Wednesday, August 24, 2011

The self-made irrelevance of the RPG

In a recent interview with VG24/7, BioWare's Greg Zeschuk spoke about the direction the influential developer has been taking in recent years, and into the near future.  Most controversial about this interview were Greg's words on RPGs, and how he viewed them in today's industry:

"I think broader. I mean – it’s funny – RPGs are and always have been our bread and butter, our heart is there, but at the same time I think – well, we had the RPG panel breakfast at GDC yesterday – and what was interesting about that was that we had the conversation about ‘what is an RPG’ and it’s a blend. The genres are blending right now, you’re getting lots and lots of progression and RPG elements in shooters – online persistence and so on.

"It’s funny because the RPG in the context of the current world is – well, it’s not specifically irrelevant, but it’s becoming less relevant in and of itself. It’s more a function of ‘hey, this game has a great story’. For us having that emotion but also having other great features like combat and persistence of character progression and stuff."

In this short piece, I'd like to take some time to break down some of Greg's thoughts on the subject, because I think that they're not only divisive, but also indicative of an ongoing - and dangerous - trend in RPG development.

History lessons

Developers have been bandying about this question for a while now, and, more than ever, I've seen the answer "it's all about story, and characters, and the world" come up.  Furthermore, a few months ago, controversy was sparked when Matt Findley claimed that RPGs always really wanted to be action games at heart, and that the strong mechanics of past games was not a legitimate design choice, but a limitation of technology at the time.  While it's no secret that RPGs are known for narrative, and that many RPGs have tried to incorporate action elements (often with little success), I think that this sort of view, that rulesets are an outdated concept and that RPGs are all about storytelling, is an inaccurate, ill-informed and poisonous one.

RPGs began as an offshoot of wargames, tabletop games designed around complex interactions of rules, where two or more players would compete against one another for dominance over a battlefield... while wargames were firmly entrenched in delivering a strong strategic experience, the pen and paper RPGs of the 1960s and 70s, most prominently emerging with Dungeons & Dragons, added a crucial and now defining characteristic to wargames: cooperation.  In order to do this, most RPGs centred around, rather than global military operations, individual characters and parties of characters, cooperating and competing with each other in order to accomplish a goal.  The design change was in many ways a necessary one, but also a popular one, and RPGs soon far outnumbered wargames in terms of popularity and success.  As narrative elements began to creep into RPGs, as players began to get attached to the characters they played as and the universes they inhabited, RPGs began to become associated with storytelling in addition to those mechanics.  While the name, role-playing game, reflected the inclusion of narrative, it still originally, and in my opinion, more accurately, reflected the fact that players had to cooperate within a predefined ruleset to solve problems, effectively serving functional roles within a setting whose narrative concepts only existed as a vehicle to structure the experience.

What are rulesets?

Though we tend to talk a bit about rulesets when discussing traditional RPGs, a moment to clarify just what these are may be in order.  A ruleset is just that: a set of rules, existing primarily to not just govern, but facilitate the interaction between players and the game (or the dungeon master).  A ruleset's strength is that it is universal: while we tend to think about rulesets as being primarily a determinant of combat outcomes, a well-developed ruleset will articulate a general standard by which the entire world is defined by.  The reason for this, quite simply, is because rulesets aren't just about determining how much damage you did to an enemy with your Vorpal Axe of Slaying +3 - they're about determining the way in which you are able to express yourself within an alternate game world, just as much as the physical rules of interaction our reality follows determine our own capabilities.

When the advent of computer RPGs came about, it initially wasn't the strong narrative focus that was carried over, it was the established rule-set, adapted to fit a solo experience on a personal computer.  Early CRPGs, including the famous Gold Box titles, were, though narratively charged, really little more than frameworks for players to enjoy a ruleset within.  And some of the most successful CRPGs ever, such as Fallout, were successful as RPGs not because they had great stories and characters, but because they created a world whose terms of engagement were well established in the mechanics in addition to the narrative, and then allowed the vast majority of the storytelling to be taken over by the player, as he or she navigated the world in accordance with those rules.

The "Golden Age"

Modern videogames are, of course, a diverse medium, and most traditional videogames don't depend entirely on explicitly defined rules, at least as far as the player's understanding goes.  Typically, these games also have a lower barrier of entry and lend themselves to a different style of play, one that relies more on reflexes and coordination than an overt manipulation and consideration of those rules.  A trend, which began in the late 2000s, and largely fueled both by newer gamers getting into game development, and a desire to appeal to wider audiences, saw traditional RPG rulesets joined with more action-oriented game design... even as far back as BioWare's Baldur's Gate, the focus of the game was not limited solely to the rules of the game world, but on being able to take part in a grand adventure, and in immersing one's self in a large and well-realised universe.  With the rise of more and more graphically and narratively focused RPGs, such as Planescape: Torment and Deus Ex, not to mention the JRPGs that preceded them by several years, the focus of RPGs on the PC began to change from the rulesets and the player's interaction with the world on a mechanical level, to the player's interaction on a social and narrative level, fueled less by mechanics than the player's own desires and imagination.

Today, I feel as if the concept of RPG, especially in mainstream understandings, has become severely diluted - what remains of the core rule-driven experience of RPGs constantly finds itself under threat of disappearing altogether, as the demands of modern game production require greater and greater budgets, and thus, the market for any particular game, RPGs included, must be expanded.  All of this, of course, is fueled by a misguided notion that RPG fans are concerned less with strong gameplay mechanics than they are by storylines.  If the RPG has become redefined over the course of the last decade, it is only because the premier developers in the CRPG space have decided to abandon the core of the genre of itself, choosing not to make RPGs founded on consistent rulesets, but to create cinematic experiences that differ from most action games only in that they tend to feature a greater ratio of dialogue to action.

A question of relevance

And here we find ourselves today, with the RPG an "irrelevant" genre, as said by one of the apparent fathers of the modern CRPG.  In light of the history I've articulated (mind you, an incomplete and highly simplified one), I think it's safe to say that the question of relevance on Greg's mind comes either from a misunderstanding or change in perception of what an RPG actually is, or from a desire to no longer make RPGs.  BioWare have, for many years, been at the forefront of delivering cinematic and story-driven videogame experiences to players... considering the ease at which many of these games can be divorced from their mechanical underpinnings, and their narratives told in a way unhindered by statistics, it becomes questionable whether BioWare are, or even have been for the last eight or so years, in the practice of creating RPGs at all.  RPGs have traditionally been about universal rulesets, and even the best narratively-charged CRPGs have governed those narrative qualities via mechanics - Planescape: Torment, Fallout, and even more recently Alpha Protocol, have all built their stories around the fundamental notion that it is the player's choices in statistically developing a character or a party, rather than around the idea that the player's decision-making be conceptualised as a choose-your-own-adventure novel.  That Chris Avellone has been involved with many of these CRPGs may or may not be a coincidence.

Of course, I do not mean, in any way, to suggest that BioWare's games are bad, because, for what they are, they are categorically of a very high quality standard.  Yet close analysis demands that we consider very carefully the kinds of games that BioWare claim they've been making this last decade... under scrutiny, I don't think that they hold up particularly well as RPGs, at least when we think of RPGs in the more traditional sense I've defined above.  The conclusion that I'm forced to draw through all of this is, simply put, that if BioWare feel RPGs are no longer relevant, part of the answer is that they have made them irrelevant.  And those fans that made companies like BioWare successful in the early days, the ones that they keep claiming they're making their games for - they're still around.  They just don't care much for the fact that the innovations in the upcoming BioWare titles have less to do with crafting mechanically sound RPGs, and more to do with alien love triangles.

Thursday, August 18, 2011

Encounter design in shooters 201

In my article last week, I took a close look at what make up the "basic building blocks" of encounter design in shooters, including both staple environment types, enemy types, and how those two can be combined and varied to produce encounters that are fun to play and engaging non a mechanical level. In this week's article, I'll be looking at how to take those basics to the next level, by offering up some concepts on how an encounter can go from just a simple corridor crawl or arena battle, into something that carries emotional weight, variety, and fits in well with the game's overall pacing.

Elevation, or how I learned to stop worrying and love three-dimensional space

If the term "corridor shooter" implies nothing else, it's two-dimensionality. People, spending most of their time on the flat ground, already have a predisposition to want to focus their view forward and keep headed in a single direction; when playing a game, especially on a console, where moving around is easier than actually looking around in a full 360 degrees, people become even more tempted to lock their sites forward and barrel forward. In fact, this priciple is so strongly rooted that many successful console shooters have actually met their success precisely because it's easier to create an "interactive shooting gallery" than truly play with 3D space. While effective and straightforward to design around, this kind of thinking can be extremely limiting to the design of your environments, your enemies, your weapons, and, subsequently, your encounters.

By far the number one way to augment an existing environment design is to add multiple levels of elevation. This can be done for a number of reasons, all of which can be valid in a given situation, but which, when used right, will always positively agument your encounters and overall game structure.

1. Keeps the player moving forward. Adding a few jumps down here or there on the path can give the sense of forward direction, while also making sure the player is effectively "checkpointed" along the way to avoid confusion involving backtracking. Usually if there's a drop down that looks safe enough to make without injury or death, the player is going to think one thing - that it's the right way to go.

In Crysis, the player is usually given the option to survey and review the playing field before fighting.

2. Allows for tactical consideration and breathing room. In a firefight, it's always a good idea to know what you're getting into. While it's rarely a good idea to give the player the high ground or a clear picture of the map in all cases, usually the player is going to want a way of viewing the entire layout of the battlefield so he or she has some foreknowledge and time to plan out his or her angle of attack.

3. Provides new opportunities on the fly. This actually goes both ways - the player can use elevation to his or her advantage, while as a designer, an interesting challenge can be constructed by varying enemy placement based on the player's own position. Take the low road because it's shorder, but risk fire coming from above, or take the high ground but work through an additional layer of enemies? Varying the path also means that enemies will continue to come in from different angles, and, when combined with the different enemy classes discussed in the first article, it means that there's all sorts of interesting combintations that can be worked with.

Call of Duty 4 varies its terrain and gives enemies multiple, and unexpected, angles of attack.
4. Gives your enemies more interesting and unique roles. While it's easy to dismiss the prospect of, say, jetpack soldiers, based on the established world you're working in (it probably wouldn't fit too well in a realistic World War II setting), the fact remains that when your levels are designed in two dimensions and extruded upwards, you're going to be limited in the ways enemies can approach the player and provide challenges - ultimately they're going to be some variation of "shoot the player in the face" or "run up and hit the player in the face". No number of new skins will change this unfortuante reality, but smart design of your environments will, and adding elevation to the mix will provide the necessary zest and vitality to keep enemy roles interesting. This also allows one to more easily pace the game as a whole - not only do levels exhibit new ideas to accommodate new enemies, but your overall level design is going to need to change and become more complex as multiple enemies begin to enter the fray.

Getting beyond two-dimensional thinking also involves the inclusion of what I'll refer to as "agumentations." These generally include environmental hazards, platforms and ladders to climb, puzzles to solve, and so on. Whether that's navigating lava and bottomless pits with precarious bridges suspended above, hopping on an elevator to gain a height advantage, dodging in and out of massive, gnashing teeth as you duck and cover your way through a giant alien organism, or blowing up a wall to proceed forward, all of these can be mixed inside of combat encounters to provide the player with a more interactive environment that can work hand-in-hand with the narrative demands of the situation.


Ultimately, two-dimensional thinking in level design isn't, itself, a bad thing... but it can't be your only type of thinking, or you're very quickly going to start running out of ideas, and a lack of ideas means a less interesting, enjoyable game, with less inspired encounters for the player to fight his or her way through. It's important to remember that, when weapon and enemy design considerations begin to enter the fray, level design becomes a holistic process. Proper understanding of all the tools at the player's disposal, and the enemies that the player will be facing, are just as key as designing a level that looks good in isolation.

Mission objectives and layers, not "kill all dudes"

It's not a very common complaint to see given the nature of shooters, but gamers tend to pick up on it when it's done poorly: while a shooter needs to be about shooting, if you can't give interesting context to that shooting, you're going to end up with a boring game. One of the most sure-fire ways to do precisely this is to offer up a single repeating goal: "kill all the dudes, then wait for the door to open." Sometimes, this can be a literal mechanism - in the case of Call of Duty, the player literally cannot open the door to go to the next sequence, and instead has to wait for their NPC companion to open it for them, which of course, only triggers after the fighting in the given area has been completed. Other times, it can be a little bit less obvious - an NPC that needs to hack a terminal, and only finishes after X enemies are killed, or an enemy machine-gun that must be disabled before progress is viable.

There's nothing wrong with any of these scenarios. In isolation, they're totally reasonable objectives to put into a shooter game - if your primary mode of interaction with the world is to blow it up, well, there's going to be a lot of blowing up done. The most direct approach, however, is in the case of shooters rarely the best, and so providing interesting, plot-relevant objectives that revolve around more than just shooting is key to providing an engaging experience, and, more specifically, engaging encounter designs.

Mass Effect 2 often builds upon objectives to provide additional intrigue.
Consider Mass Effect 2's mission to recruit the character Archangel. While in many games, this would be a simple straight-up affair of blasting through hordes of enemies, BioWare decides to build multiple goals on top of what ends up being a relatively straightforward task, leading to a more complex and entertaining scenario. First, players are able to spend some time scoping out both the environment and speaking with some of the enemy characters, and are even able to perform a few tasks in advance to optionally make the fight a little bit easier (rewiring a mech to cause it to turn on its masters). Once the push to meet Archangel begins, the player will first have to battle former allies, only to later defend the building Archangel is holed up in, from both upper and lower levels that favour different combat disciplines, and have advantages and disadvantages. Midway through this battle, the player must head down to the basement to seal up some new routes that enemies have opened up, leading to a semi-non-linear section in clausterphobic tunnels, where the battle is against time, not just an enemy horde. Finally, it culminates in a boss fight against an enemy gunship on the top floor, where the player must use heavy weapons to bring the thing down, mixed in with close-quarters gunplay. By the end, there's a strong sense of finality to the proceeding, and a "blank slate" for the rest of the story to continue from.

All of these different layers in Mass Effect 2 work together in order to provide a unique gameplay experience, all while utilising different types of enemies, different levels of elevation, different sub-objectives that aren't all directly combat-related, and additional game features like hacking, in order to provide a sequence that is fun, well-paced and memorable. Additionally, the player has to look at the environment from all angles due to traversing it multiple times, re-evaluating the same battleground as new enemies come in from different approaches; all told, this half-hour section of Mass Effect 2 gets more gameplay out of its rather small level than most other shooters would get out of a space five times as large. What's more, there is a narrative arc in here: players arrive not knowing much about Archangel other than he/she needs to be rescued, they arrive and learn the background of the situation, hints are dropped as to Archangel's identity, and midway through, the tone changes from reaching an unknown to fighting to save an old friend's life. It's a mini-narrative all in itself, and it works hand-in-hand with the variety in gameplay goals.

Strong motivations are strong gameplay

How many times have you stopped and thought to yourself, "just why am I doing this?" while playing a shooter? If the answer to that question is "ever", then the designer has failed to provide adequate emotional or rational justifications for what you've been asked to do. While I'm not here to provide tips on writing a good story, approaching this from a designer's perspective, there are certain techniques that can be built into the context of a situation in order to give the player a reason to keep moving forward... and they need to be utilised a lot more often than you might think.

To put it simply, the design goal in a shooter is to try to convince the player the game isn't about shooting. This might sound a little bit strange, but what I mean here isn't that the goal is to make a shooter with a lot of additional mini-games tacked on, or to literally make a shooter where shooting is a secondary concern. Rather, as a designer, you need to be able to give the player a strong motivation other than "my purpose is to shoot guys." Whether that's rescuing a captured NPC, hacking a terminal to gain enemy intelligence, sneaking into a base through some sewers, or driving a tank to punch through an enemy fortification, the player needs to believe he or she is doing more than just shooting for shooting's sake. It's not enough to have "cool stuff" in your game - you need to have "cool stuff" that players are given reason to care about.

Half-Life 2 avoids the worship of killing by presenting it as a necessary means rather than an end.
As an example, in Half-Life 2, the player is very rarely, if ever, explicitly told to shoot something. There's no moment where the player is handed a gun, and instructed to go out and become the harvester of souls. When shooting inevitably arsies in Half-Life 2, it's not there because it's the player's purpose to take lives, it's a logical, justified reaction to the player's situation in the game world. From there, the goal is never, ever to simply shoot more enemies - it's to rescue a friend, or protect a rebel encampment, or meet up with some allies. It's easy to look at a shooter with jaded eyes, but becoming jaded is also the first step on the path to producing a rote, by-the-numbers game, and the first step to providing your players with yet another title to roll their eyes at.

It's also worth pointing out that gameplay needs to have a natural ebb and flow for players to stay interested. One of the easiest way to dilute the impact of your encounters, no matter how well designed they may be, is to pace them poorly. Players need to have breathing room, space to think, a natural moment where they can pause (or shut off the game) and collect themselves. When a given encounter, or string of encounters, drag on for too long without enough in between, players are going to grow fatigued. In Call of Duty 4, this is achieved by providing a strong contrast and mix of high-action levels with short walks and story sections between large battles, atmospheric stealth sections which can feature minimal or even no shooting at all for extended periods of time, and even levels that exist solely to advance the story and provide an emotional impact. If Call of Duty 4 had been one straight roller-coaster ride, all the way from top to bottom, then it simply wouldn't have the same emotional resonance that renders it a modern classic... and on a personal note, it's this lack of coherent flow which has largely turned me off from all the subsequent games in the series. To borrow a phrase from the audio production world, without quiet, there can be no loud.

Closing thoughts

I realise I've covered a lot of ground in this article - everything from layering mission objectives on top of one another, to dealing with the physical layout of your environment. The simple fact is that providing a point-by-point breakdown of each and every fundamental component gets harder and harder when you start involving "softer" elements of game design, like narrative, but I hope the examples I've given do a good job of illustrating how designing encounters in shooters is, beyond a purely mechanical level, a discipline which involves an encompassing knowledge of the game, not just those fundamentals I covered earlier. If the basics are able to give you an "adequate" game, it's paying attention to the details as covered here that will elevate a game to something more than "just another shooter". Thanks for reading, and feel free to add your own thoughts in the comments below!

Wednesday, August 17, 2011

More modeling fun, texturing, and more inconsequential stuff

So, after a lot of difficulty in learning its interface after becoming used to Blender over the last while, I've begun to get the hang of Maya.  While I can't say I'm an expert by any means, I've adjusted my workflow and am finding that certain operations can be faster, so long as I don't rely on some of the old habits that I formed while using the other software.

Case in point: in Blender, per-vertex modeling is a lot easier because you can simply select vertices and draw faces between then with a single keystroke.  I haven't been able to find such an option in Maya.  However, extruding faces and generally manipulating large pieces of geometry is also a good deal easier in Maya, and once you get the hotkeys down, the workflow itself can be extremely quick and easy to manage, to the point where I'm able to perform tasks without even really thinking about them.  Maybe I'm just getting better, but I think it's also a testament to the strength of the software that I'm able to migrate over to its different interface and get the hang of it within about an hour, and end up working faster than I did in what I was used to previously.

I've been busy with modeling for the last few weeks, pretty much learning the ropes as a full-time job.  Now that I feel I've got a hang of it, and the biggest test is less my ability and more my patience to produce something decent, I've decided to move on to texturing.  I actually have some experience with this in the past, though I've never textured any of my own models before, so it's been an interesting experience learning the process.  I'm not particularly adept at hand-painting anything, but using photos, my Photoshop knowledge and compositional eye is more than enough to take me through.

My first textured model.  They grow up so fast.  Specular needs adjustment and the texture for the handle is probably wrong, but I'm actually impressed how this turned out for a first attempt.
A note about the texturing process: I'm sorry, Autodesk, but Maya's UV mapping tools leave a lot to be desired.  Okay, they're highly functional, and that's great, because UV mapping can be a delicate process requiring a lot of fine-tuning, but unlike Blender, Maya has very few decent options for automating the process.  In Blender, it's possible, once you've marked your UV seams, to literally press a key and have it auto-scale your UVs based on their relative sizes to fit your UV texture - it's not perfect, but for the most part it does a good job and the amount of manual intervention is minimised.  In Maya, no such luck - not only is the process of even marking seams a bit more complicated, but you have to manually position each and every face of your model for optimal texturing.  For something simple, like a primitive shape, this isn't too big a deal - it might take all of 2 minutes to set up your UVs.  For something like a head or character model, though... well, maybe you should just come back in an hour, I might be ready.

EDIT:  My mistake.  Turns out it is possible to unwrap meshes in the same way as in Blender, and it's a little bit more intuitive than I initially thought... that said, however, there's no tool I've yet been able to find that effectively auto-scales, and Maya has crashed on me more times than I can count trying to combine multiple UV layers into a single UV map.  While there's more functionality in Maya, and the auto-projection can work great in certain situations, it's still a slower process overall, and feels more like a chore than a quick necessary step.

I've also been experimenting with more sculpting in ZBrush and Mudbox.  While I have some experience in ZBrush, I haven't before used it to sculpt an existing mesh I've created.  I was actually impressed that the importing process actually worked without much whining and complaining - simply export as an OBJ file.  While I imagine there are probably problems involved when bringing the software back and forth multiple times, for a single export-import-export, it worked out well, and left me with some nice normals to use for my...

... chest-high wall/pillar.  Hey Epic, if you need any more environment artists, call me!
All in all, I'm really enjoying this.  While it's hard to say if my passion for 3D work will continue in this relatively early state, my first truly dedicated stab at it after my largely on-off efforts in the past (frankly, I just didn't have the patience and discipline five years ago), it's certainly been a lot of fun, and it's scratching an artistic itch that I haven't really felt since I was a full-on art student in my late days of high school.  I gradually did less and less drawing through university, simply due to my attention drifting to my coursework, so it's nice to get back in the artist's chair in a slightly more hands-on way than the level design I've been doing for the last while.

Current WIP in Maya, and my first beyond simple stuff to learn the interface, plus a departure from the norm.
I've got my hands full right now trying to finish up the System Shock 2 guide for GameBanshee, as a couple more bits and pieces I'd forgotten about were brought to light when I started trying to combine the thing into a whole, rather than separate documents.  Currently just about done, with only a few more pages to finish.  Although it's been fun, I'll be glad to have it off my hands, if only because it's hard to stay truly excited about writing a walkthrough for more than a week or two.  The guide I did for Fable III was the same way, but in the end I think it ended up pretty thorough and well-composed, and this new one is even more detailed.  System Shock 2 is a game that's pretty close to my heart, so I'm glad to contribute in my own small way to the discourse and community around it.

Hopefully, I'll be able to write a follow-up "201" piece to my article on shooter encounter design tomorrow or the day after, dedicated to something beyond the fundamentals.  If it's the basic building blocks of enemies, environments and arrangement that one starts with, then it's the context, the scenario around the combat, that provides the necessary colour and flavour to help support the game mechanics themselves, so that's what I'll be going into.

Thursday, August 11, 2011

Head models part deux

Another day, another model brought to life.  I decided to continue with building a head, since I wasn't too happy with how the last one turned out, and wanted to see if I could improve.  Due to a much smoother process in building it, I'd say the results more or less speak for themselves:


Getting a good reference seems to be key.  I worked with a much higher-res photo this time around in order to really help nail the details, and I feel that the character of the original shines through pretty strongly precisely because I was able to tweak the details a little more.  While my edgeflow still needs work overall, I'd say this is a pretty good step forward considering it's the second complete model I've put together.  Perhaps if I'm feeling adventurous, I'll try texturing or rendering later on.

Wednesday, August 10, 2011

Forays into modeling: adventure waits!

Despite all the time I've spent doing level design over the last several years, I've never really poked my head all that much into 3D modeling or using any of the related tools.  While I've picked up a lot of the fundamentals of interface and texturing due to working with Unreal Engine and others (which base much of their own design off of 3DS Max and similar), the actual process behind creating things from the ground up has been something I've been blissfully unaware of.

So, what better time than the present to try to jump in and try my hand at things?  While I've resorted to using Blender due to the lack of many free alternatives to the more common industry-standard programs, I'm finding its interface pretty in line with most others, and it's quite easy to use for a beginner, save for some annoying quirks, and features that are buried way too much for their own good.  After a few initial experiments in order to get a hang of the interface - modeling simple things, like apples, glasses, chairs, etc. - I decided to try my hand at modeling a human character... or at least, a head.


While the actual results aren't exactly what I'd call professional-grade, with some odd cheekbones being the biggest standouts, as well as a lack of detail around the neck, overall as a first experiment in vert-by-vert modeling, I'd say I could have done worse.  Total time was about six or seven hours, including learning to use the tools a bit more.  I did, of course, use a photo reference, though the one I chose was probably too low-res to really be useful for serious modeling.  I used the basic face shape as a guide, then tried to fill in my own details as best I could, with the ears especially requiring a good deal of "ad-libbing" on my part.


My biggest problem, most definitely, is that I ended up falling into the trap of adding in too many unnecessary bits of detail.  When it comes to constructing organic models, I've learned the hard way that maintaining clean edges and loops is critical; the "keep it simple, stupid" mantra applies more than ever.  Although I probably wasn't quite as bad as some other newbies due to sketching out the basic structure of the head before going at it, I ended up with way too many vertices than I actually needed, especially in areas around the cheeks and ears.

Fortunately, I was able to avoid a lot of problems with the basic organisation due to some very helpful tutorials I came across, from the good people at Blender Cookie.  If it wasn't for the help provided there, I'd certainly still be struggling to form the most basic primitives.  I realise it's not a good idea to paint entirely by numbers, of course, but when someone's already worked out a technique that works well, it's not at all a bad idea to learn by it.

The one part I'm still pretty curious about from here on in is how one ends up getting a 3D model into a game engine... converting this sort of mesh into polygons is partly an automated process, I'm sure, but I'm sure the actual optimisation work and getting something ready for prime-time inside a game is a difficult process, and probably takes just as long as the modeling itself.  While I'll be able to work twice as fast with enough practice, doubtlessly, this whole experience certainly makes me respect the professionals a whole lot more.  There's a reason development teams are half made up of artists these days.

Now, I suppose all I have to do is try this about 500 more times, and I'll be fit to model eyelids for the Madden team.  Well, nobody said it'd be all fun and games...

Sunday, August 7, 2011

Encounter design in shooters 101

As a long-time shooter fan, I've always appreciated a good firefight. Whether it's in a game like the original Unreal, with large open environments and gymnastics aplenty, in a slower and more realistic title like Rainbow Six, or the more modern cover-based variety seen in Gears of War, I'll usually find something to like. That said, I've rarely sat down to really think about why shooters are fun, how they work, and how they keep up the player's interest for hours on end when, in essence, they revolve around a fundamentally simple interaction, one whose context may change but the mechanics rarely do.

If there's one thing in a shooter besides its storyline, characters, the weaponry, and the environment, it's the encounter design. Without well-defined encounters, a shooter will most likely fail to hold sway over the player and turn into a boring slog through corridor after corridor. These days it's very hard to get the fundamentals of shooting enemies wrong, so it's going to be the combat situations the player is thrown into that ultimately decides whether or not the player continues until the end. In this piece, I hope to break down the basic components of level design in shooters, provide examples of standard enemy types, and then discuss how the two interact with each other by giving a simple encounter design to illustrate.

Environment is everything

Building a shooter, it's easy to get caught up in the aesthetic side of environment. Yes, everyone wants a shooter to look pretty, and in many cases a developer will be constrained by technology, and by the setting of the game, but without being able to get the fundamentals of gameplay down, there can be no fun, no matter how cutting-edge or aesthetically pleasing a game might be. It's in constructing the basic physical dimensions of the game levels and environments that much of a shooter's gameplay hinges on.

For the sake of simplicity, there are about four different types of combat scenarios that one can identify:
Hope you like hallways.
  1. The corridor. Used often as a pejorative, corridor shooting really doesn't have to be as boring or repetitive as it sounds. The corridor is defined by a long, often narrow passage, or collection of passages, often interlinking and intersecting with one another, passing on top and below, etc. Designing a good corridor is deceptively difficult, since it can involve so many separate layers depending on the needs of a scenario, from simply deciding where cover should go, to alternate passages, to playing with vertical space and jumping from one corridor to the next. Remember that corridors can be both good and bad in design: see the Berlin subway tunnels in Call of Duty: World at War, with their multiple routes, platforms and enemy positions on different sides for a good design of this type, versus the repetitive, closed, and highly linear spaces in Doom 3 as generally poor design.
  2. The courtyard. This is most often characterised by an open battle stage where the player is attacked by a force on one end, and must make his or her way to the other end. This is a staple of cover-based shooters especially. In a courtyard environment, alternate paths up the left and right sides of the area offering varying advantages and disadvantages are common, as are enemy positions on the upper stages of the level, which allow for snipers and suppressors to fire down upon the player, but also provide tactical advantages when the player is able to reach those spots. It's extremely common to find this in third-person shooters like Mass Effect 2, but Half-Life 2, Prey and other traditional shooters have many such encounters as well.
  3. The arena. The arena is similar in concept to the courtyard above, with one crucial difference: the player finds him or herself fortified in the middle of the level, with enemies attacking from multiple sides, typically emerging from doorways on the outer sides. An arena can both figurative and literal, both open or closed and made up of corridors; the key defining aspect is that the player will be firing from a fortified position and must defeat enemies before they get too close. Once again, there's both good and bad design in arenas. Giving the player a single fixed position to fight for is typically not too much fun, while a more open area providing different levels of elevation, cover, risk and reward, etc. is generally preferable.
  4. The turret section. This final type of fundamental encounter in a shooter is one where the player typically has his or her mobility in some way limited, if not completely removed, for the sake of providing a shooting gallery experience. These are a staple in most shooters, and provide a necessary change in pacing, often giving the player the chance to use a powerful mounted weapon like a rocket launcher or minigun. Turret sections can both be stationary and mobile, with the player either taking up a fortified and fixed position, or riding on a vehicle, which either the player or computer controls. It's important to note that it's easy to get stuck in a fixed mode of thinking with these sections - while it's totally possible that a turret section can be reduced to a simple "shoot the bad guys as they fly by" scenario as seen in an early vehicle section of Battlefield: Bad Company 2, it can also be something far more varied and interesting, like the points in F.E.A.R. 2 when the player is able to pilot a mech suit - for all intents and purposes it's still a turret section, but the ability to get in and out at will, and the movement within a small, fixed space gives the player a sense of control that makes these sequences memorable rather than routine.
While it may seem overly simplistic to break down a shooter into this number of component parts, upon closer examination, this analysis isn't at all inaccurate when one considers that the actual layouts of most first-person shooter environments all conform to either these fundamentals, or combinations of them. So, while scampering across rooftops in Left 4 Dead might not seem much like a courtyard, the actual routes available, enemy positions, and risk/reward options in different pursuits all fully adhere to the model. When juxtaposed and interspersed smartly with corridors (office buildings between rooftop sections) and an arena fight (waiting for a crane to move a platform into position as enemies attack), the actual setting almost becomes incidental; the walls of the courtyard have merely been replaced with 20-story drops, and the ceilings with sky. Not getting caught in the belief that certain environment components necessitate particular aesthetic design elements is the key to avoiding stale and repetitive encounters; with a little ingenuity in that juxtaposition and in the aesthetics, the player will never notice the basics underneath.

What's in a name? Enemy archetypes


Poor Metrocop, he's died so many times...
Now that we've got a cool environment to play around in, where do we go from here? The second key quality in designing fun shooter encounters lies in the enemies that actually populate it, both in terms of their fundamental capabilities and their positioning around the level. I've played a lot of shooters which haven't made proper use of the enemies that have been created, and were content to simply place the same legions of assault rifle-toting soldiers over and over throughout environments, to the point where even if the design of those levels was creative, the end gameplay was still boring because there was little variety in the interaction with and combat within that environment.

Like the environment archetypes, there's a number of fundamental enemy roles which just about every competent shooter needs to include to allow for sufficient variety, use of those environments, and interesting scenarios. It's worth pointing out that it's common for games to provide newer, stronger iterations of the same enemies, either as literal re-skins or as ones who share the same function but look a bit different, or do so in a slightly different way; the "super" versions of these guys fall into the categories below just as much as the basic ones.
  1. The goon. This is your standard foot soldier, the troop that the player encounters most often on the field of battle and who compromises the most basic challenge in the game. The goon is, in general, going to be of a medium level of intelligence, with the ability to take cover, melee on close contact with the player, and perhaps throw a grenade or two while firing on the player. The key to the goon isn't so much that any one of them is a major challenge, but that when combined in numbers and with the rest of the enemy types in the game, they provide a natural barrier or wall, in order to impede progress and make it more difficult to fight off other enemies. In other words, it's fairly easy to destroy a sentry turret or tank on its own, but it becomes much harder when you've got gunfire from other sources eating away at your health bar.
  2. The sniper. Another staple, the sniper is usually equipped with a sniper rifle or other long-range weapon, and, while relatively weak when faced head-on, can provide a serious challenge in certain situations, especially when the player's attention is occupied by other enemies. The sniper is a situational enemy, of course, but including at least one or two in most encounters, especially those made up of courtyards and arenas, is bound to make an encounter a lot more interesting. In fact, the sniper might be the number one enemy type to really change the nature of a fight, next to all others. A sub-type of the sniper is the turret, who is also stationary and typically long-distance, but can be disabled easily in a similar manner; the difference is mostly the danger posed (rate of fire vs. damage).
  3. The tank. What more needs to be said? The tank is the beefy, powerful, kick-your-face-in enemy type that's a pain to fight, but whose large size and speed ultimately ends up being an impediment, with the player able to out-maneuver it while using the most powerful weapons in his or her arsenal. Examples include the Rumbler from System Shock 2, the Antlion Queen from Half-Life 2, and so on. All of them are formidable on their own, but when they're combined with other enemies, the situation changes from mere single combat, to a puzzle.
  4. The rusher. Also referred to by an alternate title, like the brawler or flanker, depending on the particular game, the rusher's role is to get in the player's face and generally keep his or her attention while other enemies pelt him or her from a distance. A rusher is almost always going to catch the player off guard, interrupt existing actions, and generally annoy and make life difficult. If your rusher gets your player angry, this is actually a good thing, as the primary challenge isn't so much one of sheer prowess, but of wits and management, with the player weighing the time and resource investment to take out the rusher before it ruins his or her day, against the greater threats. Good examples include the Fast Zombie from Half-Life 2, or Wretches in Gears of War.
  5. The assassin. Contrary to the actual name, the assassin's role isn't so much to assassinate the player, as it is to provide a constant feeling of tension during the combat, the sense that the player is never really quite as safe from harm as he or she thinks. Bonus points to go one if it's got some sort of creepy or off-putting quality, but even so, the assassin is less about real threat and more about perceived threat, and scaring the player into running, changing tactics, and generally giving the other enemies to do their work. For examples, look to the Protocol Droid in System Shock 2, the Poison Headcrab in Half-Life 2, and the Hunter in Left 4 Dead.
  6. The spawner. Spawners are effectively powerless on their own, but rather they exist to create new enemies for the player to deal with, either on their deaths or under some other condition, usually brought on by the player's mistakes. Spawners may have their own separate offspring, or they may release other types of existing enemies, but the main mechanic behind them is to give the player a sudden incentive for more careful play to avoid provoking more harm than good. More obvious examples include the Pregnant from Dead Space, the Flood from Halo, an, in a minor variation, the Boomer from Left 4 Dead.
Of course, like the environments mentioned above, combining these funadmentals into different enemy types is a common practice, and sometimes it's hard to pinpoint precisely what role an enemy fills. Moreover, different enemies can have redundant roles depending on what stage of the game the player finds him or herself in. The most important thing about these fundamentals is that they all have a clearly defined role that the player can predict and use his or her knowledge, the environment, and his or her equipment to overcome in a predictable way. The real challenge from enemies shouldn't so much be in the individual dangers they provide, but in what sorts of capabilities they have, and how those capabilities interact with one another to give the player an obstacle to overcome.

Composition, and putting it all together

So, now we've got both a theoretical environment to work in, as well as a number of archetypes to put within it in order to design an interesting encounter for the player to get through. But how does it work? What's the best way to use these given enemies with an environment? How can we vary things and provide a unique experience to the player with each encounter?

The first thing to remember about using building blocks is that they're just that: building blocks. With the basics, it's never about the basics themselves, but in what you're able to create out of them. Much like learning a new scale on a musical instrument, that scale is largely worthless unless it can be integrated with other musical elements and used within a full piece, in new and interesting ways each time. Like composing a song, composing an encounter in a shooter should be treated in a similar way: taking those familiar elements and putting them in places the player either won't anticipate, hasn't seen before, or making them perform tasks that provide an interesting new way to play the game.

Thanks to Epic Games for providing this example from Gears of War 3.  Click for larger size.
Above is a pretty typical example of the courtyard layout from the upcoming Gears of War 3. Courtyards are a staple of the Gears of War series, but they're also a great example to use in combining different enemy placements and arrangements, as well as hypothesising different routes for the player to take when moving through the level.

There are a few fundamental features here: a large, central fountain which provides both cover and serves to break up the left and right sides of the level; a raised position which can be fortified, but is also exposed from both the left and right sides; and two partially covered approaches on the right and left sides of the courtyard (only partially in frame here). Additionally, buildings on the right and left sides also provide potential positions for enemies to attack, and pieces of cover dotted around the area provide the player with a means of traveling from one side of the courtyard to another without being completely exposed.

Already, we've got a number of different options for the player: sneak around one of the sides to reach the fortified position, brute-force and use the fountain as cover while slowly making the way forward, sit back from a distance and try to snipe enemies, maybe making use of one of the adjacent buildings to gain a tactical advantage, etc. If we were to simply throw a few random goon enemies in here and call it a day, the player would still probably have a decent five minutes of gameplay here. But how do we make it more interesting?
With a few varied enemy placements, this scenario is more interesting to play (click to enlarge).
Here, I've added a few different types of enemies to the mix. The player will start somewhere at the back of the courtyard, and the end goal will be to reach the top section at the end. The yellow "G" signs represent goons, the teal "T" is a tank, the blue "R" stands for rushers, and the pink "S" is a sniper. Please excuse the roughness of the example.

The number of options available in getting through this environment as described above are still here, but instead the player needs to contend with a few different threats. First, there's a number of goons stationed around all sides of the courtyard, mostly to provide suppressing fire and impede the player's progress a little bit. The player can't just run past and rush up the stairs; he or she will need to be smart about the approach and in taking them out, without becoming too exposed. On the left side, there's some added cover, but it's watched over by a sniper, which could ruin the player's day if he or she were to run blindly forward. On the right, a pack of rushers will assault the player, impeding what is otherwise a straightforward run to the end in a way that could be fatal, as the rest of the enemies take advantage of the player's preoccupation. Finally, at the top of the stairs, a tank waits for the player, firing rockets down below and as the player approaches its domain.

Overall, while not the most unique combat encounter ever, most players will find themselves having to make a few choices based on play-style, weaponry available, skill level, and willingness to expose one's self to danger. It's also completely possible that, with a few change-ups in the enemies (for instance, moving the tank to the middle of the battlefield, adding another sniper or two, removing the rushers and adding a spawner along the narrow side passage, etc.), that the combat here could take on a different dynamic. Moreover, a similar enemy composition could likewise be moved to a different environment, for a different type of challenge. And finally, while the objective given here is very simple, if it were compounded with a sub-objective, like the requirement to flip two switches on opposite sides of the courtyard before moving on, then the player would need to face both threats on either side, once again choosing to tackle them with a different approach, be that in direction, weaponry or tactics. Suffice is to say, things could look very different, and play very differently, with only a few changes.

Closing thoughts

What I've given here are just some basics in creating an interesting combat scenario. Obviously, with a lot more time, effort, as well as more interesting environments and more particularly-defined enemy types, the actual second-to-second nature of gameplay can be further tweaked. For example, I haven't defined what the precise firing intervals of the different enemy types are, I haven't suggested any non-combat or stealth routes, I haven't defined how easily the player is able to take on a standard goon enemy, I haven't addressed the issue of game difficulty levels, and how those might change the nature of the situation. All of these will change depending on the game, and what might be an option or non-option for one, may be available or unavailable for another.

I hope that this breakdown has provided a better understanding of some of the fundamentals in building encounters in shooters, and helped to better conceptualise the nature of encounter design in both existing games and for those in development. Please feel free to contribute thoughts, and point out any fundamentals that you'd like to discuss, or dispute. Thanks for reading!

Tuesday, August 2, 2011

It's good... I think.

My review of Pirates of Black Cove has gone live, for anybody interested.

It's interesting. The developer, Nitro Games, have a history of making strategy titles with deep, complex mechanics. While I haven't touched them, from what I've seen of East India Company and Commander: Conquest of the Americas, it's pretty clear that they're aimed at strategy nuts, not casual audiences or even the mainstream. What they've done here is essentially made a super-friendly, super-simple version of Mount & Blade with some Monkey Island humour on top. And, to be honest, it's hard for me to sit there playing it and think, "why didn't they go all-out"? Was it the $20 price tag and their perception of what value to deliver? Did they think their intended audience would find more complex mechanics confusing? Was it just a genuine misstep, and they didn't have time to add in extra features? All the ingredients for a great game are here, and yet it seems to squander them with its own lack of ambition

It definitely works, but it's a hard sort of review to write because, even if it does have legitimate flaws, it's hard to tell just how many of them are due to the nature of the game and my own preferences colliding with one another. I don't write full reviews on enough games I have mixed feelings about to find myself in this situation often, but it's interesting here because it's got all of the elements that I'd love to see in a game, completely pared back to the point where only the most fundamental concepts remain... as if a fan of Civilization had to go and play Civilization Revolution instead, or a fan of military simulators like ArmA suddenly being asked to give an opinion on a Soldier of Fortune game. Try as you might, it's hard to put yourself in the shoes of "someone who might like this game" without feeling like your thoughts aren't really welcome.

That said, I did enjoy Pirates of Black Cove for what it was, but writing a review for an RPG-focused site, it's hard to maintain a level head when in every paragraph you've got to throw in a qualifier like "for a $20 game" or "for RPG fans who don't want a good RPG."

Currently I'm paying attention to E.Y.E: Divine Cybermancy, a deliciously homespun game made by a tiny team of French mod developers. The game has the usual rough edges: near-broken enemy AI, strange performance issues, and one of the most awkward translations I've seen since the original STALKER, but the underlying mechanics are possibly the best I've seen in an FPS/RPG since System Shock 2. Granted, it's a much faster-paced game with a greater emphasis on combat, but the sheer number of options, paths of progression, statistics, and the openness of the levels really invites creativity. And, although I'm not really one to emphasise presentation too much, it's a damned gorgeous game, with a brilliantly bleak and alien art design that nevertheless doesn't feel brown or boring. There are few games that have captured the cyberpunk imagery well, and this game completely nails it in both its soundtrack and visuals. If you've been looking for a game like Deus Ex, but with a co-op focus and can ignore a poorly-told storyline... this may well be the best game of its type in ten years or more.