General Category > AI War Classic - Strategy Discussion

Need Strategy Advice?

(1/9) > >>

x4000:
If you're a new player, please make sure to start with the four in-game tutorials.  They walk you through everything you need to know to get started.

If you're still having trouble, or if you're trying the harder difficulties for the first time and finding them overwhelming, take a look at our ArcenWiki or our online Video Tutorials.

If you still have questions, ask away right here on the forums!  Arcen Games staff, or often other AI War players, are always happy to help newcomers.

Admiral:
Having just completed my first, easy, multiplayer game (2 players, 40 planets, level 3, 4 Easier Random AIs), I'm about to embark on a 2-player, 80-planet, level 7 Moderate Random AI game. (I did do the tutorials and watched the video, but some things take a while to sink in at my age.)

Apparently the strategy I used in this game (take everything, kill everything until you kill the home planets) doesn't work in the "real world." I ended up with a (1.008-style) AI modifier close to 600.

From reading the boards, I have surmised a typical advanced game strategy. I would appreciate it if the community would let me know if this is a reasonable strategy to form the basis of a "normal" difficulty and size game?

Key idea: Leave AI Modifier as low as possible while clearing more and more of the universe's planets of enemy ships.


* Take only certain planets. Planets that have either a strategic value (e.g., choke-points) or material value (e.g., lots of metal or crystal) should be taken. All others should be cleared and have the command posts and warp gates left alive.
* Destroy only certain warp gates. I'm not really sure what the point of the warp gates are (despite the intermediate tutorial) but apparently enemies attack periodically through them. Leave them be except on strategically important planets, even though some attacks may come through. (This suggests a ship/station idea, "warp suppressor" which prevents a warp gate from operating or reduces its capability.)
* Link up with ally. Take the planets between your home bases so as to provide a solid, safe base of operations.
* Build defenses. The enemy is going to send waves at you regularly, so make sure you pay attention to the wave alerts and have appropriate defenses in place. (What do the wave alerts mean exactly? Do all attacks only come in waves, or can enemies just move through wormholes like normal?)
* Have a "mobile front." Don't build everything in your home planet and then send it to the front; the world's too large. Periodically pack up and move so your supply lines are shorter.
* Research. Raid planets for research and carefully consider what you and your partner research. (Giving ships to partner: What are limits on that? What if they can't build them themselves?)
* What else?
Thanks!

Cheers...

x4000:
Just going to respond to the places where I have a difference of opinion.  Otherwise, the above is great!


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Key idea: Leave AI Modifier as low as possible while clearing more and more of the universe's planets of enemy ships.
--- End quote ---

I wouldn't try to clear every planet of enemy ships.  Just whatever is relevant for your goals.  Otherwise, even if you leave the warp gates and command stations alone, and the AI Progress stays nice and low, you'll be in for an incredibly long game with 80 planets or more, when it doesn't really need to be that way.  More on this below.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---
* Take only certain planets. Planets that have either a strategic value (e.g., choke-points) or material value (e.g., lots of metal or crystal) should be taken. All others should be cleared and have the command posts and warp gates left alive.
--- End quote ---

Yes, just take whatever is of strategic/material relevance (material value, choke points, as well as data centers, advanced research stations, advanced factories, or other special ships that you find that are desirable to capture).  For the others, there's not point in clearing them -- they will rebuild themselves over time if you leave the command posts there.  That's what lets them do the reinforcements to the planet (those, and special forces command posts).



--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---
* Destroy only certain warp gates. I'm not really sure what the point of the warp gates are (despite the intermediate tutorial) but apparently enemies attack periodically through them. Leave them be except on strategically important planets, even though some attacks may come through. (This suggests a ship/station idea, "warp suppressor" which prevents a warp gate from operating or reduces its capability.)
--- End quote ---

Warp Gates let the AI warp waves into your planets (either the planet of the warp gate, or the neighboring planets).  You'll want to destroy most warp gates near your planets, when possible, and warp gates that are not near your planets have no relevance to you, so are probably a good thing to skip.  There's some more info in the wiki about this:  http://arcengames.com/mediawiki/index.php?title=AI_War_-_Gate_Raids

Also, this in general about priorities:  http://arcengames.com/mediawiki/index.php?title=AI_War_-_Starting_Out

The primary way to prevent warps against your planets is by destroying the warp gates, which comes at an AI Progress cost.  If you want another way, there is indeed already a Warp Jammer style of Command Station (you can unlock it, then build it with your colony ships).  That comes with a steep metal/crystal cost, though, and you only get a few of them per player.  All in all, an AI Progress of 600 is not bad for the end of an 80-planet game, though it is a bit on the high side for a 40-planet game.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---
* Build defenses. The enemy is going to send waves at you regularly, so make sure you pay attention to the wave alerts and have appropriate defenses in place. (What do the wave alerts mean exactly? Do all attacks only come in waves, or can enemies just move through wormholes like normal?)
--- End quote ---

AI players can come through like normal, if you destroy the ships they are guarding, or if they are just patrolling around.  However, the primary way they attack you is with the waves, which are new ships coming in from outside the galaxy.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---
* Research. Raid planets for research and carefully consider what you and your partner research. (Giving ships to partner: What are limits on that? What if they can't build them themselves?)
--- End quote ---

You can't give a ship to your partner if they haven't unlocked that ship yet, or if giving them the ship would put them over their ship cap.  Otherwise there are just far too many avenues for abuse, as my alpha players demonstrated for me. :)

Revenantus:

--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Take only certain planets. Planets that have either a strategic value (e.g., choke-points) or material value (e.g., lots of metal or crystal) should be taken. All others should be cleared and have the command posts and warp gates left alive.

--- End quote ---

This is absolutely true. As destroying Command Stations and Warp Gates, along with a few other structures, increases the AI progress level, a planet is only worth conquering if the benefits gained from such outweigh the increased strength of the AI. As you've already suggested, the main factors affecting a planet's value are;


* Economic Value, measured by the number of resource points a planet contains. It is also worth considering the metal:crystal ratio of a world if your resource gathering/demand is somewhat biased.
* Position. Is the planet an effective base for launching future scouting missions? Would holding the planet form an effective buffer between you and the AI, reducing your number of fronts and allowing you to concentrate your defensive forces in a smaller area? Would taking the planet bring you closer to the AI's homeworld? Does the planet hold logistical benefits in linking areas of your domain together? Additionally, you must consider how difficult it will be to defend the planet once it is taken, does the planet contain links to a large number of hostile worlds?
* Special Structures. The presence of Advanced Research Stations and Advanced Factories immensely increases the value of a world. Although a double edged sword, the presence of Ion Cannons can increase the value of a world in many cases, given that they can make defense of the system considerably easier if captured.
Unfortunately, there is not always a clear choice of world to attack, in these cases you just have to accept the lesser of the evils.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Destroy only certain warp gates. I'm not really sure what the point of the warp gates are (despite the intermediate tutorial) but apparently enemies attack periodically through them. Leave them be except on strategically important planets, even though some attacks may come through. (This suggests a ship/station idea, "warp suppressor" which prevents a warp gate from operating or reduces its capability.)

--- End quote ---

If the AI has a warp gate in a system adjacent to one you control then it can warp ships into your system via the corresponding wormhole. Consequently, it is often worth launching Gate Raids, in which you specifically target the Warp Gate on the AI world and then leave upon its destruction. This is an important reason for considering how many adjacent AI worlds a system has, so you can determine how much you would have to raise the AI progress level in order to make the system secure. Bear in mind that destroying a Warp Gate does not mean that no ships whatsoever will come through (especially against Special Forces type AIs) and so you will likely still need some form of defense at the wormhole.

You may also choose to launch raids against Special Forces Command Posts and Astro-Train Stations.

A strategy that can sometimes work well is to only leave a single adjacent AI world with a Warp Gate. This way you will always know exactly where the AI's warp waves will come in. A minor caveat on difficulty 7 and up - destroying all adjacent Warp Gates will cause the AI to warp it's waves into one of its own systems, and then fly them to attack your worlds. Since the AI is not warping ships directly into one of your systems you will not receive any warning that the wave is on its way.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Link up with ally. Take the planets between your home bases so as to provide a solid, safe base of operations.

--- End quote ---

This is beneficial in that it generally reduces your number of fronts, and allows your forces to support each other. Good tip.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Build defenses. The enemy is going to send waves at you regularly, so make sure you pay attention to the wave alerts and have appropriate defenses in place. (What do the wave alerts mean exactly? Do all attacks only come in waves, or can enemies just move through wormholes like normal?)

--- End quote ---

AI ships can, and will, come through wormholes regularly, even when not being warped in via Warp Gates. For this reason I generally place a few turrets on each wormhole, including a couple of tractor beam turrets. Placing a few ships in free-roaming defender mode (V+Right Click) can also be very handy in this regard (Once we get our Garrison Commands this will be even easier :D). Remember that as the AI progress level increases your defenses will have to be upgraded accordingly - planets that were safe at the beginning of the game may begin to falter under the pressure. Consider building Harvester Exo-Shields on worlds that are under sustained attack.

Placing engineers in systems will help to maintain your turrets, particularly from those evil Astro-Trains on the higher difficulties. If you place the engineers into free-roaming/attack-move mode they can maintain a number of wormholes by flying between them, however, flying between wormholes takes time and it may be necessary to deploy multiple engineers. Investing in the Tech II teleporting engineers eliminates this problem. Engineers cannot repair themselves so they function synergistically in pairs. You could also invest in counter-negative energy turrets to protect against Astro-Trains.


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Have a "mobile front." Don't build everything in your home planet and then send it to the front; the world's too large. Periodically pack up and move so your supply lines are shorter.

--- End quote ---

Placing space docks in multiple locations is useful for emergency defense. I'll quote a post I made in a previous thread that addresses a couple of the issues here - any criticisms you may have of the argument would be appreciated.


--- Quote ---My argument against your engineering shuffling strategy is as follows;

First, let's assume that I have invested in 3 engineers and I am accelerating ship construction at a dock in system A. I now want to move production forward a single system to system B. I therefore construct a space dock in system B and start flying my engineers there. My engineers reach the system and start accelerating the second dock. The problem is my engineers have been doing nothing useful in the time they have been flying to system B. If I had decided to continue producing ships using my engineers in system A and then flying the ships to system B, there would now be more ships in system B. No matter how long I continue to produce ships in system B the total number produced will always lag behind what I could of had by continuing production in system A. The point here is that engineer travel time is detrimental to production. Suggesting that the engineers could just stay in the one place is again a net loss as using the raw resources to directly construct more space docks would have been more efficient in that case.

Consider the alternative strategy of placing a space dock in every system you control (or even just building all of them in your home system, although this slightly dubious, as I will come to in a moment), it follows from the above argument that constantly constructing ships at each dock and ordering them to your front line will result in more ships than could be obtained by shuffling engineers forward. Having a space dock in each system is also useful for defensive purposes should a scenario requiring emergency reinforcements arise. Even if I hit peak production capacity using this method I will still have achieved it more efficiently than if I'd used engineers for assistance - a net gain.

The argument against this building strategy is the reduced level of reactivity, in that it is harder to quickly change your ship mix, as the resulting changes from editing dock production schemes will not come into effect immediately as a number of your ships are having to travel to reach you. Your reactivity is reduced more significantly the further your docks are from your front line, which is often a long way from your home world, hence massing all production there is questionable. Fortunately the effects of this reduced reactivity can be greatly mitigated by good planning.

Now, I will almost always have an engineer on a planet that is adjacent to a hostile world for maintenance purposes, and if he is not doing anything else then he can assist the docks - this is useful but does not affect the points above. Constructing lots of engineers early and dropping them off one by one on planets ties up a large number of resources early that could have been used for military ships, and the huge amount of resulting engineer travel time is very damaging. Also I often have engineers with the fleet to repair ships, these can also be used to accelerate docks when not doing anything else but again don't affect the points above because any engineer that were to leave the fleet to become a production assistant would require replacing, which would cost more than a space dock etc...

By the time I reach the end-game the cost of constructing a few extra space docks near the AI home world to boost the fleet is negligible next to the cost of all those unused engineer hours caused by shuffling them forwards.

--- End quote ---


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---Research. Raid planets for research and carefully consider what you and your partner research. (Giving ships to partner: What are limits on that? What if they can't build them themselves?)

--- End quote ---

Knowledge is by far the most valuable and scarce resource. 'Guerilla Research' is a highly effective tactic. Bear in mind when doing this that Mk I Science Labs are not immune to Ion Cannons, whilst their Mk II counterparts are. The Mk III Forcefield excels in this area, although it is extremely expensive. While it may seem obvious, be extremely careful what you spend your Knowledge points on. Investing in something that you don't really need and then realizing that those upgraded bombers would have been pretty darn useful right about now is a pain (trust me, I know).


--- Quote from: Admiral on July 04, 2009, 08:46:16 am ---What else?

--- End quote ---

Be wary of targeting the AI Command Station in a system too early - it can occasionally be a good way of drawing all the AI ships towards you for quick destruction, moreso early on, but when it goes wrong it goes really wrong. As you may already know, when the AI Command Station is destroyed there is a 50% chance that the AI ships guarding command posts will desert and attack you. Where possible, they will travel through the wormhole network and wreak havoc on your worlds. In a particularly powerful AI system this can spell doom. There is also a 50% chance that AI ships will desert their posts when your own Command Station completes construction.

Become familiar with the relative strengths of each ship type. Effective positioning and targeting will greatly minimize your losses. A good example of this is ordering groups of fighters to fly off and attack AI cruisers before they come into range of your main fleet, thereby protecting ships such as bombers. This tactic is especially important when the range of the relatively strong ship type is shorter than the weaker one - if the fighters were to stay cocooned in the protection of the main fleet they may never come into range of the cruisers that are attacking your flanks.

Where possible, capture Ion Cannons rather than destroying them.

Build forcefields over Advanced Factories, having these destroyed in bombing raids is dire.

Admiral:
Very interesting post, Revenantus, thank you. Makes me wish the Wiki was open to everyone so we could start building a set of "player strategy" pages. (I'll leave it as an exercise for the Moderator on how to integrate the SMF and MW user databases for SSO.)


--- Quote from: Revenantus on July 04, 2009, 10:47:28 am ---A minor caveat on difficulty 7 and up - destroying all adjacent Warp Gates will cause the AI to warp it's waves into one of its own systems, and then fly them to attack your worlds.

--- End quote ---

Interesting. So, the waves will come no matter what; they will just go into some (unknown) system and fly to attack me if there are precisely zero warp gates adjacent to all of my planets?



--- Quote from: Revenantus on July 04, 2009, 10:47:28 am ---I'll quote a post I made in a previous thread that addresses a couple of the issues here - any criticisms you may have of the argument would be appreciated.


--- Quote ---My argument against your engineering shuffling strategy is as follows;

First, let's assume that I have invested in 3 engineers and I am accelerating ship construction at a dock in system A. I now want to move production forward a single system to system B. I therefore construct a space dock in system B and start flying my engineers there.

--- End quote ---

--- End quote ---

The main thing that I don't understand here is why do you fly the engineers there at all? Just destroy them and rebuild them in the new place. That will happen very quickly, but cost 90% of the resources to build them.

My "forward mobile base" I envisioned as perhaps 6-18 engineers and 6-12 constructors. When I want to move them, I'd just destroy the old ones and build the new ones in situ. This way would lose only a few minutes of construction and cost a bunch of resources.

Starships, given that the later ones take upwards of an hour to build, would be built on a safe, well-defended planet and then flown to their destination.


--- Quote from: Revenantus on July 04, 2009, 10:47:28 am ---Become familiar with the relative strengths of each ship type. Effective positioning and targeting will greatly minimize your losses. A good example of this is ordering groups of fighters to fly off and attack AI cruisers before they come into range of your main fleet, thereby protecting ships such as bombers. This tactic is especially important when the range of the relatively strong ship type is shorter than the weaker one - if the fighters were to stay cocooned in the protection of the main fleet they may never come into range of the cruisers that are attacking your flanks.

--- End quote ---

This is a great tactic. I sometimes do this as it is, although not with such a fine grain understanding of the strengths and weaknesses. Usually I'm sending the fastest ships off against something I really need to get killed right now.

Thanks, and cheers!

Navigation

[0] Message Index

[#] Next page

Go to full version