AI researches, similarly to how s1234567890m did it?
 
Last edited:
Is this a troll??? Any country with a lot of states NEEDS political thought for NFs (which are absolutely necessary)

b) the 'right' tech choices are both subjective and contextual, so there's no definite 'right' choice which everyone would always agree with. What you think the 'right' techs are for a given set of circumstances will be precisely the wrong ones from someone else's point of view.

Even the political techs for NFs are subject to this rule. You might think that it is 100% vital that you get those NF techs up to your limit for population. That's not always true, for the following reasons:

1) If you plan your economy right, you won't need NFs to encourage industrialization anyway. I can get plenty of craftsmen without ever using a craftsman NF.

2) After Nationalism and Imperialism, those political techs start to look really bad in their inventions. Yes, you get an additional NF, but at the cost of lots of rebel organization gain, which only gets worse as you go further into the line. Depending on your game, two extra NFs might be pointless in the face of hordes of communist and fascist rebels, since you end up killing so many people. And since you will never solve unemployment, you will never stop having POPs with high militancy and radical ideologies. The problem just never goes away.

3) NFs are useful for spamming colonization, but if you aren't going to do that, then they become less vital.

4) While I can speed up clergy creation in the early game using NFs, tech rushing political techs for NFs still isn't as helpful for literacy as grabbing those literacy techs. Getting Darwinism really early solidifies efficient literacy increases in ways that NFing for additional clergy just can't compete with. After all, you can't really expect to get much out of more than 4% clergy in a state, and clever players will note that high literacy techs mean that even 2% clergy in a state will get it close to 100% literacy long before the end of the game.

5) If I need soldiers, NFs will get some quickly, but high military spending will do it over time regardless.

6) Ditto for officers.

7) Ditto for crats, although the efficiency advantages to POP promotion mean that you want better efficiency sooner, rather than later. It's not like you can even get 7 NFs early enough in the game to get tons of crats and get that admin efficiency up. You might as well NF as best you can in the 50s and let it go with high spending.

8) You can achieve 100% literacy, 100% admin efficiency, and maximum effective leadership gain without ever going past Nationalism and Imperialism.

In other words, in several games as Russia, I have dominated the planet without going too deep into the political line. In fact, I find it easier to do so when I am working towards certain goals. (Not all the time, though.) I can watch AI nations with lots of political techs burn in the fires of revolution while I sit back as the Tsar, sip a refreshing beverage, and laugh at them while my own people line in prosperity and contentment.

I wouldn't program the AI in AHD to focus too much on political techs. Past Nationalism and Imperialism, the AI is setting itself up for revolutionary failure.
 
There are really only two no-brainer techs for the beginning:
(1) Ideological Though (for pluralism + NF).
(2) Idealism for RP boost.

Unless you expect a war within the first 5 years, these shoudl ALWAYS be the first two to research.
But from that point on it really depends. I agree that education techs are really neat, but additional NF early on allow you to promote up to 4% clergymen fast, which is WAY more powerful in terms of literacy increase than researching the early education techs.
One should try to grab Biologism fast, but I guess even that 50% boost is less worthwhile than the early NF.

As for the higher tiers, I agree partially. The NF techs beyond N&I are not too great. But: They are in fact reducing rebel power until fascists kick in - and only failing nations will have to deal with those anyway. :)
 
There are really only two no-brainer techs for the beginning:
(1) Ideological Though (for pluralism + NF).
(2) Idealism for RP boost.

Unless you expect a war within the first 5 years, these shoudl ALWAYS be the first two to research.
But from that point on it really depends. I agree that education techs are really neat, but additional NF early on allow you to promote up to 4% clergymen fast, which is WAY more powerful in terms of literacy increase than researching the early education techs.
One should try to grab Biologism fast, but I guess even that 50% boost is less worthwhile than the early NF.

As for the higher tiers, I agree partially. The NF techs beyond N&I are not too great. But: They are in fact reducing rebel power until fascists kick in - and only failing nations will have to deal with those anyway. :)

I would add Medicine for the pop growth, possible large prestige, and colonisation.
 
Indeed it's obvious that idealism is a no-brainer, but the Russian AI doesn't seem to know that. Can you people help lobby paradox into fixing this major AI bug?

It looks like nobody cares here if the AI is shit, or about gamebalance, instead preferring to ask for more fancy features like new countries...
 
Last edited:
Is this issue resolved in HoD? I'm thinking of going back to victoria 2 after the exams
I had a quick switch out of interest in my most recent game, in 1886 Russia is completely up-to-date with the idealism techs, has two education efficiency techs (no Biologism), has Nationalism & Imperialism, 75% education investment, 2% clergy, 21.5% literacy, 31 research points. Pretty decent overall, I reckon.
 
As for the higher tiers, I agree partially. The NF techs beyond N&I are not too great. But: They are in fact reducing rebel power until fascists kick in - and only failing nations will have to deal with those anyway. :)
Have you seen how powerful the fascist-boosting events are?

They're on a par with the liberalism-boosting events that fire during Liberal Agitation, and they don't stop happening.
 
They are province based, so they are really neglectible for most medium-to-large sized nations.
 
Is this issue resolved in HoD? I'm thinking of going back to victoria 2 after the exams

A quick trip to the files will tell you that PDS has redone all AI tech weights in HOD. AI may still choose another tech (there's a random factor to AI's decisions), but it shouldn't be as extreme as to prevent Russians from getting Empiricism throughout the game. They now know this is a very powerful tech.