Specialization

I usually specialize all three (or more if i capture one ) capital worlds (a must), then I usually have a few more specialized economic worlds with high pop (depending on how big the galaxy is), and then the rest are fairly balanced between tech and manufacturing (no advantage to specializing these really)

specialise your largest planets for economy/military production. low pq planets should focus on research.

The logic behind it is this. Your economy is the most important thing in your empire. If you're not at 100% spending with enough cash left over for purchasing expensive factories, wonders and ships you'll struggle. A good cash farm needs a high population and therefore a lot of farms. High pop means unhappy people so lots of entertainment is needed. then you need to find space for lots and lots of stock exchanges. If you can get a PQ of 20+ this should be your economic capital as well.

research can be spread out as all spending goes in to one big project so no need for large specialist research worlds (except one for your research capital and omega research facility).

You'll need a few dedicated manufacturing worlds to churn out top of the line ships especially once you get expensive techs such as nano rippers and the special good alignment defences.

Apart from this your decision on what to do with that new planet comes largely down to the bonus squares on it. A low PQ planet with a precursor mine can make a great manufacturing capital if no bigger planets are available.

I do the following:-

1) Every planet has the equivalent of 4 factories, making every planet a industrial one.

2) Then a planet is a tech planet or money planet. 2-1 in favour of money.

3) 1 or 2 planets usually of low quality r reservered for building 'specials', trade goods, galactic achievements etc...

It's 'painful', rearranging conquered planets, as the AI is still very poor when it comes to planetary development, seems to not bother building anything on some planets now or building at a very slow pace.

I only build achievements on my high industrial planets, otherwise they take too long to build. Like Sxt178 I build nothing but research facilities on low PQ planets (except I will start with an industrial site but later decommission it). Large planets mean large populations and large incomes. I had a PQ21 planet with a 700% research tile as my research capitol in my last game. I was generating over 300 research per week. Most of my middle PQ planets focus on industrial capacity with just enough stock markets to make them income positive.

My big concern about specialization comes in if my planet is invaded. My economy could be ruined, my research in shambles, and my production superpower crushed.

High population capacity planets don't need entertainment modules. Why waste a tile when you can harvest the population for ....troop transporters...

My message to my populace is this...be happy or you get sent to the battlefront

unless you play on the smaller maps that shouldn't usually be an issue. that said, you should always concider how exposed a planet is when you decide what to do with it. maybe that size 5 is a good place to put a wonder, but if it is on your border you prob wouldnt want to put something there that would criple you if you lost it. also keep in mind you can always adjust a planet later.

personally, my base planet build is a eco planet. i put 2 factorys on every planet and a starbase, the starbase often gets built over mid to late game tho. most wonders go on the 8 and under planets. i build very, very few industrial planets. usually only when i luck into a starship bonus. this is because i try to control the pace of the game so i can take my time building up my military. if your style is brute force tho that obviously is not an option. small planets are good for both starship factories and research. and remember that focusing a planet means you can use one type of improvement to augment the production of another. that means that a certain amount of generalization can be quite helpful. that is why i tend to build 2 labs worth on almost every planet. so my eco planet of size 10 will usually end up looking something like 2 factory,starbase, 2lab, farm, 3 markets and colony. with only one farm i rarely need to bother with the morale improvements, but that changes in a hurry if i build more. 1 farm-0 morale. 2farm-2morale.3 farm-4 morale. you can build a monster economy with just 1 farm per planet tho.

if you do build research or production planets dont sweat the income on them but do put a bit of research on production planets and vise versa unless the planets are very small. a size 8 production planet should have no farms(unless you need the population to resist invasion), no markets(unless you built a farm ), 1-2 labs and the rest factory or manu capital. just that little bit of diversification can add a lot of flexibility to your empire, and flexibility is your main advantage over the ai. ai has efficiency, you have adaptability. you don't put the market on these planets because they will not help when you focus.

i only make one research planet unless i just happen to get a crazy bonus tile planet after i already built my research capital, but the 2 labs per planet is enough to keep me in the research game. that said i would still recommend filling those 6 and under with labs. it is just about the only way to make good use of them.

never forget that it is the money that counts. with a fat bank acct you can rush build a factory if you need to get your shipbuilding rolling, or just rush build the ship. you can take any of your eco giant planet and turn it into the most gawdaweful shipyard in just 5-6 turns if you have enough cash socked away and if 95% of your planets are eco planes you will barely feel the pinch. this methodology is doing well by be at masochistic setting.

It's 'painful', rearranging conquered planets, as the AI is still very poor when it comes to planetary development, seems to not bother building anything on some planets now or building at a very slow pace.

Actually, I've found the planets are fully developed with only the occasional improvement taking over a bonus tile that could have been better used. (playing on Tough) What I do find interesting is if the planet is going to flip over to my race then most of the improvements have been destroyed prior to the "revolution"....specially mid-game and later. If I invaded, I usually have to upgrade or buy an entertainment improvment and add a farm.

I'll specialize loq pq planets as research. They get one factory and the rest research (though occasionally an entertainment improvment if there's room). I still don't specialize higher PQ planets, but will weigh them towards manufacturing, or research (depending on any bonus tiles) after adding a couple farms w/entertainment.

Still, if it weren't for the couple extra survey ships I build finding extra money anomalies, I'd be bankrupt quickly.

I'm still relatively inexperienced, but FWIW, I've had several conquest wins so far. I run the planets in concert. I jockey factories and research centers around as I need them. I usually start with an emphasis on factories. Once my planets are fully developed, I upgrade all but a couple factories per planet to research centers and blow through the tech tree. Then I upgrade the research centers back to factories for military production and the conquest win. I also jockey the spending distribution sliders on the domestic policy page a lot. I set the overall spending at 100% as soon as my economy can support it. I've been using diplomatic tactics to keep opponents off my back while I'm doing research and my military is weak. In the end, it's very cool having all your warships equiped with a doom ray, black hole eruptor, and black hole gun. Kicks the pants off the enemy.

One thing I noticed while spying is that the AI never uses "focus" , so the social production for most colonies is so low it ends up taking >20 turns to finish even one plantery improvement.

I on the other hand, tends to use focus (on social) a lot, so my stuff gets built fairly quickly (less than 10 turns typically for each improvement), still I suppose it might be more efficient just to increase the overall social spending...

ai has efficiency, you have adaptability.

You must be joking. The AI sucks at efficiency...

nope, not joking at all. what the ai does it does perfectly, no effort is spared. it is not subtle tho, nor especially perceptive. it applies what it knows with infinite patience. it sees one thing at a time and that is as far as it can adapt. it CAN adapt, but for the most part it adapts globally while you adapt locally AND globally, with great subtlty(strategically speaking, not tactically). you can never match it's efficiency. you may KNOW how apply your evolved strategy more EFFECTIVLY, but you do not apply what you know more EFFICEIENTLY. but it only knows what Brad told it and only within the limits of his ability to teach it to the ai(programming skill AND limits of the language and hardware used). furthermore effort has been made to differentiate the different races, they don't all use the same technique. that means less programmer/hours spent perfecting 'the' ai for the sake of having a varied(better) gaming experience.

so yea, i stand by my statement. it has been true of every game i have ever played and i beleive it applies here. the nice thing about this game it that they chose to work within this inherent limitation, most(say 99% most) games just straight out cheat to keep up with the player.

Even a child knows how to use capitals more "efficiently" then the AI.

Your point is, things it knows how to do, it knows how to do 100%. Things it doesn't, it doesn't.

But that is not the same as saying it is more efficient.

oh yea, capital placement seems haphazard at best. somehow i feel abused when i take over a size 8 planet only to find a capital there, lol. what i suspect is happening is that it has no capacity to save the capital til it colonizes a planet worthy of it, building it as soon as it gets the appropriate tech. perhaps even choosing the planet based on how quickly the capital will become functional(emptiest build queue) dominates the decision. ah well...