GalCiv2, SotS, SEV: a 4X Comparison
Review - posted by Jason on Sat 16 December 2006, 01:11:08
After the fiasco of Master of Orion 3, which certainly didn't do any good for the genre of Space Strategy games (or 4X games), three new titles finally appeared this year: Galactic Civilizations 2 (GalCiv2), Sword of the Stars (SotS), and Space Empires V (SEV). Among these, two are obviously sequels and this ratio can easily be taken as the magic formula which is currently the trend in gaming industry: No more than one third true innovation.
The aim of this comparative review is to give anyone interested in such games an insight into what are the weaknesses and strengths of each so that he can vote with his money. However, I think that anyone truly interested in the genre will end up with all three of them, simply because each one is far from being the disappointment MOO3 was.
We start with a short introduction to each of them and then discuss the major aspects in detail by comparing them between the three games.
1. Galactic Civilizations 2
I will begin with Galactic Civilizations 2 because it was the first one to be released this year. First, a short description of the basic gameplay for those not yet familiar with it: GalCiv2 has a very fitting name because it differs least from Sid Meier's Civilization series. It plays on a rectangular gridmap (which is essentially the same as the tile-based map in Civ). Suns and planets each occupy cells on this map and the empty cells in-between can be used to set up space stations or travel between the planets.
Development of colonies (equivalent to cities in Civilization) on planets is done by waiting (for population to grow) and producing buildings on the surface of the planets. This includes a little map where you can choose where to put what, depending on which bonuses there are on a given tile. But other than that, there are no gameplay implications and this feature appears slightly superfluous currently.
Speaking of superfluous features, there is another one in GalCiv2 - the ship building. In this game you can design the your ships using a very neat and easily usable 3D construction program. You can place where to put your guns, shields, wings, engines and everything else. In fact, this part of the game is executed so well that it plays as a game within the game. Unfortunately, apart from determining how many weapons (and other parts with an actual in-game function) you put on that ship, all that designing is completely irrelevant. You could stack up all you beam weapons at the tail of your ship, pointing right *at* your ship or put the engines in front of the ship. It just doesn't matter at all. This is because in GalCiv2 you get no tactical combat other than a cinematic show which only hides a dry formula to compute the winner of the battle. This is, by the way, also the biggest difference and weakness compared to the other two games, SEV and SotS.
So if there is no tactical combat in this game, what else does it have to make up for it? Well, quite a few things. Contrary to the other two games, it does have a working diplomacy, a slightly interesting economic model, and a really good AI. And when I say really good, that's not an exaggeration. The AI doesn't need any cheating at all to beat even a seasoned GalCiv2 player. Although someone really determined will, of course, always beat any game AI, the amount of work to do so is a lot larger than for any other game out there, except professional chess programs. The AI also tries to play slightly human-like: It recognizes masses of ships at it's borders as a threat, it will try to counteract influence attacks, and it will try to sneak in troop transports, just to name a few. There is also different code for the different AIs. This is supposed to reduce the uniformity of each enemy civilization and partly succeeds in doing so. However, while the AI is good, it still is very dull. The 'goodness' of the AI comes from it being very optimized with respect to building up a highly efficient economy. It won't be creative or at least surprising by suddenly changing it's strategy. But this is a general problem with the current understanding of what an AI should be like in a game.
Nevertheless, it is nice to see that the AI is mostly bug-free (by now, as of version 1.4, but it wasn't bad in 1.0 as well). Thus it can never happen that a player begins a game, plays some 50 turns only to see that he has already essentially won, because neither of the other AIs were able to colonize more planets or somesuch. This is something that can easily happen to you in SEV currently.
2. Sword of the Stars
The second game to be briefly introduced is Sword of the Stars because it was released as the second 4X game this year. Contrary to GalCiv2, it plays out on a set of stars, freely positioned in space. Between the stars there is literally nothing except for the fleets later on. The stars are placed in a truly three-dimensional manner and the game offers a camera which can rotate and move around in this space. However, I noticed that most players still tend to play on flattened maps, such as the spiral galaxy. It simply is too confusing to attempt to memorize 150 solar systems which are placed randomly in a real 3D ball. Especially since, except for the stars, there is nothing to help you remember certain areas of the 3D map, such as nebulae, wormholes or any other non-solar-system feature and since the implementation of the camera controls could have been better.
Personally, I think that the biggest innovation SotS offers us is the way space-traveling is implemented. The four races each have very different ways of traveling. The Hivers can build one teleporter in each solar system. The more teleporters they have globally, the more ship-mass they can move through. This means that they usually can move one or two full fleets from one end of the galaxy to the other. This powerful feature is balanced out by the fact that in order to arrive at another system they have to go there at sub-light speed, which can take quite a while. The humans, on the other hand, must use a system of wormholes which connect the stars with each other. So if there is no direct link from A to B, then there is no feasible way for them to get to B. The Tarkans, on the other hand, can move freely to any star, but do so slightly slower than the humans. Incidentally, in the gameplay this is also the only feature which really helps to make the different races feel different. Apart from the visuals of their ships, there is literally nothing else that noticeably helps to feel like I am playing the Hivers currently.
I heard (or read) someone say about SotS that the strategic overlay is only an excuse to put the player into as many as possible tactical battles. This remark has some truth to it. It is obvious that the most effort went into designing and implementing the tactical battles, whereas the strategic gameplay is reduced to researching and seeing to it that the right fleet is at the right spot at the right moment. The economic, cultural or diplomatic aspects of gameplay are either streamlined into two nearly impact-less sliders, or non-existent altogether. Simply put, the strategic gameplay is not much more intricate than one of those very old 'conquest' games where you just move numbers of army points back and forth in your empire.
SotS tries to shine with its tactical combat. Here you are given the possibility to control the warships you previously designed in the hopes of outwitting the enemy. Except that usually 'outwitting' means to bring in the bigger guns - as simple as that. The tactical combat comes with an attempt at Newtonian physics, so a ship sent to fly to a point over there and suddenly ordered to rather fly over here will turn around and begin to accelerate into the new direction, while still continuing to drift into the old direction. While on paper this might sound like a neat idea to someone, in the game this is simply frustrating. Against the AI it also results in players usually employing the 'sitting fortress' strategy: Simply mass your ships right at the starting point into a group and wait for the enemy to come over here. Hopefully not with all ships at once. Of course, this usually works. Even if the enemy has many long-range fire ships and the player does not, it still would send at least those non-long-range ships to be blown apart by your group.
The ships in this game come in only three possible classes: Destroyer, Cruiser and Dreadnought. Each class can take on from about three to nine ships of the next smaller size on it's own. Specifically, the player can design which weapon is placed onto which hardpoint. Also, the ships are constructed out of three parts, such as the engine or the mission part. However, all that is almost in vain. Usually, it's just plain better to put the best weapon into all slots. Some goes with the parts - there are four different engine parts, but you'll always take simply the latest one. The only noticeable difference during the design of your ships is whether you put long-range missiles on your ships or whether you go for short-range beams. Apart from that, whether you put one more torpedo instead of a blaster on the right side of your new dreadnought design simply hasn't any impact on the later tactical battles at all. Additionally, due to the lack of feedback the game gives you during tactical battles, you never know quite for sure whether, for example, taking the ultraviolet laser is better or worse at intercepting enemy missiles than the mass-driver based point defence. The game designers want the player to simply know such things by heart and think that exploring such things without any feedback is fun. There is some truth to that, but it just doesn't seem to quite work here.
3. Space Empires V
The third 4X game released this year is Space Empires V. But before continuing, I should add that 'released' is not quite true. The current state of the game could be more precisely described as 'open beta'.
For those not yet familiar with this game, it plays out on two maps: A map of the galaxy involving from tens to 255 solar systems connected by warp points (the only interstellar traveling here). And the individual hexagonal maps of the solar systems, containing planets, suns, asteroid belts, storms and other such things. And your fleets, of course. This duo of map types feels quite pleasant, as it gives the feeling that there is a lot of stuff to be explored and the resulting effective game space can be very large.
The strongest aspect of SEV is (and has always been in this series) the level of detail. You get to research hundreds of different techs, design your own war ships, fighter, troops, transports, mines and drones. You have to keep track of the supply levels of your attack fleets, lay and sweep mines, deploy drones, build facilities on planets, handle cures or riots on planets and much more. The interface, though cursed by some, at least tries to give you all possible information (very unlike SotS) but needs some more work, especially by introducing fast means to achieve certain oft-needed operations. All that micromanagement is also quite a lot of fun and is clearly quite the other extreme, compared to the overly streamlined SotS. However, with all that fun comes one strong disadvantage: the AI, but I'll come back to that later.
There are several fundamentally different playing experiences possible in this game. You can play a pure diplomatic game here and the detail of diplomatic agreements is so deep that it's easily possible to conclude a deal which forbids you to generate any intelligence points, but at the same time allows immigration of those aliens which make those barely habitable worlds (by your own species) much more useful. Or you can sell that one space station in that god-forsaken system for a new tech. Or you can sign a full subjugation. Or ... the possibilities are almost limited only by your imagination.
You will get an entirely different experience by going the xenophobic turtle way. Building lots of mines, drones and satellites to defend the warp points from any intruders, research up to sphere-world construction, build sphere-worlds around all your stars and then use the newly acquired humongous industrial power to squash the other civilization, who most probably seem like petty barbarians by that time.
Or you can go the sneaky way, investing a lot of resources into intelligence. Once someone decides to attack you, he will have to deal with the problem that all his fleets, facilities and just about everything else begins to randomly and mysteriously explode for no apparent reason all of a sudden. Turn by turn until he has nothing left to attack you with.
Alternatively you might try the aggressive way, building lots of ships quickly and overrunning those pesky aliens before they have time to lay mines or do other nasty things.
All this really is possible in SEV and the only real problem this game has is AI. As the current state of the art wants it, the AI is designed in a very straight forward manner: It comes with schematics of which types of ships to fill with which kinds of weapons. Once you understand that, you can easily build ships that counter the predefined ones and the AI can't cope with that. Further, the AI comes with a defined set of possible "strategies". It will always play out all of the above-mentioned playing styles at once and thus not excel at any of them. Once you understand that, you can easily go for any of the extremes and the AI, again, cannot cope with that at all. Finally, the AI comes with a predefined set of specific actions to do in the solar systems. It will set up lots of star bases over its planets, fill the wormholes with fleets, mines and fighters and once in a while send an invasion fleet. And again, once you understand this, it's too easy to exploit that. Thus, for example, even if the AI sees you building a huge fleet of minesweepers, it will not be able to counter that by abandoning the mine-laying and building lots of fighters instead. All that unfortunately detracts quite a lot from the replayability of this game. But do not despair! The good news is that the AI is highly modifiable. In fact, the game comes with a complete script compiler and editing GUI. This means that sooner or later, if the sales of the game itself achieve a certain critical mass, good AI mods will begin to appear. A first indication of this is that there are already several mods tinkering with the AI.
4.1. Designing ships
Probably the most simple and fun part of each of these three games is the possibility to design your own ships. Yet, all three manage to implement that feature very differently. The nicest implementation is the one in GalCiv2. Here you get a fully-blown 3D construction set where you put together the pieces that make your ship in LEGO-style. Just better. However, the most useful implementation is clearly the one in SEV. Simply put, the one in GalCiv2 is all nice and beautiful, but totally useless. The only things that matter are the number and type of weapon. A simple spreadsheet could've served the same purpose. In SEV a lot more things matters. While you cannot influence the looks of your ships at all, you do define whether your ship comes with more supplies (because it's a scout), with more ammo (because it's an artillery type of warship), with few guns but lots of armor (because they are supposed to ram anyway) and much more. It's simply a lot more fun in SEV to put your newest researched shields into your battle-proven defender ship design and again decide whether swapping some shields for more armor wouldn't work even better.
In both respects, SotS loses ground. How you design your ships doesn't have a lot of impact as long as you keep putting the newest guns into all slots, nor can you influence the looks of your ships. You may ponder for a while about whether to rely on long-range weapons or short-range weapons, but that's about all of it. Quite disappointing because the design interface looks nice for the first moments.
Now, the ideal solution would be a combination of GalCiv2 and SEV, but currently the ship designer in SEV wins in the long run, because usefulness always wins over beauty in the long run (did I really just say that?!).
4.2. Tactical combat
After having designed your ships, you'll want to see them in action. This is where GalCiv2 disappointed many: It doesn't have any tactical combat at all. It has a cinematic battle-resolve show, which you'll turn off very soon. However, the other two games do not excel either, each coming with their own weaknesses. SotS suffers from, how to put it, strange design decisions. The above-mentioned attempts at Newtonian (and thus realistic) physics don't go well with the utterly unrealistic scale of things: A single dreadnought is bigger than a moon! Also everything about the tactical combat has this strange not-quite-right feeling. Instead of at least a health-bar or ideally a schematic showing the damage status of ships you get - nothing. And the designers are reportedly proud of that. Well, you don't get nothing: If you look very hard, turn up your gamma correction (because the default is way too dark) and zoom in to that ship you are interested in, you may notice that certain parts of the ship look damaged. That's nice. But if you have 8 cruisers and you want to know which one is most damaged, zooming in and looking for 'broken' textures or effects isn't very practical and I found myself simply not trying to find out (and make use) of such information.
The tactical combat in SEV is less realistic than in SotS as it doesn't even attempt to be realistic. However, it is more 'tactical'. Because it's rather simple to control your ships and their status (unless you try to do that with designated fleets of ships), it is possible to effectively use a variety of tactics from spearheading into enemy fleets to flanking, surprising with the deployment of satellites and much more. However, currently the tactical combat suffers from a general unfinishedness. Sometimes enemies will suddenly turn around and flee. But because your ships and his ships have exactly the same speed, such a chase is potentially endless. Also, the size differences between the ship classes are not as fleshed out as in SotS. The means that you end up just not caring which part exactly was destroyed on that twenty-fifth destroyer over there or whether this missile frigate over here still has ammo.
Essentially, the tactical combat in SEV is usually more fun even though it us uglier than the one in SotS. It's fun to test you new idea of a mine-laying ram-ship and see how it disables that new dreadnought-class attack ship of the invaders while leaving a trail of death between your fleet and the enemy. Something like that is just not possible in SotS and it's also not fun having to guess most of the essential information as well as trying to keep control over those drunken Newtonians.
Diplomacy is where GalCiv2 clearly excels. Even though the diplomatic options in SEV are slightly more detailed, in GalCiv2 it is much more polished. Diplomatic messages actually do have some meaning usually and they do reflect the current 'mood' of the other civilizations towards yours. It is even quite fun to see those Drengins being enraged about your new military starbase right in front of their homeworld - simply because they will actually identify such a provocation and really say something about it. Or just send a fleet. But then there is the galactic council where you can choose to participate in or not. This council will vote on some "issues" from time to time which results in slightly changed overall game rules, if a vote is passed. Such as a tax on space stations in foreign territory. The council allows for an interesting game mechanic, where weaker races can band together in order to try to control stronger races, although that rarely happens. Overall, diplomatic races actually do have a chance to outwit more militaristic races, but the time-frame to do so is rather narrow. Once an overly militaristic civ becomes too big, there is no way it will let itself be stopped by some pesky council.
As mentioned, the diplomatic system is also extremely detailed in SEV. However, it is one of the most unfinished features currently. In fact, since it really allows almost any imaginable treaty and the AI has not been properly instructed to make use of that, the AIs often tend to end up in ridiculous treaties with each other such as the dreaded 'research lock' where two enemy civilizations cripple themselves by agreeing to not research anymore! To make things worse, they usually stick with it or a similarly bad treaty for quite a long time and these civilizations tend to lose due to that. Other examples of unfinished features are trading techs that might produce log entries which are switched or you receiving techs which you cannot receive and hence re-receiving them every turn.
SotS shouldn't be mentioned here at all, because, well, it does have a simple diplomatic system, but it serves the sole purpose to define peace or war between two civilizations with only a few extras.
Clearly, the winner in this category is GalCiv2 and I think even after the final patch for SEV some two years from now the diplomacy in SEV won't be as good. The diplomacy in SotS was never meant to compete.
Apart from building and deploying fleets and using diplomacy, the economy is another cornerstone of the 4X genre. The build-up of your civilization should be fun and competitive. It should allow for experimentation and unique solutions. Additionally, it should provide meaningful context for the battles: Do I really want to risk half of my entire fleet to hold that worthless low-quality world over there? In this respect, again, SotS wasn't even meant to compete. The economy here has only two variables for each planet: population and industrial build-up. Both get fairly fast to their maximum for any planet and then that planet acts only as another, well, planet producing ships. If you have 10 planets then you basically have only half the production capacity compared to someone who has 20 planets. As simple as that.
In GalCiv2, on the other hand, things are much more interesting. Each planet comes with its own map, where some tiles can be filled with buildings. The tiles may have bonuses and the buildings enhance certain aspects of the planetary productivity - either it produces lots of money, research or ships. The output of a planet can further be drastically modified by starbases near that planet, which adds a bit strategic depth to their positioning and the positioning of your defending fleets. Additionally, there are happiness and global sliders where you may adjust whether you currently need more research or rather more ships.
However, given that the AI in this game knows all these rules perfectly, there is not much room for unique solutions. Basically the player has to maximize the output of planets very tightly. Building a world full of entertainment centers because the player is trying to role-play a bit will very soon remind him of the fact that he just wasted a whole world. Falling behind economically in this game means an almost certain unhappy end. This is also where the useless ship designing comes in: In SEV the role-playing player at least has a chance to pull out some crazy thing such as building lots of ramming drones and spam the technologically overwhelming invading forces with them to gain some more time to finish the warp point destroyer to buy even more time to finish research on sphere-world construction. In GalCiv2, however, once you fall behind economically, you also fall behind technologically really fast. And then suddenly (also due to the lack of tactical combat) it's not possible to be inventive just to buy a bit more time.
Additionally, GalCiv2's economical model still doesn't manage to do away with micromanagement. Even worse, for new players this doesn't even pay off. The micromanagement comes from several factors: You have to either *know* how many morale buildings are necessary to keep x population in check or you have to constantly check happiness. It also comes from changing situations. Perhaps you realize you are in a prolonged war? Then just setting the spending sliders to military production won't help enough. Getting rid of a few research facilities in favor of a few more factories isn't made easy in any way and many clicks are the result.
In SEV the economy is simpler, but more detailed. Here, what you see is what you get. There are three resources. For each it's possible to build mining facilities and each planet comes with a value for each resource that tells how much of the resource a mining facility will be able to produce. Then there are more facilities such as research or intelligence. Again, each one produces a set amount of the corresponding 'resource' (i.e. research points) which depend almost only on the current population level of that planet. The only real gripe one can have with that is that the three resources are not diversified enough. Basically, you just have to have enough of them all, because lack of any of them means a halt to any type of production. A more diversified model would allow to build at least ground troops, for example, while not having radioactives any more.
As a final verdict, the economical model in GalCiv2 is much better thought out, but also for some reason quite boring. Probably because it is necessary to maximize returns here too much and there are only a few good variations to do so. Also, the economy model in GalCiv2 exacerbates the balance problem: Once one civilization has twice as many planets as some other civilization, then that other civilization hasn't the slightest chance to catch up.
The last major aspect of the three games is AI. Both GalCiv2 and SotS try to excel here, but only GalCiv2 does. Sadly, SEV, by its other features the most interesting game, doesn't even try to be good here, it seems. Even in the current patch-level (1.17) I fear to play too well because on turn 100 I often find myself having as many planets as all other civilizations together. Perhaps once the many bugs in the AI scripts are ironed out this will change, but I remain skeptical simply because even the last patch for SEIV, the predecessor, didn't help the AI to become noticeably smarter. These comments, by the way, apply both to the strategic AI, as well as the tactical AI.
SotS, however, especially because the entire game is rather streamlined (or shallow), has quite a good AI. In fact, on the harder difficulty levels it is indeed hard to compete in the strategic gameplay. The tactical combat AI, however, is rather mediocre. It won't do too many stupid things, but it'll always try to charge you head-on, for example.
As mentioned above, one of GalCiv2's strongest points is the excellent AI. The AI here is so much better than in any other game except for chess computers, that there isn't much to be said about it at this point. GalCiv2 is and will always remain the clear winner here.
In the end I don't think there is a clear winner here. SotS seems to be the weakest title of the lot as it doesn't have any particular strength over the other two. The other two games have their weaknesses and strengths. If you like to experiment a lot and role-play your empire, your preferences are clearly with SEV. If you want a challenge and your general sci-fi fix, go for GalCiv2. I personally like both SEV and GalCiv2. For some reason or another I keep returning to SEV once in a while. But after having played GalCiv2 excessively I never feel the need to play it again. The problem seems to be one of focus: Whereas in SEV you may focus successfully on a certain extreme style of play, in GalCiv2 you cannot afford to neglect even one part (such as economic, technology, or military build-up). The other races will mercilessly exploit that. That's good, of course, but then again, perhaps fun in games is not only being smarter than some admittedly smart algorithm!