Friday, April 20, 2018

Making a Better 4X (Part 2)

Continuing where we left off...

It can't be a 4X without the "eXpand" part, and in a space opera setting that means colonizing worlds.  The degree to which players are tasked with managing these operations varies a lot even within the sub-genre.  It can be anything from moving some sliders around on the overview screen to dragging and dropping individual units of population on a tile-like grid.  More complex systems of management tend to give the player the option of having an A.I. assistant take the reins although this tends to be a sub-optimal choice.  Personally, I don't think the level of detail is nearly as important as where that detail centered on.  For an interstellar empire it's not really about what's on the planet, but rather what the planet can get up into space.  To clarify that last point a bit further the total population of a planet is a lot less important than how many trained personal can be brought up to (and sustained in) orbital habitats.  Construction materials and equipment follow a similar line of reasoning.  Raw resources are another major point to consider.  For all intents and purposes it's a lot easier to mine whatever is needed from asteroids and small moons.  A fuel refinery in low orbit around a gas giant is far more useful than one down in the gravity well and dense atmosphere of a planet's surface.  From the player's perspective a planet's space infrastructure should be all that really matters when it comes to managing their interstellar empire.  Farms, factories, and power plants can be abstracted out of the picture because the things that really matter on the ground are space elevators and rocket launch/recovery facilities.  Meanwhile, orbiting stations that serve as shipyards, supply depots and living facilities count for far more by virtue of being readily available for use.  Having said all that, I think making planets entirely irrelevant would be a bad idea.  Quite the contrary, giving each colony and outpost a distinct feel adds a lot to the 4X experience.  Aside from biome type, world size, and mineral richness, most 4X games give planets little "tags" to help make them feel more distinct.  In Master of Orion it's things like artifacts (which boost research) or crystals (that raise colony revenue).  Stellaris is a bit more interesting with tags like titanic life or the discovery of a previously unknown underground civilization.  These sorts of flourishes are great, but it feels like there needs to be more of them.  Ideally, there should be a bunch of "tags" for each kind of planetary biome.  It would also be nice if habilital worlds were more defined by their native flora and fauna.  After all, if the planet can support life then chances are there's life on the planet already.  Whether that entails nothing more than primordial ooze or a sentient race of primitive natives, what's already living on the planet can dramatically affect what the player can do with it.

In Spaceward Ho! there is quite literally only five thing that are ever available for research; speed, range, weapons, shields, and miniaturization.  All scientific efforts in the game are directed toward war.  Master of Orion and it's sequel have six distinct categories of research which are further divided into individual "techs."  Stellaris does something similar, but reduces the categories down to three.  In Spaceward Ho! technology is a linear progression, level 3 shields are followed by level 4 shields and so on.  It's pretty bland, but functional.  Master of Orion: Conquer the Stars opts for a tech tree with more exoctic sounding labels.  In all but the first game in the series there was an attempt to create more gameplay variety by limiting the player's research to one of of several options within a given category.  The problem with that approach is some techs are more useful than others, plus it never really made sense to me why you couldn't simply go back and research something that had been previously passed over.  Stellaris does a much better job with its "deck-of-cards" approach.  Although I still find myself annoyed by the fact that certain techs will vanish from the list of options only to reappear later.  Stellaris also has some kinds of rare technology exclusive to certain empires based on government ethics or special events.  Overall, it is the best take I've seen yet, but it still has room for improvement.  As reluctant as I am to break that ever-so-important sense of immersion, for the purposes of comprehension it might be better to simply categorize fields of research into three clearly label subsets such as "ship," "colony" and "society" based improvements  Other than that, I see some interesting ways the Stellaris research system could be more heavily customized for a given species (depending on their natural tendencies).  Take, for example, the Bulrathi; they are tough, ecologically minded wolf-bears from a high-G world.  Based on that information I can come up with interesting setting questions like, do they refrain from heavily polluting industries?  How about terraforming?  Are they pro cybernetic and see it as a way to enhance their innate strengths, or do they view the tech as a kind of spiritual corruption?  High-G tolerance sounds like a boon for spaceship crews that need to withstand maneuvers with rapid acceleration or deceleration.  Does that make them better in space combat?  What about boarding actions?  There are more questions I could ask, but I think I've made my point.  A species having bonuses or penalties based on their nature is great, but I would also be curious to see what potential can unlocked over the course of the game.

In Spaceward Ho! population equals production.  Master of Orion uses a combination of population and factories to measure industrial output.  Stellaris turns everything into various currencies which are spent over time, or in a lump sum to achieve a desired goal.  If a player wants to build a cruiser then they need to pay a certain amount of mineral points and wait a certain amount of in-game days.  Once the warship is finished the player must pay a monthly energy credits fee to keep it operational.  None of these economic systems are particularly elegant.  They get the job done (so to speak), but as I mentioned in the colonization section, I think a decided focus on what can be brought into space (rather than what's on the ground) is the way to go.  The exception to this is terraforming.  In Stellaris they initially made the mistake of pushing this tech  out to the late game in an attempt to avoid the Master of Orion pitfall of having every planet end up with the exact same biome over time.  I see the issue here, but I honestly think terraforming should be available from pretty much the very beginning of the game.  I would also dump the prohibitive costs usually associated with utilizing the technology.  Instead, terraforming should simply be a huge time investment.  Say we're playing humans with the starting planet of Earth.  The first planet we colonize is Mars and from the get-go we start to terraforming it for a minimal expenditure of resources.  Over the course of the game the planet transforms from a barren world to a tundra-like environment.  By mid-game it's an arid desert world, and only toward the end of the game do we finally see a blue-green Mars.  The Jovian moon of Europa might travel a different terraforming path, starting off as a ball of ice, then becoming vast arctic sea, and at last a temperate ocean world.  Of course as habitability improves the colony could redirect efforts toward things other than simply sustaining itself.  The more prosperous the world the more it can contribute to the interstellar empire as a whole.  What that contribution entails can be left to the player's discretion.  Maybe they want one colony to work on establishing a star-system wide mining network?...or perhaps a steady supply of new space-adapted recruits.  Maybe the colony is geared toward research and scientific discovery?  Construction ships are nice and all, but much like merchant fleets, I don't see why the player needs to micromanage them.  Just assume they are there ready and able to do whatever task the player wants unless it's some kind of extra-special project like building a gateway between star systems.  Also, if we're going to have construction ships at all, I'd prefer something more along the lines of a mobile dockyard that moves from star to star gathering up materials and personal that are needed to build whatever the player demands.  It's sort of like a giant 3D printer in space, or a "Mothership" from the RTS series Homeworld.

Next up is spaceship design...

Saturday, April 14, 2018

Making a Better 4X (Part 1)

I've been bouncing back and forth between Stellaris and Master of Orion: Conquer the Stars over the last couple of weeks and it has made me think long and hard about the genre.  Pretty much every 4X has its own strengths and weaknesses.  Might it be possible to combine the best of each entry in the genre to create an ideal 4X?  Galactic conquest strategy games tend to follow a predictable gameplay pattern of species selection, galaxy creation, exploring stars, colonizing planets, researching technologies, manufacturing, ship design and after encountering rival empires things like diplomacy, espionage and warfare.  So, let's examine these core aspects of space 4X design and see how they are handled or see what, if anything, can be improved.

Species Selection:
Every 4X begins with the player choosing who they want to be and who they want to play against.  Spaceward Ho! (the first 4X I ever played) had no differences between empires except for the kind of hats they wore.  Evenly balanced, to be sure, but there was a distinct lack of variety.  One of the nice things about having a pool of diverse alien species to choose from is the dramatic increase in replayability.  Being humans from the planet Earth is a very different experience than say taking the role of a hive mind of giant insects from a chthonian world.  That said, balanced gameplay is oftentimes hard to maintain with a more exotic selection.  Ask anyone who has played a lot Master of Orion and they'll probably tell you that the Mrrshans and Bulrathi are some of the hardest species to win with because their advantages are minor compared to other races.  Stellaris is the most robust option currently out there in this regard with the ability to choose from a list of pre-made species and custom create new ones to play as (or against).  Players can even randomize the selection process or have the computer generate species whole cloth.  This definitely adds an extra layer of mystery when they player journeys forth to the stars.

Galaxy Creation:
On the surface this seems like a fairly straightforward task.  Astronomy ranks all stars in the universe on the OBAFGKM scale of classification with each having subsets like "giant" and "dwarf" within their respective light spectrum.  This is all easy enough to simulate in a digital environment until you account for the numbers needed.  Realistically, the smallest galaxy in the observable universe has 250 million stars.  The only game I know of that grapples with numbers of that size is No Man's Sky, which as someone other than myself astutely observed has the width of a vast ocean, but the depth of a tiny puddle.  More stars does not equate to more interesting gameplay.  Thankfully, there is an elegant way meld plausibility with playability.  Stars tend to form in clusters.  Representing the area in which the game takes place as one of these clusters allows for a realistic arrangement without melting your CPU and possibly your brain.  Of course it also begs the question, what is preventing the player from sending ships outside the cluster?  That's something I'll explain in the next section.  For now there is another potential numerical nightmare I want to address when it comes to galaxy creation - planets.  The discovery of large numbers of exoplanets in recent years has definitively proven that our solar system is unexceptional when it comes to orbital bodies.  In fact it appears that many stars have planets in significant numbers.  4X games have been, and still are, stingy when it comes to allocating planets around stars.  The original Master of Orion had a maximum of one planet in each star system.  Spaceward Ho! simply had planets suspended in the void.  Master of Orion II: Battle at Antares along with the most recent one, Conquer the Stars, rectified this to some degree, but the average is still only 2~3 planets per star with a maximum (that I've seen) of 5 planets plus an asteroid belt.  Stellaris pushes the number of planets per system higher still in addition to having colonizable moons.  However, the game isn't able to represent our own solar system accurately without modding support.  It's certainly possible that Sol has an exceptionally high number of planets compared to most stars, but even so I think any given 4X should be able to generate an eight planet star system.

Now that the game has begun in earnest we come to the first "X" in 4X games.  So, how does one go about it?  We're going to need FTL (short for Faster Than Light) travel.  Scientifically speaking it currently only exists in the world of mathematics.  Sci-fi authors, on the other hand, have come up with inventive ways intelligent life might be able break the ultimate speed limit.  Generally speaking FTL can be distilled down into three basic forms: Warp (Star Trek), Gates (Mass Effect) and Lanes (Wing Commander).  Most early 4X games used Warp, but more recently the genre has shifted over to Lanes and, to a lesser extent, Gates.  Stellaris had all three, but recently embraced Lanes, similar to what Conquer the Stars did in comparison to the the rest of the Master of Orion series.  So, why do developers like Lanes so much?  The reason comes down to A.I. programming.  The more possible moves available the harder it is for a computer to make good decisions.  This is why your PC can play a killer game of chess, but is barely competent at Go.  Compared to Go, chess has a fairly limited moveset; hence the reason it's good at the latter and not so hot at the former.  Lanes also create some interesting strategic options that are lacking in the other two forms of FTL.  That said, I don't think Warp and Gates should be excluded entirely.  When you get down to it trying to travel from one planet to another, even within the confines of a single star system, can be a real headache both logistically and in terms of resources needed.  Therefore, I think having some kind of Warp drive, even if only capable of sub-light speeds, would be an extremely useful component of any interstellar empire wanting to accomplish things on human timescales.  As for Lanes and Gates, I think combining the two technologies could create some interesting choices for the player.  Assume for a moment that each star system (on top of planets, moons and asteroid belts) has a bunch of nodes that connect to other nodes in neighboring systems.  Building a gate at a node opens up that's the catch though, it's a one-way trip until a gate is set up on the other end.  So instead of sending scout ships or a science vessel (depending on the 4X) the player must mount a resource intensive expedition complete with a survey team and construction ship to set up the gate on the other end.  In this way it becomes very important for the player to consider where they want to focus their "stellar cartography" efforts.  Some nodes with more distant connections require more advanced gates to utilize which can only be accessed through technological advancement.  In this way new routes will become available as the game progresses, creating shortcuts and unlocking new areas for exploration or avenues of attack while still keeping things within the star cluster generated at the start of the game.  In fact there's a lot this sort of FTL arrangement adds to wartime strategies, but I'll discuss that in more detail in the section on warfare.

Whew!  Long post...expect more in about a week where I'll continue this space 4X examination by continuing with the next item on the list - colonization.

Tuesday, April 3, 2018

What I'm Watching

I wanted to embed a couple of game related videos that it thought suited the themes of this blog so without further ado:

Wednesday, March 21, 2018

Sticking the Landing

I've recently been reading through "Swords v. Cthulhu" - an anthology of short stories.  For the most part, it's an interesting collection of fantasy/horror fiction.  The one big exception being an annoying tendency for many contributing authors to fumble the conclusion.  It's not a problem exclusive to creative writing.  Far from it, most forms of entertainment media have the same kind of issues to varying degrees.

Sadly, when it comes to video games, I can't say it has been a recent problem.  Many older games have had notoriously awful endings, in no small part due to the assumption that few players would actually ever make it the finish before moving on to something else.  Hence, developers rarely felt the need to put real effort into that last bit before the credits roll.  A couple of games that I think concluded on a strong note are Legend of Zelda: A Link to the Past, Vandal Hearts, Ghouls and Ghosts, as well as the original Valkyria Chronicles.  Contrary to expectation, a large number of highly regarded story-driven games kind of drop the ball at the end.  Pretty much all the Silent Hill titles have overly obtuse finales, and the Soulsborne series (for all the attention I give it) unanimously finishes up in a manner far too abrupt for the amount of time and effort it takes to make it there.  That's not to say every game needs to conclude with an hour long cutscene, but when you look at the amount of building up the Mass Effect trilogy did you're left wondering what the heck they were thinking by slapping on those half-ass red/blue/green endings.

The problem can be so pronounced within the industry at times I sometimes think that, when coming up with a narrative arc, game designers should figure out the ending first and then work their way back from there.  Otherwise this whole player-not-finishing-the-game-because-of-crap-endings becomes a self-perpetuating loop.  One way I've seen developers try to circumvent the issue is by playing up the idea of a trilogy, or at the very least a definite sequel.  Everything from triple-AAA titles like Halo 2 to indie games such as the Banner Saga 2 try to pull it off.  Occasionally, it works out well enough, but more often than not we get Half-life 2.  Some games don't even make it that far.  The Order 1886 just stops abruptly at what would normally be the second act in a three act story with no sequel forthcoming.

These sorts of screw-ups are why I prefer self-contained plotlines that give a sense of closure even if there's a sequel in the cards.  From a business perspective, I get it.  Publishers are convinced they'll make more money off their IPs if they leave the customers wanting more.  Sometimes these sort of monetization techniques can get blatantly exploitative; such is the case in Dead Space 3's real ending being paid DLC.  Obviously, a non-trivial number of players got sick and tired of all these half-baked story arcs to the point that companies like EA and their ilk decided that narrative driven experiences are no longer profitable.  Gee...I wonder who made it like that...?  Maybe it's not good business sense to trash an integral part of game design just to make a quick buck.