Despite this game being rather old and known, curiously enough, talks about very basics of playing versus AI still appear time from time, with walls of noobish text resulting, and eventually making it harder to get real info for somebody actually trying to. While there is already many game info available, its a subject of drowning in the aforementioned flood eventually, and as result, there indeed is no place with at least most basic stuff covered, so i decided to wrote small faq-like stuff about it, so somebody who plan to write or read another wall of lame stuff about moo2 could check first, and ensure if he really need to. Usually its comes not from some evil nature, but simply because of lack of benchmark to compare for that, player, who eventually honestly count himself as good enough to "help" others, or "disagree, because have different opinion" to them. So i decided to gather some most basic benchmarks in one place. There are also a places where players are gather for a real-time talk, and they are valuable the most; there are also a posts and videos about actual strategies and how-to - all of them are easy to find via search engines, but too long to list here. To summarize an intro: this stuff is about game basics of game vs AI, if you (reader) is see something new there - likely he shouldnt write a walls of text about how-to in moo2 yet, and better spend a time on a learning first. If you know all of it already - again, its not necessary means that you good yet, as the listed things are only most basic facts about game, just very minimum to check yourself if you at least above zero-skill level already or not (here wouldnt be anything new for a skilled player ofc). There is some longer way to become actually good after that, but sadly most people stop at zero-skill level, not even growing to basic one, and not because they hopelessly bad, but because they misleaded to think they are good already, and stop there. End of intro.

Ill cover here only the really redundant topics i've encounter in various variations of variations over a years, as they usually eventually will rise in most discussions, and would take alot of wasted words with them, "explaining" and "disagreeing" on a stuff known for decades already, so dragging discussion down and not letting it to go further to some actual points.

So below is some stuff about playing versus AI, in a vanilla settings of game. Human-to-human games are not really covered, but if you will see something new here, dont try to mishope yourself with stuff like "but i won Billy and Joey with easy, and they were human players, that means alot". They were human players, who play worse than AI, that happens on a zero-skill level, and it dont really counts (if you still think its not true - join some place the actual players meet to agree a game, you will learn fast there. actually join even if you think its true, as here is only most basics). Mods are also not covered, as they are definitely not basics. The one of main point of mods is to address a known basic vanilla game design faults, but it could be of use only if player if know those faults already. And this text is about them mostly, and the fact that many lame and new players are still unaware of them (time to define "lame" and "new" player in this context: "new" means a player, who spent relatively few time in game, the stuff we're about is honestly new for him; "lame" means a player, who played for a relatively long time, but instead of learning it, do actually not, but still bother with "answering" a questions about it (without this habit we couldnt be able to detect lame player ofc)). There also will be "quote examples", they didnt belong to some certain person, but contaminations of similar quotes ive seen over years. Thus common redundant themes explained and respective benchmarks to them set below:
-

1. AI in moo2 is very weak.
Absurdly weak. If the setting is {player can pick any race to play vs AI} (if its not - it will be covered in #2) and
a) you think there is a need to propose something like "i play race X {description of race} and do it this way {description of some bright ideas about importance of research labs and building battleships for victory}, and im usually win" - you have to realize that the description there is of use only for "help, what im doing wrong?", as its add nothing new or important anyway. As player should always win AI with playing races about stock ones, and ofc with playing any race better than they are.
b) you think there is a need to propose something like "i play race X {description of race} and do it this way {description of some bright ideas about importance of autolabs and phasors for victory}, if im like to roflstomp the enemies, and its so absurdly easy, so its boring" - you need to realize that right race will always win the longest traditional setting (7AI players, huge, impossible, tactical combat (strategic combat is longer, but its not traditional)) less than 100 turns from pre-warp without diplomacy. Thats it, always, and way faster with other settings, and not require a single colony ship or base to be built. So if your race is still preparing to roflstomp some AI in some turn 140 (and its already +50% for average final turn) - its actually not good for a setting, its very slow. Surely it can easily kill AI, but all pre-constructed races can too, and likely you suppose your race to be better than they are. Likely it is, but also likely its not worthy to be present as good here, as it just one of very many races that can do it, and its not a best in them (some of you could even realize that even on turn 180, that is twice average the time needed to kill all AIs their "roll over AI, and so powerful, that even feel like cheat" race is still "preparing to early mid game, to unleash an unstoppable high-tech battleships" lame stuff).
c) AI is really weak, not only all 7 of it always should lose pre-turn100 if player do it right; alternatively player dont even need to spend a single research point to acquire a tech, and of course will not trade with AI for it in a whole game neither steal it (assuming playing not from pre-warp (playing pre-warp player could stop researching right after a few essential for a space travel techs)). Thats is, player dont need anything above nuclear missiles and bombs to always win a usual setting above. Of course it will be slower, than if player will actually research a more optimal tech, but still. So, if you plan to write something like "my gameplay is centered about surviving till i research {some shining tech of victory}, after that im unstoppable, and my race is really good in letting me survive till it in most of tries on a larger maps (i prefer larger maps, as small feel to random for me)" - you have to realize that right race dont need anything above the techs you got at start of game, and if your race need something above not for additional speed, but simply to be able to win - its weak race, and you doing it wrong.
d) AI is so weak, its really not even actually finished by developers due to time constraints. Keep it in mind when use a Diplomacy with it. While technically not a cheat, but cheap exploit, so viable, dont fall into trap of "but my play style is really shine when its about trade pacts and getting needed techs from AI". If you really need it to win - you doing something deeply wrong then yet. That part of game is just not finished and buggy, and in any case bonus from it is weak. Dont forget, that if you still playing after turn 100 you doing wrong already, and there is no real need to talk to AI prior too. Diplomacy is close to save-load scumming or cheat codes in this sense - it both can help to learn, but is also a danger of pitfall of dependency on them, resulting in stopping of acquiring actual skill.

Summary of #1. There is no deal to win AI in moo2 on its own if you can chose any race you want to play. Its already given for a player who slightly above zero skills. For your playstyle\strategy\race to be good vs AI reliable speed of about 100 turns for 7 AI impossible huge pre-warp is required. For smaller maps or more advanced techs its faster ofc. If your result is slower than that - the only good thing left about it then is its ability to win AI, and there is nothing special in it at all, if simplified - any stock race is able to do so under this conditions, so your playstyle\strategy\race is just one of myriad sub-par versions of it, and not really worthy mentioning, as all it can do, could been done better by better race. Alternatively no tech above nukes in needed to win a game, so if you plan to share "ill tech to something, then i can win AI, its really work, try it!" - you have to realize that you dont have to tech till that something} to win AI at all, so there is nothing special in it on its own beside a possible speed it offer (so its viable only if other {somethings} are slower).

DarzaR, I think you are trying to be helpful and not just ranting - and if you like I'll delete my post in a while to clean up the thread for your "FAQ" idea - but almost everything you say here is either completely unhelpful (like that only basic techs are needed to win without showing how), or says in 100 words what you could say in 5. If you want to make a reference to help people avoid common mistakes and repeating questions or comments you find absurd, make your points quickly and neatly.

I'll add that your "2. Special home rules" analogy of basketball is simply wrong. For most people this is a solo game, and the pleasure of using different strategies is part of the fun of that game. So when people post in this forum with questions about such strategies, this isn't "off topic". This is part of playing the game, where winning is NOT the only thing.

No, its fine, as it could help with wording, lets try to see if your main concerns about it could be solved. Assuming you did read the quoted part, and see its as non-reference still. I see a 2 alternative ways personally then:
a) i'll go and sift through all the links i have and did read\wrote in past to provide some perfect summary to you. This honestly seems as too much work for no reason (its rather a long post already), if you so drastically in need and dont want to do it on your own - hire some data analysts, so they will provide a memo.
b) as way more simple approach i'll just list only a links to certain forums instead - "read stuff there, and you will find all the explanations about". But i) the very same links will be obtained with right search engine request and ii) actually im already used the proposed approach in past, and it resulted only in "there is too many threads, i will not read them, just write it again to me personally" (while there in some thread is answer, that is written on behalf of a similar request already, so its really recursive) or "i did read them all, there is no info you were about" (inability to read or direct lie).
So i just summarized the data here, and if you curious enough to go deeper - you go deeper, search engine is your friend, if you not really want to go deeper - you still can compare yourself to some benchmarks, i did search for you as a gift. Deal here is that often when somebody can do some X, and he spent some time to be able to do that X, he stop there, assuming that X is good enough; its common learning pitfall. So, when he can do X, he tend to ignore reading about the theme X belongs to, as supposedly he have nothing new to read there, as he can do X already on his own. But its way easier if he able to compare X to actual decent value of Y without need to read something really deep, so, if interested in going better from X to Y can spend own time on it, instead of just staying at X as some imagined "top of". Or, worse, to waste own and others time, explaining how to reach X, while trying to be "helpful".

And its sorta an exercise in a non-native language for me, so im trying to practice the stuff id prefer to practice, so the point about style is most certainly right, but thats the way im write by now, maybe it will improve eventually. Here it could be my fault in using a team game example instead of some single game like some golf you mean? Whole topic is about discussions about, not pleasure from gameplay and how players should or shouldnt get it, thats their personal buisness only. Its about not derailing into offtopic while presumably trying to be helpful, say proposing some answer that is redundantly wrong, revealing that "helper" is didnt bothered to read a question, or simply ignorant in question he supposedly "solving" (so i covered most often cases of here). I used non-moo analogy, because sometimes, reading moo threads it seems that people there would have more easy time if example is actually not about the game they supposedly are, as it turns they could be not familiar enough to get a same-game example properly.
Moo2 analogue, if its more clear, probably its better to change the basketball one.

Some thread somewhere in forums called "Moo2: whats the best Creative build under conditions {detailed description of conditions}"
A says [its build Z, its good because Y]
B says [no, its build X its good because W]
C says [creative is crap, i play with my friends and we never use it, its simply weaker than build N, let me explain you why it so]
D says [C is right, creative is crap, but only lamers play N because there is M]

C and D derail a thread into offtopic here, despite follow the general moo2 theme still (no point to explain why Creative is weak in comparison to something else if rules essentially state the Creative is mandatory). But special home rules of A&B and C&D differs, and C&D ruin discussion with their answers, because they belong some thread like "Moo2: whats the best build under conditions {detailed description of conditions}" instead.

This one seems longer tho, but if its more clear, tell please. Actually im slightly lost here. You mean there are people, who like to use some strategies, thats dont let them to win? If so, whats the questions they can have about it (i think i just really misread something here, actually), sorta:
"Hello , i play moo2, but im not interesting in winning, so i put all my pop to farmers at turn 0 and add all new ones there when they spawn, and keep losing to AI! Help me to win, somebody, but keep my strategy the exact same, as its part of my fun in game!". < you cant mean something like this, right? And meaningful questions about own modified home rules are at #2, they good and valid if described clear.

MOO2 with its weak AI is a game that lends itself very well for sandboxing, allowing players to just exist in its world for many turns, and (be it either consciously or subconsciously) exploring inefficient strategies and tactics and/or pursuing very suboptimal development paths. A 'do what you like' kinda place, "where winning is NOT the only thing". Most players probably play this way and from time to time they get pummeled by the AI, causing them to reach out for advice, that basically translates to 'I am sandboxing, but don't wanna lose the game, what do I need to do to prevent it'. In such cases, the answer should not be to follow this-or-that better strategy but simply to lower the difficulty setting to Easy or Tutor.

In addition, this type of sandbox play explains the popularity of the Creative trait since, while Crea races are not the most powerful, nor do they facilitate efficient play against the AI, e.g. winning in least #turns, they do make it much harder for AI players to win. The weak AI already has a tough time attacking planets defended by reinforced-heavy-tritanium-armor, class 1 shielded, ecm jammer star base plus missile base, buying the human player decades more turns of sandboxing play.

Linked youtube video showcases a 59 turn win against 7 AI on huge galaxy with pre-warp start. I'd recommend to read Matthew Bouyack's "Baffling Tactics FAQ" in the video description explaining to what lengths he went to get this awesome result.

https://www.youtube.com/watch?v=_tqp1FVbQSU

Note: I posted without refreshing, so didn't see Rocco.40's typically concise and helpful points. Sorry!

DarzaR, thanks for the considered response. I did read it, and I think I understood your intent, but I think your point #1 boils down to:
"Every race can beat 7 AI at impossible/huge/pre-warp within about 100 turns without doing any research or diplomacy. Do a search for basic strategies, and don't bother posting unless you have a new idea that is better than this baseline."

But your summary is 8 lines long, and your full point #1 is a wall of text. It's an awful lot of writing to say "stop wasting everyone's time with your lame strategies, noobie". And it doesn't come across as any nicer than that.

Sure, people are often lazy, and it is annoying. And some people really are very bad at searches. This will never go away - those people aren't going to read a FAQ telling them to stop posting, I'm afraid.

Maybe... maybe if your post provided some useful links they would check, but this is not guaranteed, and you're right - it's a lot of work for you, for things that people should be able to look up themselves. But that would be better justification for the length of your post (lazy people are even less likely to read your FAQ if it's four pages long!).

My suggestion, though, is to make each of your points in two or three lines, maybe four. Then perhaps follow each with a "quoted" or indented example to illustrate your point, but I'd keep that short too. And to be honest, I think your points would be better as "rules" for a moderated forum you control, but are unlikely to have much impact otherwise. Sorry.

(basketball, etc.)
I get what you're saying now, thank you for clarifying, I hurried a little and thought you were complaining about people asking QUESTIONS regarding certain limiting rules. But instead, you're appealing for people to stop wasting everyone's time with ANSWERS that don't match the conditions of the question. So, that's nice advice, but I lost it in your example. I guess here again I'd say: can you make your point in two lines? That has a better chance of being read. And honestly: we know that people will still ignore it, and make those replies, sometimes because they don't read carefully. Which I was guilty of, there!

For your point #3, again a quick summary might be: "People asking strategy questions need to be really clear what conditions they are using, and anybody answering should make sure their answers reflect those conditions". Well, my point doesn't really matter, because I was talking about questions, not answers. But yes, in fact there are people who deliberately play with non-optimal strategies, and would like to know how to win despite those handicaps. And there's nothing wrong with having those discussions here - I think we agree on that, right?

For example, if a person is "roleplaying" a pacifist race that won't ever attack first, that's going to affect their strategy without changing the game rules themselves. Of course the person has to be really clear about this when they ask their questions ("Oh no! Nobody is attacking me! How can I provoke the AI into attacking my peaceful empire, without attacking them first?").

Except you say this: Your last comment about a "regular moo2 thread" could be read to mean this sort of question is weird and inappropriate here. But i hope that isn't what you mean.

(Edited to add: Rocco.40's Sandboxing comment above is better)

Thats way more simpler: i just placed those benchmarks to some public forum, so i could easily provide a link or share a quote while being any place, if i need it, instead of need to retype them again and again in case (so it slightly long to cover the most redundant ones). As it could looks somewhat weird - i added some description, so it also could be seen as a some good spiritual reading for somebody passing by if he want to. No, its very wrong summary of it. #1 is long enough to separate cases you put together and mixed. Re-read original if you actually bother about and not trying to misguide a possible reader. Thats fits farmers example above well. While there nothing essentially bad in it, still it just some junk discussion of no or very little actual value to learn from it, like "whats your favorite color?", they just take up the space like spam does. Also bothering other people asking about "i invented some own stupid rules from boredom, are you want to stop that are you doing now and talk with me about them, no, i dont have a decent explanation why those rules are needed, just i have a fun from them?" is close to cultist people behavior, waste of other's time. Not a something grave same time still. Sorta my bad, not highlighted enough, it meant to be read as "thread about regular (without some own modified rules) moo2" . Regular game rules, not regular thread, sorry if it written muddy enough.

I re-read it, and in particular I re-read your summary. While you give more detail (talking about custom races too), I still get the same basic message I described. Apparently that's not your intent, but that's what I see... it's a communication problem, and I'll take some responsibility for it, but not all the responsibility.

To be frank, I don't understand your trouble seeing that other people's objectives when playing MOO2 can be different than yours and still be valid; you seem to think that other opinions are silly or stupid ("stupid rules from boredom", really?), and now you even suggest that maybe I'm deliberately trying to misguide people, which would be insulting if it wasn't so weird.

Have you considered Rocco.40's comments about "sandboxing" at all, or do you imagine that there is no such thing because it's not your way? Anyway, it seems we don't really have common ground to work from. Your MOO2 is not my MOO2.

Just to finally clarify it - you mix some cases into some wrong one. Ill try to see it as some proposal to simply clarify it tho one more time, and not as intended harm.
a) its not any race, its any race above some level directly stated there.
b) the "any of races mentioned in a)" should only always beat AI under stated settings, no timestamp given for it.
c) some better races should also always do it not slower than 100 turns under stated settings, or
d) alternatively, some better races should also always be able to do it under stated settings, deliberately not researching or trading or stealing anything above the techs, already known by player at average tech level.
e) as additional to d) remark, other cases also doesnt rely on diplomacy to be involved.

Its easy to see:
only subset of all races a) is mentioned, unlike your "every" you pulled from some unknown place.
we can also designate subset of b) equal to a), as
subsets c) and d) include b)'s features, but not necessary include them by\from eachother
or see b) as [a) - (c)+d))], so treat as b) only not included in either c) or d)
additionally subset a) is fully included in subset e)

Now when you mix them together in one they looks like:
Suppose A and B are discussing a way to get to some place C located couple or miles nearby to them, being in some place D now, because E asked how to get from D to C.

A propose to do some work, earn lotta money and create own private jet air company, build 2 airports, in D and C, build a plane, take a flight and get there "really fast".

B say its not needed, because he can (and roughly everybody else too, E too) do it either:
1) by own foot in some time and absolutely for free, but relatively slow (still faster than A's proposal).
2) on a some public transport train, as the ticked to it really cheap, so its not essentially free, but way faster.
3) on a taxi, its will be way faster but come for more price than 1) or 2) (still nothing in comparison to A's proposal).
4) with no committing of crimes in process of any alternative way.

And A's proposal its indeed "silly or stupid" opinion, of not helping here. Note, not because A's proposal in inherently bad, its could be good, for other settings (obviously), but its useless in a given ones. If A honestly believe the stuff he proposed, the he should learn some very basic basics of transportation instead of "helping" others on it. So learn stuff he pretend to know enough to waste others time on explaining that very stuff (he actually dont know), instead of wasting others time sharing own bright useless fantasies. In moo2 case i will look like "hey, i play this game 20 years, let me help you {write some really lame\wrong crap}". There is no yet a real deal if one doesnt learn a game he care about for years enough. Deal is when such person is starting to misleading others, while claiming he's competent in it, and pretending to be helpful. Spend a time to learn it instead, assuming you do care about it indeed.

Now you pretend to add some defense for A via "misreading" it into "haha, but it boils into claim he can go from D to C by train, but same time riding on a taxi while in train and also completely for free, and without any crime (imagine poor taxi driver, who have to put a whole train with passengers into his car, and also have to make that ride for free, haha)", actually A's idea is at least equally plausible, they both weird!". And when being pointed that you describe it wrong and misleading trying to say something like "but that basically same message of your, that I described!". No, you didnt.

Now try to put yourself on a E's place as some thought experiment practice ("real-life" experience). Do you really will enjoy A's proposal while having a need to get to that C? You can call that A's behavior by various terms you like, but it actually not helping to E, only wasting of his time.

And yet, the AI seems better than about any other more modern 4X game AI that I've played against.

Low bar, I know.

I'd be awed if anybody took the time to prepare a Google deep learning agent for this game

Bump.

I have never won huge impossible galaxy by Turn 100, so that probably makes me a lame player. I will give it a go tonight.

One of my friends wanted me to play something that is repulsive and uncreative, and what you post indicates that uncreative is not as terrible as advertised, because we still get all the general tech, just like creative, but with 12 race points to spare.

This race will run into an issue if we do not get deuterium fuel cells (range 6) and there is no place to put an outpost ship. In that case, I will have to reroll the galaxy - or choose something other than uncreative.

My picks will be unification, +2 production, telepathic, rich home world. ¡Vamos!

You're lame if you play this game rather long time already, and only if you think you're good in it despite of it. Otherwise you either new player or aware casual. But as Dunning & Kruger teach us - its really hard to make lamer aware what he's a lamer. Thats why benchmark is for afterall.

Its not that terrible in vanilla game, just it provide no real advantage over other options, unless you add some additional rules. For example it would be very good (and even outright better than Creative) if you play an average tech game with 0 RP spent etc.

It cannot run into such issue, as in non-modded game such universe generation is impossible (unless there is a bug, unknown for me for all the years exist ). Worst case could be monster-blocked system as the only reachable in 6 parsec distance. Player just have to kill that monster, its surely drastically slow a run, but thats all.

Sadly after a few tries i have to confess i cant decyphe that line. Its about winning in less than 100 turns? Its about uncreative? Its just abstract info to share? Beside it, in non-modded game this race is simply impossible, as it use 20 pick points, while only 10 is available to player.

I meant:

6 - Unification
6 - Telepathic
6 - +2 Production
2 - Rich Homeworld
-6 - Repulsive
-4 - Uncreative
--------------
10 Picks

Version 1.40b23 is what I downloaded from here. Is there a later version where this changed?

My first run took a little over 212 turns and I think my flaws were:
1. Waiting for MIRV-Nukes before taking out an AI.
2. Only running around with a single battle group.
3. Lost focus while continuing along with 2 - while having thousands of BC in the bank.