So I'll guess there's only one "right way" to play this game?

[Buttons840]

The problem is, even in Chess you can do whatever you want as long as you obey the rules of the games. This doesn't seem to be the case in AI War. I can try alot of strategies, and then getting punished vicious for doing so. From what I've read it all comes down to engage the AI as less as possible skipping as many planets as possible and maybe neutering some.

To both of these I have the same response, and it goes along with what Chris said. Sure, in chess you CAN do anything within the bounds of the rules, but, there are any number of strategies that if you try them against a good player, you'll be crushed, and they just aren't viable strategies. They only time you have any chance of winning against them is to play on a lower level, or against a weaker chess player, as the case may be.

Well said tigersfan.  Deleth, you said, "I can try a lot of strategies," and indeed you can; you can do whatever you want within the rules of the game.  But again, this doesn't mean every possible strategy is effective.

Perhaps one of your frustrations is you've reached a stalemate where the neither side can make progress, and the game just kind of stalls in this state for many hours.  It can be frustrating.  I consider this a weakness of the game and there has been some ideas from the community on how to fix this.

I've played on 30 planet maps and completed games in 4 to 6 hours at 7 difficulty (even against harder AI).  This is how I like to play.  You might like playing with fewer planets, or there are many other game variables you can experiment with.  Considering starting a new game and adjust some of the settings.

[keith.lamothe]

As far as different strategies:

1) Take all the planets.  As Chris said, this can take waaaay longer than usual and either lower difficulty or very smart play and/or favorable map (e.g. Snake).  Basically the AI will continue to get more and more powerful to the point where you can feel pretty "railroaded" : either have a completely impenetrable defensive chokepoint or die.

2) Take 15-25 planets, accepting moderate spreading-out (skipping some planets).  This seems to be the normal range, and the AI gets pretty dangerous by the end but not overwhelmingly so as long as you can defend your homeworld without having to simultaneously guard too many "chokepoints".

3) Take 5-10 planets, accepting major spreading-out.  This can allow extremely-low-AIP games, particularly combined with long-range "sniping" of data centers and co-processors (the super terminal can be used to great advantage in any of these, of course).  So much so that we introduced the Core Shield Generator mechanic to increase the minimum number of worlds that had to be captured before you can win.  That can be turned off, though, which was a relief to several players who saw that as more of a "railroading".  Even with them on, there's a lot of variation in other things.

4?) And that's just the question of how many planets you take; then there's strategic questions of what you spend your knowledge on, which planets you prioritize, and many tactical considerations on how you actually execute a given attack or defense, etc.

How you deal with spreading out is one of the points where strategies vary.  I tend to play games where the AI launches a lot of exo-waves, so I just get used to the idea that any of my "outlying colonies" may get killed off on very short notice.  I sometimes long for an "auto-re-colonize" feature because I keep rebuilding those planets, but it gets done and I'm able to make more progress, etc.  And generally I set it up so that any AI attack gets "blunted" a little bit by each such planet it takes so that when it hits my serious defensive perimeter around my core systems there's much less of a threat left.  Other players adopt a strategy much more focused on never losing planets.  Other players never allow spreading out, though that tends to run high AIP.

Another point of variation is how to deal with the planets you don't take, and how to "manage" AI reinforcements.  If you're going to skip a planet you can "neuter" them of guard posts to keep down reinforcements, at the cost of ship-time.  That way any "reinforcement points" the AI spends on that world don't produce as many ships, though it can still hit the "population limit" for that world and start producing barracks and then carriers (but it's still easier to clean out).  A more advanced approach is to keep some out-of-the-way cluster(s) of AI worlds on permanent alert to "siphon off" their reinforcement points to a part of the galaxy you aren't planning to do much in, and that tends to prevent the massive threat buildups and carrier buildups, etc, because the AI isn't generally hitting the "population limit" on individual planets.  More generally, later on it can be good to have a lot of AI worlds on alert so the AI doesn't "chokepoint" you, or massively over-pile a world and produce lots of carriers.  And, of course, one strategy on this point is to just not bother about AI reinforcements and use all the time saved thereby to go for a quicker win or some other advantage to make the win more likely.

A word about carriers: the AI needs a few counters to "impenetrable player chokepoint syndrome", otherwise a player can set up a nearly un-loseable situation simply by picking a map that lets them set up a one-world chokepoint and dump every turret/ff/etc in the galaxy on it.  Carriers are one of those counters.  Some chokepoints can melt even carriers, of course, but it tends to involve strategic opportunity costs from all the knowledge dumped into defensive types.  Or superweapons, which have their own costs/tradeoffs.


Anyway, the main problem in your current game seems to be that you've gotten AIP higher than you're equipped to handle smoothly (armor ships are perfectly viable bonus type, by the way, but their strengths may not gel with what you're trying to use them for) .  You said you're still making progress and you might still win, but basically you've fully woken up the bear and it's not happy   There are ways to face a grizzly hand-to-hand and win, and if you want to play the game that way you have to pick a strategy suited to that.  Other strategies depend on killing it before it's fully aware of you.

As Cyborg said, you might enjoy the Fallen Spire campaign better than the "normal" game.  Worst to worst, it's a very different way to play and win the game   Though you don't have to pursue the campaign to its "logical conclusion", and can still win the normal way, etc.

[Lee]

It also sounds like you've picked more than one starting planet, which can up the difficulty considerably.

[BobTheJanitor]

As an aside, on the subject of overwhelming odds. I read that and it got me thinking: Most games that represent themselves as this type of game are really lying to you. It appears to be overwhelming odds, one man against an army, etc. But in reality the one man is a super man, who can take thousands of bullets and has some sort of mystical regenerating health or is able to instantly heal wounds by touching health kits which are clearly made of pure magic (and are lying on every other end table). He can also carry around 500 pounds of heavy military hardware and ammo, without even slowing down his ability to sprint everywhere, and he appears to hold all this gear in some other dimension, from which he can pull it at a moment's notice. Now if such a person existed in real life, clearly to have him go up against an army could only be considered overwhelming odds from the point of view of the army. They would be much wiser to lay down their arms and seek a truce as soon as they heard he was in town.

AI War is perhaps the best game at really representing overwhelming odds. The enemy has an endless stream of ships to draw on, and holds every planet in the galaxy. The only reason the remnant of mankind survives up to the start of the game is that they haven't been noticed yet. The whole game is built around the idea of not being noticed and surreptitiously gaining just enough power, but not so much that you trip alarms and get overwhelmed.

And for what it's worth, it seems to me that there is plenty of freedom in how you do this. The only thing you can't do is an extreme conquest approach (well, you can, but it's hardly recommended) and even then, if you're doing that, then you're not really playing the game as designed, you're attempting to play some other game. It's like making up your own variant on the game and attempting to play that. If one wanted to do that, more power to them. But complaining that it's not balanced for your invented game isn't fair. It's like complaining that tennis is too hard because you are trying to hold the racket upside down and hit every ball with the handle.

[c4sc4]

I think the solution here is to turn down the difficulty. If you want to play a massive conquest game, maybe try a difficulty 5 or 6 opponents instead of 7s. Sometimes if you want to play a game in a way different than the optimal win strategy, you have to play at a lower difficulty setting.

For example, in SWAT 4, I can't go in guns blazing killing every suspect in sight. If I were to do that, I would lose, either because I ended up getting shot or because I had too low of a score. If I want to play that way, I can play on the easiest setting and it is a viable strategy. Now if I want to beat the game on the highest difficulty, I have to use less-lethal weapons because if I kill one or two suspects, I will fail. Really, I don't think it is possible to beat the game on the highest difficulty setting if you give your squad mates lethal weapons because they will kill someone.

Also, if you had fun for the first 10 hours of the game, why not just start another one and play for another 10 hours, sometimes losing is fun. I don't think I have every really beaten a game of AI War, I have tons of saves where games are at 10 hours or so. Although this may be caused more by the fact that if I leave a game for a week or two I am completely lost in what is going on and what I was trying to do in the game. Even though I haven't beaten the game, I still had lots of fun playing it.

EDIT: Sometimes a game won't let you play a certain way. It would be like after playing Burnout, going over to Forza Motorsport and being mad at that game for not letting you crash into other cars. Sure you can do it but you are punished severely for doing it.

[Wanderer]

On the subject of carriers, there's actually a reasonable counter to it but you've got to prep your whipping boy/defensive points for it.  Firstly, you'll need to use a 2-system approach, letting the carriers go through to the second system while your first does the anti-ship work.  In the second system, keep a few of the heavy beam cannons around, level 1's usually work fine unless you're getting overloaded via exos, near the gate.  Have these primary target carriers.  This will usually burst the carriers pretty well, and then have the second region mop up the fleet ships.  My second region usually packs a light smattering of turrets for anti-polycrystal and a pair of Fortress 1s.  If a serious wave is punching through I'll get the attack fleet involved for finishing the cleanup.

With that and grav turrets to slow down the rest of the fleet, you're usually pretty defended and you've allowed for a defense in depth scenario while still keeping a viable chokepoint.  This usually works best if you can get a set of systems in a line, but if you're careful about it you can even let the 'rear' system be a single system chokepoint and use a system one out as the whipping boy, just be careful about Warp Gate Guardian respawns.

As to the original posting, I don't find the game to railroad me that badly.  Sure, it's an RTS with some tower-style defenses, but it's an RTS.  Unlike a lot of 4x games, however, instead of just gaining more power for each system you take, the AI is getting more powerful for you doing it.  Instead of just getting vague economic bits on its tech tree and some falsified econ adjustments, though, it's a direct improvement you can understand and track.  If you paused for 3 hours, the computer will never get 'stronger', though it will reinforce more.  This allows for a more controlled aggression increase from the player's perspective.  If you're going to start doing 30+ system grabs, you're in for a heck of a fight.  I usually aim for about 20 or so before I start hammering supply lines into the core worlds.  15 or so in core economic build behind the chokepoint (if I can build one), 5 more or so grabbing and defending high priority targets like ARSs, Core Shield Generators, Fact IVs, and worthy Golems.  The last 5 to 10 should only be really for supply lines to the core worlds to neuter the homeworld defender worlds and then to finally pop the AI Command centers.

That said, I recently did a 120 world completionist on a snake map and my AAR is kicking around here in that forum, so it CAN be done, it's just not a pleasant way to play.

[Professor Paul1290]

I think part of the issue here is that you're playing against two defensive AI types. By doing this I think you might have set yourself up for quite a stalemate.

I usually play random AI and I remember getting similar combinations a few times and they often ended up being very frustrating to play against. Thankfully it hasn't happened too often.

There are some settings, the AI types included, that can affect the pace of a particular game significantly. I prefer to play against random AI and bonus ship types, but even then sometimes you run into bad luck and sometimes the Elder God of Random decides to screw you and spit out just the right combination to cause all sorts of pain and hell for your playing style.
It's an inevitable consequence of the game giving you so many options to tweak the campaign, there are going to be some combinations that are just plain bad for your health and sanity.

[KDR_11k]

Not just defensive AIs but also apparently a map with a single massive chokepoint. And then he went and pissed the AI off. That's trench warfare and it's a brutal slog filled with losses as WW1 has proven. The chokepoint works both ways, the AI has an easier time defending against your attacks as well. If you give it time then the AI will inevitably win (you only grow stronger by conquest, the AI also grows stronger over time). Less chokepointed maps lead to the AI spreading its forces more or getting way more aggressive (it won't attack when the only path into your territory is defended by an overwhelming force but on a wide front there will always be weak points where it will strike).

[sol_ilya]

There is one thing that dramatically changes game flow. Superterminal. In my previus game it was on South in six systems away from my home station. Moreover I first went to opposite direction.  Usually I carefully choose what planet I shall take which lead to situation when almost every my planet except home station can be under attack by AI.  I have got advanced factory, several fabricators. When  at last I scouted superterminal and made border with it I already had 23 planets, 475 AI progress and fights vs AI in 3-4 my  systems at once. In such situation I was very afraid of going for superterminal becouse I heard about a power mark V ships it spawns and I was not sure if I can hold them.  Shall I mention that I did not used it before in my games.
I had wrote report of last my hour of surviving.  http://www.arcengames.com/forums/index.php/topic,9632.0.html
That was a lot of fun game.

Now in my current game I had superterminal bordrering my home system. And I must say it is not so dangerous thing. I placed there only 400 turrents II level and two forstress of I level. I did not sent any of my ships there. Only turrets II level they are cheap in research points. First superterminal spawned level II ships that were easily destroyed, afterwards III, then IV levels. I already got at that time very nice redution of AI progress. When I got 1000 mark V ships I destroyed my station. Then several times built and destroyed the station. In this game I have reduced by datacenters, coprocessors and using superternminal AI progress by 600. And that is completely different game now. When I had also 23 planets in this game I had AI progress 147.  Just compare with 475 AI progress with the same amounts of planets. This game became too easy and less fun that my previus. I am really missing loosing and retaking back my systems. The most fun begins when AI progress reach 400+. Well, superterminal allows you making more conquest type game.

Can you post screenshot of your map? I'd like to see what situation you have.