« Unevenly Distributed: Los Angeles | Main | Occidentalism's Cost: War? »

August 12, 2010



One of the best series of games out today is the "Total War" series. Napoleon: Total War is one of the newest installments of this series, and to which I was rather addicted to over the last two months.

The AI sucks--Playing as England, I took Paris with a single army in the Winter of 1806. When I was able to do that, on the hardest setting, I basically stopped playing. But, the units are historically based, with ship designs guided by the National Maritime Museum, London.

Even the best warfare/statecraft based games are nothing like krigspiel. To make the game harder or easier, generally a greater strategic depth is not given to the AI, rather the ability to raise armies or keep moral up in the armies are affected--but the game never gets any smarter.

The biggest loosing factor in my mind regarding Napoleon: Total War is that naval combat is precise. Move a fleet from the Chatham Docks to Trafalgar and it will be done in three turns, no matter what.

If this damned technological singularity would happen already, I think games might actually get interesting.

Joseph Fouche

I always hated the Starcraft AI, especially when it was controlling your own units. How many times did I shout, "No you idiots, stay in formation!" when a stray Zerg unit would draw a posse of advancing marines and tanks after it while I was commanding units on some other part of the field. When you're AI's default move is a mindless and disorganized frontal charge into hostile fire, you're screwed.


The stimpacks were also the only way to make the marines really useful.


Also Lucien, the whole singularity thing is lulz in and of itself.


I ran across some space-based strategy game that would log the moves you made in the game and add them to a server. The server would then compare that to what other players were doing, and adapt the games AI to your strategy. It would also find the most successful strategies other players used against it and in turn, use those against you.

I didn't have the computer or the money for the game at that time. But, I am rather surprised not to have seen more games go in this direction. The adaptive algorithms are there, they can be licensed. The Steam network exists and could easily compile the information. But, it has not been implemented. A shame.


That's interesting. Most of my favorite videogames tend to be RPGs like the Final Fantasy series or survival horror like Silent Hill or Resident Evil. I do play a few turn-based strategy games (some Civil War sims) though.


The Air Force actually developed an airpower strategy game based on the StarCraft engine that I was 'trained' on for one of my courses. Basically, they had gotten their hands on a developmental version of the game and reskinned all the vehicles to look like actually platforms (F-15, B-2, etc) and tailored all the functions (endurance, HP, weapons range, damage, etc).

All of that would have just been superficial, but they also tailored the system where multiple players could be on the same side, basically replicating a COCOM air staff (Defensive Air, Offensive Air, Logistics, Base Defense, etc). Each of these functions would have 2-3 people. On top of that, they placed a 'Commander' function. The interesting thing about this role was that he couldn't actually control forces. He had to work through subordinates (just like real combat). Overall, the 'teams would be ~10-12 people working toward a common goal.

They did a pretty good job of creating dependencies in the system. Bombers couldn't make multiple runs without coming back for munition resupply and couldn't loiter for too long without refueling, so if your logistics guy sucked you could run out of fuel.

Overall, it was an interesting concept and fairly well executed. It definitely increased the chaos of the system and went beyond the all-powerful General model that is normal for these kinds of games.


Really? That's pretty awesome.

Laura Donovan

Your roommate sounds like a pistol. I'm glad you moved out. Excellent post.


It's a good point that SC2 doesn't really simulate a pure struggle for strategic and tactical advantage. But it's also not a surprise, because it's not really supposed to. It's not that kind of game.

The reliance on micromanagement in Starcraft (and Starcraft 2) is a design decision, not a flaw-- or rather, only a flaw as far as you disagree with it as a stylistic choice.

In the 10 years since Brood War, the genre has done a lot of innovating, including some great variants that did indeed get rid of a lot of the micromanagement in favor of squad-based command, point control, and terrain features.

SC2 decided to ignore it all and go back to micro, which means they definitely did it on purpose. The reason is that SC isn't chess. It's not speed chess. It's speed chess while log rolling in an archery range during a fireworks show under a strobe light. The entirety of the challenge is making anything resembling a decent strategic decision in the face of all the distractions.

There are some elements (for those familiar, the Protoss Chronoboost, the Terran MULE, and the Zerg Inject Larva) that make no sense except as a deliberate test of one's multitasking-- a small-scale strategic decision for the player to remember and execute periodically.

So, you're absolutely right that SC and SC2 aren't good war simulators. If you don't like the dynamic, they also might not be enjoyable games. But the focus on micromanagement isn't a failure of imagination or a failure to accurately simulate being a battlefield commander. It's a success at trying to overload the gamer's capacity to execute his strategy.


I don't see where I say that it was a failure of imagination or a massive flaw. My point was that 19th century kriegspiel users would not be impressed by it as a war simulator.


The game that comes to my mind in this is Gratuitous Space Battles. You design your ships with whatever various weapons, armor, and engineering functions you want, arrange the fleet, give each ship orders, in terms of what types of enemy ships to engage first, at what range, holding formation or not, etc. And then you hit fight and your job is done. All you get to do during the battle itself is watch the pretty alpha-blended laser beams and explosions.

It's definitely a different experience. And, IMHO, a fun one. Especially for the price. I recommend it to anybody (Google it).

Ginger Yellow

As mentioned above, the extreme micromanagement in SC2 is a feature, not a bug. There are, however, many tactical strategy games that do incorporate delegation, to greater or lesser degrees and with varying levels of success.

Staying within the conventional RTS genre, Men Of War (great title, eh?) gives its units considerable autonomy, while still allowing lots of micromanagement (down to the inventory of individual soldiers). Left to their own devices, units will seek cover, attack the enemy and so on. This means it can present battles on a scale rarely seen in RTS, as the player can focus on a small part of the battlefield and more or less rely on the AI to be sensible with the rest of the units. They're not going to launch a coordinated offensive on their own or anything like that, but neither are they going to stand there uselessly.

Moving more into wargame territory, the Take Command series (discussed at length in this podcast: http://www.quartertothree.com/game-talk/showthread.php?t=60797) models delegation quite comprehensively. You have to rely on your junior officers to carry out your commands, and they may not follow your orders.

In the grand strategy world, WWII sim Hearts of Iron 3 allows you to delegate entire theatres to the AI. I haven't played it though, so I can't really say how it works out.


I've played Hearts of Iron 3, very good stuff!

Mike HEss

My Starcraft is pretty rusty, but I think that if you watch those Korean grandmasters and listen to the commentary, there's a subtle battle between "micro" and "macro" management between two Starcraft players. "Micro" players control their units obsessively, and use huge numbers of commands to finesse their marines around incoming missiles or tentacles, increasing their effectiveness. "macro" players certainly still issue commands to their units, but spend their time maximizing resource extraction, base expansion and unit production. Their units tend to be less effective in combat because their "micro" isn't as good, but they bring more troops because their "macro" is better.

In good matches between a "micro" and a "macro" player, the "micro" master has a certain number of minutes to win the match with superior unit tactics before the "macro" player overwhelms the map with units and structures, and wins by weight of numbers.

Maybe, watching that, those Germans might be more impressed.

Ralph H.

Nice post & interesting thread. My wargaming these days is limited to occasional bouts of The Operational Art of War, which I enjoy but generally wind up on the losing side to the AI (e.g., in the Crusader or Kasserine scenarios I've never beaten computer-Rommel). I would greatly welcome a similar game where delegation was built-in, & perhaps one could vary the competence/initiative levels of various commanders in the hierarchy & issue general orders & objectives.

JimmySky, back at USAF Squadron Officers' School in the mid-70s I participated in a game like the one you described, with our section of a dozen officers simulating a battle staff in a humanitarian intervention/limited war scenario. By virtue of having spent a couple of months in "Blue Chip," the 7th AF command post at MACV, I was designated the CO. The only key combat decision I recall making was to cancel an intercept of an unknown aircraft (proved to be the correct correct decision). A very realistic exercise for all concerned, IMO.


Ralph, that's a fun game too. The Kasserine scenario is rather rough.

Mike, that's still primarily a battle command perspective--and the friendly AI are still automatons. Other war games, like the ones mentioned in this thread, have a larger view.

online jobs

why are you always so much blame new technology, I do not understand ochem you here so many have written, you'd better tell you about a new discovery or something like

The comments to this entry are closed.