AI War:Micromanagement

Why Doesn't The Game Automate Scouting or Reactor Management?

Q: I can derive a lot of metal/crystal savings by closely monitoring and enabling/disabling energy reactors. Why doesn't the game just automate this for me, disabling whatever reactors are not needed at the moment, and then enabling them as my energy needs grow?

Note:	In the current game version, the energy system has changed and this article no longer applies to game play.
	This article has been left intact so that the developers thoughts on game design are still available.

Response from Chris Park, AI War's Lead Designer:

My thoughts on the reactor auto-management are as follows:

• 1. The game is built around expecting players to build enough reactors to maintain all of their ships.
• 2. Some players like to micromanage the reactors to gain an added advantage. This is fine, but I'm not going to go out of my way to make that easier; they are deriving an advantage out of this that is not really intended.
• 3. Sometimes when things are going really poorly and you just lost a bunch of ships, you need to turn off excess reactors in order to save on the metal/crystal costs. This is expected, and is pretty infrequent, and we recently made it really easy to do with the shift-clicking the reactor entries in the quick-button menu at the bottom of the screen.
• 4. Therefore, I don't see a need for a reactor automation node at all. The only thing that it would do is help all players derive an advantage from auto-microing their reactors, in which case I would then have to rebalance the game around that, in which case you're back to square one of it not mattering at all.

Things like this are always an arms race. When you give the AI a new ability, the game gets harder. When you give the player one, the game gets easier. Etc. And in the middle are various micro-intensive activities that players can optionally do in order to gain some sort of specific advantage if they are so inclined. Keyword there is optional. These are not really intended strategies that the game expects players to use in order to win, but rather are sideways solutions that players work out in order to maximize their effectiveness given the game rules (at a cost of their time and attention to the micro).

Examples include:

• 1. Micromanaging scouts to get really far with low-level scouts.
• 2. Micromanaging unit tactics and positioning in battle.
• 3. Micromanaging reactors.
• 4. Micromanaging low-power/full-power ships to minimize energy footprint.
• 5. Micromanaging engineer positioning and usage so as to reduce the number of engineers needed in the galaxy.
• 6. Micromanaging the positions of shield and munitions boosters to provide maximum coverage.

And so on. These are things that I have no intention of making easier, because they are entirely optional (in the sense of not required to win on standard difficulty levels) and things that some players enjoy doing to gain an added advantage. If those become automated and mainstream, there is no longer any advantage to them. In other words, the opportunities for clever sideways solutions to tactical and strategic problems goes down, and down by a lot -- which is bad. Under normal play, none of the above are required (though some degree of tactical management of battles is always a good idea).

However, in extreme circumstances where a campaign would otherwise be a loss, players can either opt to A) concede defeat, since they have lost; B) get really tricky and do all sorts of micro-intensive activities to claw their way back in. This is very much like regular warfare, and most players will simply opt for option A. Which is fine. But for those players with a certain personality, having the ability to choose option B is the difference between a shallow strategy game and a robust one. Given that I am often in group B myself, that's not something I'm going to sacrifice. Given my above notes about the game being an arms race of difficulty between the AI and the humans, it wouldn't accomplish anything except to make games more homogenized -- again, not a goal at all.

By contrast, reason I previously automated the building of metal and crystal harvesters is because there is no strategy to them at all, it is simply a "click to replace" type of activity that is short and pointless, something that is a time-tax more than anything else. Removing the need to build them doesn't affect overall strategy at all. I don't see the above list of things as being at all the same, because there are plenty of alternatives to all of the above:

• 1. Use larger groups of scouts, or higher-level scouts.
• 2. Build large numbers of ships, build really effective ship mixes, use Mobile Repair Stations, etc. (And, to an extent, tactical micro as an option is a big part of the genre).
• 3. Take more planets in order to increase your metal/crystal income, and thus reduce the load of the energy reactors on your economy; or, find a Zenith Power Generator or similar (in the expansion, anyway).
• 4. Same deal as #3, basically.
• 5. Same deal as #3 and #4.
• 6. Simply throw them into your group and they will provide a benefit commensurate with what the AI typically gets out of them.

AI War:AI War

Why The Emphasis On Taking Many Planets?

Q: Following the above topic, and other comments made elsewhere by Park, it seems clear that there is an emphasis on not having players take very low numbers of planets in a campaign (such as 6 out of 80 planets rather than the "expected" 20-30 out of 80 planets). Why is that?

Response from Chris Park, AI War's Lead Designer:

Continuing the discussion from the topic above: In the end, what the automation of energy reactors boils down to is that some players want to capture as few planets as possible, and thus make the game easier via having a lower AI Progress. That's fine, to a point, but again that's not something I'm going out of my way to make easier -- rather, I've explicitly made that more difficult by adding the concepts of Supply and Reactor Efficiency. The only use I see for reactor auto-management is to maximize your economy while minimizing the number of planets you hold. If you're taking option B from above and clawing your way back from defeat, then you'll appreciate the ability to tune your energy reactors by hand to get the desired result. If you're trying to play a super-low-AI-Progress game, then you're doing it to yourself by fighting the normal game flow. That's certainly allowed, but if you don't like the micro that causes, then don't try to circumvent the game by not taking enough resources to support yourself.

That might seem harsh, but we're back again to the arms race: if taking a low number of planets is easy to do and convenient, then everyone should be doing that in order to keep AI Progress low, right? If that's the case, then there's never a reason to take many planets at all, and the game is artificially easier and in need of rebalancement. My take is that taking very few planets is less fun, in a general sense, because you get to do less capturing and you have fewer options overall. Also, it is harder for the AI to be very interestingly effective against you, because there are so many fewer ingress points. So you wind up having less fun of a time, and a less interesting AI opponent; sounds like a bad deal to me.

But, given the current game design, some players delight in having a super low planet count simply because of the fact that this is a unique, difficult, off-the-beaten-track way to play, and it helps them be experts in an unusual specialty. This, again, is fine, and something I like to support as much as possible. So as I've added mechanics to the game such as Supply and Reactor Efficiency, I have always been balancing the needs of the standard game against at least making the low-planet-count games possible, if not easy. That sort of style of play is not something that I personally enjoy, but, since some others do, it's not something I want to kill because it does add variety and is fun for some. The problem comes when a tactic like that is too much of an exploit (which it once was), or when the players who enjoy that style of play start recommending it to all newbies -- who then take it at face value that this is the normal and expected way to play, and then come away feeling like that is a frustrating and micro-intensive way to play (and it is).

Bottom line there is that if you're the sort that enjoys a micro-intensive challenge specifically to thwart the normal play conventions of a game, then there are things you can do such as having a super low planet count. That's a certain kind of person who enjoys that sort of metagame, and I wish them well with it so long as it doesn't turn exploitative to the point where it is legitimately the one best way to play (currently it is not). For anyone not fitting the above description, they should generally instead focus on playing the game the way it was designed, which is much more fun for most people.

AI War:AI War