I like to play on impossible mode with max races and the largest galaxy, and find that my best combos are:

Positives:
Subterranean
Population Growth +100
Food Production +1
Large Home World
Charismatic (very important)

Negatives:
Defense: -20
Ship Offense: -20
Ground Combat: -10
Feudal

I find that these are the best picks to succeed industrially and I can quickly end up dominating the game from a fairly early point in play (first or second galaxy election). If I start near Alkaris, Gnolams, Humans or Psilons, the game is too easy even on impossible. But if the game starts with Silicoids nearby, I am toast. What's your setup?

not that good at all. Read https://www.gog.com/forum/master_of_orion_series/moo2_ai_playing_stuff (assuming you not pretending to be worse than you actually are)

Ok. What are your preferred race picks?

Best race setup is the one that leaves the most points to pick
(must edit the config file to allow lower value, as default is not enough;)
and to take only what you think is sufficient to win the game the way you want to win it.

For tha asap win that would certainly be transdimensional

Thats really questionable statement at best:
1. There is no reason to talk about some imaginary mod you have in your head only in a thread about default game.
For example: I could say that the best race pick setup vs AI is [Dict StealthyShips + nothing else] (its obviously not the best one in default game). Then later "explain" that i "obviously" meant some "mod of my own that i'm playing only" (must... edit the config file... to allow... ...) and that mod is set the way StealthyShips cost -10 and every other race options cost -11 with -10 max negatives. So, while i actually indeed would name the best race pick setup for my mod, it would be very useless and misleading to ones who talk about default game (and who do not care about some imaginary mod any lamer can do with config). So if you talk about "edit" - you have to share all the actual editing values you have in mind at that very moment, if it supposed by you to be at least discussable by others.
2. From what you get the idea you're about. Cutting that your odd "editing" theme with that you tend to drag every topic i noticed you in into offtop - you basically claim that [Repulsive -Grow] and [-Food -Prod -Sci PoorHW] are both equally best race pick setups in vanilla game? And that they are of same strength etc? Beside it, why unused points matter at all according to you? They do not affect even the highest victory points, as player can get any amount of them, so probably you should mean something else then? If you just like to leave "most points to pick" as a matter of habit - its definitely not enough to claim it as "Best" overall. Playing vanilla game - definitely. (must edit the config file to allow lower certainly, as default is not enough; (actually "asap win" had to be defined too: certain rules could be about other definition of "win", say {get as much as possible RP at turn 20}, in this case TransD would be not really useful). Assuming that "Ok" as that you already did read and got all the stuff on that link its weird you asking it. At least you should got already what that question is meaningless without all the game settings named. Assuming otherwise its a waste of time, as you not only managed to skip all the info about game that come in last 20 years or so (as seen from original post), but also do not show the change in it.

upd. Well, in case its actually just a literally weird question about my abstract preferences (?) - well, it would be of not huge usability due to odd nature of question. But they should be ones like TransD, Telepathy, Feudal, Lucky, -(minus)Science, Uncreative, RichHW (not in particular order), tho i doubt it will help much.

For me there are a few setups I like to play depending on mood and universe settings but most of them are far from best and totally suicidal to try using in a multi-player game.

I like the combo of large + rich homeworld in more crowded universes. I like subterranean even if it makes for longer games. I never could use democracy well in an unmodded game where you are only allowed a maximum of ten penalty points.

Try many different picks in different settings to get more out of the game. If you already play on "impossible" you must have discovered how lacking the AI is. It is sooo easy to find a winning recipe and stick to that, game after game after <yaaaawn> sorry I fell asleep. Been there, done that and put the game away for over three years.

To read about how there have been research into which race is best and they have been tested in multi-player and against the computer may kill your enjoyment of the race customisation.

I may try that race listed in the OP and aim for diplomatic victory just for fun. Not a pick that would come naturally to me. Not on impossible though.

Best race setup?

I always play as repulsive, I don't bother with diplomacy. I like to play compact and efficient, so the custom race I've come up with is this:

-4 (BC) -6 (Repulsive) +7 (Democracy) +8 (Creative) +3 (Artifacts World) +1 (Large World)

That leaves me with 1 surplus pick. No matter. Makes for higher score multiplier if that's what you wish.
With this setup I research androids by turn 120, IIRC, and then you simply spam those for an exponential growth in RP/Production. You'll be master of orion for sure. This without even expanding yet.
The only this is you need to use the optimal research order, but basically you beeline Androids. Of course I only play impossible, and with this setup; huge galaxy, 8 players and pre-warp.

Efficient gameplay there should mean player capture all AI players already pre turn 100, so you need to redefine actually. Using "optimal research order" and "beeline Androids" in same phrase is a something close to lunacy, unless there is some negation also used in it. Using -BC and Creative while playing pre-warp you probably joking tho (in odd case you somewhy actually honest now - then why you bother answering while being so lame in game, and not reading about it instead that time spent?).

Ha! This reply certainly has a familiarity to it, from our last encounter. This time though, there's no ambiguity, you do like to be very sure about stuff don't you - like "efficient" can absolutely only equal "quick conquering".

Unless I'm a lunatic as you put it, and my judgment simply cannot be trusted, then *efficient* in this context does not exclusively refer to "quick conquering", and is not what I had in mind either.
I mean efficient in terms of micromanagement. An easy, lazy type of play if you will, but still almost guaranteed win. I think the word *compact* should have hinted to my mindset here. A single star and planet, and about turn 100, when the AI begins to get pesky, you're basically unstoppable (all their attempts bounce off, and you're free to build your empire. Mostly).

And BTW, getting androids this early is part of what makes you unstoppable. It's an integral part of this particular setup, so your criticism does not apply. Not everyone plays that trans, telephatic, feudal style. It should go without saying.

I don't know what you mean by "negation", but you don't simply beeline Androids, as I've alluded to. You need the optimal order which will enable you to research a very high science tech very early. In short you get Labs first, then auto factory, farms, soil/cloning center - then you start towards supercomputer, and you stay on that division until Android Farmer/Scientist/Worker.

Of course you have to know to utilize the techs you get along the way, to further bolster your research and production output.

You have to redefine "efficient", exactly. In terms of micromanagement too. Its very easy with benchmarks, actually. If races A and B utilizing one star and planet, in a way of easy, lazy type of play, but race A is perfectly unstoppable from turn 0 and already won with finishing screen under given settings from it by the time race B only supposedly become a "basically unstoppable" - race A is strictly better that B in terms of effectiveness. Because race A can research Androids in turn 120, and finish game the same turn, but also its can already finish game without them way prior. While race B only can research Androids at that turn 120 and still had to work all the way later. Problem is that race B is offer no advantages over A. The stuff about some player dont like to play it or so doest change it. See, you basically answering a question "whats the race you do like to play" instead, while actual question is about "best race". And its obvious derailing.

" And BTW, getting androids this early is part of what makes you unstoppable." - thats really disastrous. 1 - its not any early, as the proposed race is horrible (creative from prewarp, -BC; no production overall etc). 2. - preconstructed races are already unstoppable on this setting if played right, and they dont even have to get androids to reach it. To write something like that you had to skip all the reading bout this game that come in last 20 years. While there is nothing bad in it on its own, its weird you decided to drop that habit right now, and not only read some odd thread, but decided to "share knowledge" in it, thats what puzzle me.

"Not everyone plays that trans, telephatic, feudal style. It should go without saying." - its not about playing something or not. Its about pretty certain question of whats better, not about what somebody personally like.
The way you doing it now its:
suppose there is some thread bout skis, people discuss Rossignol and Madshus and why one's peculiarities should make its better than other. Than somebody come and say "I just dont like skis, they are not fun, i can do the same distance by foot, its slower, but it work too. Not everyone use skis style. It should go without saying" .
Sure, but in this case there is little sence to drop into thread with that "breakthrough idea". People who use skis already know how to pass distance barefoot, its somebody, who dont know about skis, not reversed. Its somebody, who cannot contribute there. By negation i meant phrase like "In vanilla moo2 optimal research order is not to beeline to androids". So it both utilize the needed words and the same time is true. Player dont need anything above nuclear missiles to win AI there. AI is that bad. If you planning to do androids just because you like it, while you know its not needed actually - its misleading to others to claim it as "good way". If you do not realize it, and thinks you need androids to win - you managed to skip even basic knowledge about production races yet. That doesnt mean that player cannot win AI the way with androids. Surely player can do, but its true roughly about any tech in game, and there is nothing particularly special in androids in this sense. And its true essentially due to AI being so bad, that one even can research androids from prewarp with horribly designed race that miss even basic knowledge about how to buld effective race, and still reliably win. As you dont actually need anything above preconstructed races and nukes to win AI.

"And BTW, getting androids this early" "
Im slightly miss the timeline there. First one should be about that odd turn 120 with androids and "This without even expanding yet". Common prod races have many colonies already that turn with strong research already, and also can research androids that time if it would been of any actual need in reality. So i dont get thats the novelty here is. That its a one planet strategy instead? And goal is "to get to androids playing from one planet as fast as possible from pre-warp"? Then its not even close to be a suited race for this goal. And dragging it to question of "Best race" is pure lunacy ofc. As its can be named "best" only if you decide not to use all the better ones.

" You need the optimal order which will enable you to research a very high science tech very early. "
What is "very high science very early here"? Its not turn 120 i think? And its even not an optimal way for a race you provided.

"Of course you have to know to utilize the techs you get along the way, to further bolster your research and production output."
Im really lost here. You mean that players (oh), actually have to build first a buildings they researched to get an output from them? That info doesnt really seems to be fresh even for 1996 then.

Summarizing stuff:
Its wrong to answer "this is race i like to play" on a question "whats best race setup". If that race is not best, its simply misleading. If it best - its still correct only accidentally, as you actually answering some other question nobody asked you about, so its misleading too.
Its also wrong to answer "this is race what is best under my own rules, as i ban all other more powerful races" on a question "whats best race setup". If that race is not best (overall), its simply misleading. If it best - its still correct only accidentally, as you actually answering some other question nobody asked you about, so its misleading too.

Seriously, you seems to even skip reading https://www.gog.com/forum/master_of_orion_series/moo2_ai_playing_stuff yet, as it seems im already redundantly rewriting some parts of.

You certainly are one despicable dude. Did you even read the thread besides the title "Best race setup"?
It's evident from reading the first few posts that the OP is not assigning the same meaning to that question as you are.
You have a bad case of tunnel vision.

Why don't you pester the OP about being so stuping and ignorant as to ask such an obviously well researched and discussed question, in that case.
And I get that you consider *efficient* to only mean *quick*, and I guess you only play MOO2 like 10mins each game, or maybe you only play online multiplayer. Well guess what, not everyone does that, and they don't have to.

Thats because OP is bailed out after his crap has been called. Maybe even went to actually read something instead and become better. And you still somehow managed to add in after. I thought you hoped to get some explanations as you missed something or so. Why else you wrote the stuff you did wrote if the question in case is such an obviously well researched and discussed already then? Also its doesnt really matter if OP got his own question wrong etc, as its common and redundant theme overall. So i could c\p stuff i wrote here later to some other place and save own time in case.

Efficient is efficient. It could mean quick, could mean something else, depends on goal set. And ofc if you set a goal and the one way of playing let you reach that goal faster than other - its more efficient, assuming equal reliability. The amount of fun player got from game, on the other hand, is not about efficiency directly. Looks like you occasionally mix them, and got ired when being called on talking about "a way you like to do it" while pretending to talk about "a way of doing it proper way". Its not about you should do it only one possible way. Its about not pretending that some suboptimal way you occasionally do like is actually good enough to share it in a loosely related discussion.

Maybe more clear example:
You know there are some basic and well known chess openings. There is some limited amount of them, and people use them. Then somebody come to chess forums and say:
"Well guess what, not everyone does that, those openings, and they don't have to. Personally i like to first jump my left knight out of position and then back into it for 10 first turns, and then do same with my right knigh for next 20 turns, and i still win my opponent eventually!"
Problem is that it not actually offer anything valuable for others. Guy is playing somebody who even worse than he is, and thats all we can get from the message. Not only the shared info cannnot help other players to become better, the somebody himself is would spent his time better reading something about those very basic openings instead (as the way he playing now is good only due to lack of actual opposition), and become way better in game, instead of playing some odd way of his. Or at least not waste own and other's time with sharing it in case of lack of intention to actually become better.

And well, i played this game for various goals to achieve, and some games has been pretty long in time. Im not sure why you think its related at all for a start, but decided to clarify that you guessed wrong, if it helps.

Well I'm sorry for not taking your cue and responding directly to the OPs question, in line with his intention.

No, I don't mix them.
The reason I play the particular way I described, is because it's efficient - from my perspective.
Say you have a certain amount of work to be done. If you can do that within a small area, instead of a large one, where you'd have to move around a lot to do certain things, which you could do more easily if the distance was short. That's more efficient for me, my body. Similarly, playing MOO2 and scrambling to send out ships or such, to beat the AI quickly, is not efficient from the body perspective, especially since I'm not concerned with simply winning, or winning quickly (which I agree is easy, even on impossible).

Yes, with my way there's probably more total work spent. Though it's over an extended period, where at any one time the amount will not cause stress. Whereas it would if it was concentrated in a short time span. Thus making the gaming experience non-relaxing.

And I never said my way was the universal or mathematically "best" way. As I said, I was responding to the OPs original question and intention.

Well, you seem to be of the opinion that if one doesn't beat the AI within turn 100, you're doing it wrong. So I inferred that your games must be short.

Beside obvious underachievement expressed by OP, its clearly seen that expressions like "best picks to succeed" "can quickly end up dominating the game" and "from a fairly early point in play" are taking up a quite amount of space in OP. Unless you will pretend now what you as a some teep felt OP's intention that they should meant something else than they mean - they are means stuff like "best" "efficient" "fast" among similar other. Not something close to "its best because i like to play it that way cos its not require moving a ships despite later i move a ships anyway, but its other story". But i have to get meaning of message from the words used by author, not an intention felt, so could be wrong here ofc. In this case ofc it should be a right way to intentionally answer on a direct question about the universal or mathematically "best" way with something that supposedly require less mouse movements (and im really afraid its still suboptimal there, as optimal playing from prewarp including about 30 turns of merely pressing a turn button).

See, thats what im about here. Its all technically valid points, right. Just they are totally irrelevant to actual question. Its like answering "could be any race picks, but only a gray colour should be used" to it. Its a valid position, but it not contributing at all.

No, im of the opinion that if one doesnt know how to beat the AI within turn 100 under default setting, where goal is to win the game - he have to learn it at least. So one later could answer about best race if needed etc. Im not of the opinion its the only stuff player have to know about game and stop there. Im of the opinion that if one doesnt know how to beat the AI within turn 100 under default setting, where goal is to win the game - there is really not a significant probability that that player do actually know more complicated things, so likely will not be able to help with them too. So im of the opinion that that player would do better not only to himself by learning a game basics first, but also to other, whom he could help later. Yes, im not of the opinion that redundant sharing of wrong and misleading info instead of learning a correct one will serve a common good. As other way its a race to bottom, where people express ignorance first directly; then indirectly, by defending own right to express ignorance instead of beating it.