Monday, June 12, 2017

The First Time's the Best: The "Civ Problem"

There's a problem in most computer games. It first came to my attention while playing the Civilization series. Once you've isolated it, though, you see it occurring over and over.

It has to do with the way certain things -- from building a city to opening a treasure chest -- become by the middle of the game rote, boring, and annoying, with too many button clicks and a long animation to sit through. Paradoxically, in the early game these tasks are fresh and exciting enough you as the player find yourself wanting even more. More choices. Longer animations. More detail.

In my opinion, the choice to try to strike a balance is the wrong one. I think games need to do something different. The question is how the design team can afford it.

At the moment I'm contemplating the crafting system in Skyrim. Many people have chosen to do a "crafting" character build; to downplay other play aspects and even ignore the main campaign in order to specialize as an Alchemist (aka Herbalist), Blacksmith, etc. This is popular enough that third-party Mods are out there to allow such things as Tailoring, or a Bard character build (where you collect sheet music and instruments and buff your performance skills, paying for all of this with money you earn performing at taverns).

Here's the problem. That first cycle; of finding iron ore, finding a smelter, getting leather, starting to craft weapons and armor increasing your skills until you've unlocked the Riddle of Steel, is fun.

It isn't combat fun or barrow delving fun or talking to locals fun, but the game world is big enough to support many play styles. And the details and animations and so forth are enough to give you the feeling of actually doing that work, of swinging your pick against a vein of ore or stropping a hide with a knife.

Thing of it is, after a mere half-dozen iron daggers you are able to graduate to making "fine" quality weapons and armor. At a bit over a dozen you can create "superior" quality (depending on materials and perks).  These aren't just words. They translate to monetary value and combat effectiveness.

It takes forty-three iron daggers made by your hands before you can graduate to working on Elven and Dwarvish metals. By that time, even watching the animation is starting to pall. Gathering those forty-plus iron ingots and twenty-odd leather hides has become more of a chore than not.

To buff your skill to the point where you can make Dragon armor (the best stuff in the base game) you would have to craft 574 iron daggers. (I know this because other players have experimented and logged their results).*

So this is the "Civ Problem." In the Civilization series, you start out as the leader of a wandering stone-age tribe. The game begins as your people domesticate local plants, build permanent structures -- in short, translate to an agricultural economy. Aka your tribe moves to the civis, to a city-based existence, and the game starts with the founding of that first village.

So the early game is a lot of fun. You learn about the importance of storing grain. You make the decision whether to build a granary or prioritize on other pressing problems (there's this little thing called "Writing" which is really starting to catch on.)  You are learning how to do new things, and experimenting with them, and you are learning a little history as well (the games give you some background about the history and social consequences of whatever technology you have just discovered, and have all had some form of encyclopedia where you can learn more.)

Your civilization grows. Eventually the village is so large you send out colonists. Eventually the satellite towns are large enough to need lebensraum themselves. By mid-game, you are managing tens of burgeoning cities.

And every single one of them is clamoring for your attention in deciding whether to build that granary, or that temple, or those barracks, or a rapidly growing list of structures. The Tech Tree is also sprouting, with new developments coming at you so quickly you no longer care to read the deep explanations about the historical reality this represents, what impact it will have on your growing society, etc.

As of Call to Power the series has added some level of management, so you can template some of these decisions or hand them off to a local representative.

My opinion, though, is that this doesn't go nearly far enough. As you go through the game, "you," the leader of this particular tribe-polis-peoples-nation, get addressed in-game by titles that reflect the expanding society; "Chief," "Lady," "Queen," "Madame Chairman," etc. But I would like to see the game itself change. That when your single recently-settled tribe is learning about farming you the chief actually goes out there and paces the fields. And when your village needs defending you are personally one of the strong arms that does it.

But in mid-game, you should have a staff. By late game, a cabinet. No village should be asking you about where to build a granary. But also, your chief of spies and army commanders and Surgeon General should all be working independently, meeting with you to keep you updated on scientific research and urban decay and police actions in far-off lands, but leaving you free to focus on the big picture. No...the US President doesn't fly a fighter jet personally (even if aliens are attacking).

What this game would look like I'm not quite sure. Something more like the much-anticipated but generally panned Spore by the same creator. Spore changes the game completely, multiple times; each time the focus changes, from tidal pool to interstellar civilization, the tasks change, the gameplay changes, even the interface changes.

This is a problem everywhere in games. The problem boils down to things only being new once. In most cases, the freshness is insufficient to last through the number of times you eventually end up having to do it. It is both too much in the later game -- and too little in the early game.

This is probably built into the structure of games themselves. A modern computer game -- especially an AAA game -- works on an economic model where the game costs from $40 to well over a hundred dollars, and is expected to deliver in turn at least thirty but into the hundreds of hours of play in return. The economics break down to the designers being able to afford a staff from a dozen to half a hundred bodies to work on it for two or three years (and they are shamefully underpaid).

And, yes, there are expectations about graphical and sonic richness that pretty much require much of that money be farmed into detailed models and textures, extensive voice acting, musical scores (and other costs, like licensing and/or developing the game engines that will push graphics and AI behavior and so forth above that of the competition).

The solution set to this model is to create essentially one set of assets. One (or a relatively small number) of treasure chest models, animations of opening them, sound effects, a central loot system that determines what can be found in the chest, an overall inventory management and availability model that provides game balance. And then it is copy-paste to fill enough world to fill the requisite hours of gameplay. This of course in a close fit with a essentially rote model of play, where the same cycle of interaction with these assets is repeated for as many times as the player can stand it.

One interesting thing is that games will usually chose to provide variety by providing a multiple of tasks. Not a multitude of approaches to a task. So the same chest opens in the same way every time -- but there is also a plant that can be harvested (the same way every time).

Actually, though, there is an exception to this. Melee and magic. There are usually a large variety of spells and weapons and (often) fighting moves. Each having distinctive animation. And this makes sense from a economic vantage. The average player avatar is likely to swing a sword tens or hundreds of times more often than they open a treasure chest.

(It is also because the rock-paper-scissors of even a complex melee system can be relatively easily broken into discrete game-able elements. The truth table of block, attack, power attack does not lead to hundreds of cells. Plus of course combat is so central to the philosophy of so many games, there is a rich heritage of existing solutions to draw on.)

But this is just one part of a whole problem, to wit; a single animation, as well as a single asset management system, a single skill progression system, etc. Over the player avatar's Hero's Journey they will inevitably accumulate history, choices, inventory, and so forth until what was in the early game sufficient is in the later game annoying. One size does not fit all.

As just one example, the third-party content is replete with better inventory management mods for the late-game character (or experienced player). It is the Civilization problem in a different form; managing fifty cities the same way you managed one village doesn't work. And an inventory system for tracking your half-dozen iron daggers and some strips of hide falls down when you have three hundred magical weapons and full suits of armor plus magical gems and various keys and treasure maps and...

What I would like to see is scaling: even as I recognize the economic difficulty of it. Games are forced by their economies to make one asset and stretch it over as long as they can get away with it. This asset stretching leads to all sorts of oddities that we generally shrug off in the honored, "it's just a game" manner.

Such as health packs and ammo. Both asset minimizing and the pressure to design a fast-moving style of play means that the same damned first-aid kit or ammo drum, with the exact same visual appearance (aka in-game model) is found everywhere, very much including those places where it has no right to be.

(I want to take examples from Tomb Raider: Underworld and particularly Half-Life2 for challenging this. The former in the smallest ways; health packs morph into mysterious containers of healing liquids when you enter ancient ruins. In the later, there is good in-world justification for every supply cache you find.)

This makes the game world less rich. In Tomb Raider 2013, for instance, there was a wonderful variety of clutter, making Shantytown or the ruins of lost Yamatai visually distinct. But the same stock box of shotgun shells appears in both. This leads to a visual paradox. One's eyes end up passing over all the other detail, because it doesn't matter. By mid-game, you know there is nothing new to discover. All those mouldering books and arrangements of candles and ad-hoc sculptures of fishing weights and flotsam are non-clickable, essentially non-interactive. You've been trained that the only things you can pick up and use is that same damned box of shells and its brothers in arms.

(It also leads to Gamer Eye Affliction. Spend six hours playing Skyrim as an Alchemist, and when you take a walk outside in that strange VR with full HDR rendering you find your vision locking on to every clump of red or blue flowers you pass.)

The modern FPS/RPG has also shot themselves in the foot by inevitably activating some sort of icon, cursor-change, or hovering "pick me up" label. So once again there's no need to actually look at the pile of stuff. No need to engage with it visually, to explore it with fingers and pattern recognition. Just wave the cursor around until the icon appears, then click to pick up.

It breaks immersion (one of the few exceptions I can think of is the Deadspace series, where all of the annoying clutter of floating icons and, "pay attention to this!" HUD sounds are diagetic. They are part of the claustrophobic, information-dense environment Isaac Clarke lives in.) And worse, it changes focus. It is the same essential problem with QTE's (Quick-Time Events, aka a movie you watch while sometimes having to make a timed button press). The game may be showing you Lara Croft leaping from a burning helicopter while an undead samurai is clawing at her leg, but that's background noise. Where most of your visual attention is, is on whether the icon that just flashed up is the left action key or the right trigger. Instead of playing an exciting sequence, you are playing a lame version of Guitar Hero while an action movie plays in the background.

This is only one aspect of the re-use of stock assets. A peculiar one crops up in the game I seem to love to single out: reloading the weapons happens frequently enough the animation is necessarily made short enough not to be annoying. So, as it happens, is searching a dead body. Leading to the ludicrous spectacle of Lara Croft riffling enemies for more ammunition while in the middle of a firefight with them!

Which segues limpingly to one of the things games do do towards making that first time special. They pop up a tutorial panel the first time you do something. It goes without saying this breaks immersion. But it is worse.

Especially if you have managed to stray even slightly from the expectations of the designers, this tutorial moment tends to unfold this way; your character was just attacked by a dragon, who with one taloned swipe took out most of your hit points. You fumble desperately for that health potion you pocketed earlier in hopes of living long enough to come up with a real plan of defense. And the game pops a tutorial pane. "Do you know? You can restore health by using a Health Potion. Health Potions can be purchased...."

The panel is quit (in the worse cases -- looking at you, Tomb Raider 2013 -- automatically after some seemingly random interval) and dumps you with your fingers in the wrong place, inventory not opened after all and health potion not in your hand, as the dragon opens his jaws and flames you back to a distant Save Point.

Well, I hate to say it, but if you wanted to make the first time more meaningful, that's actually the route to take. Even more scripted, even more on rails, forcing the player to do this thing for the first time ever in completely controlled conditions.

Which is how essential play mechanics are communicated in the better games. Sort of. A new mechanic or weapon or whatever is introduced in a controlled "training" setting, or a training wheels situation. In Half-Life2, there is the famous, "Pick up the can" sequence, but a better model for how cleverly the game approaches tutorials is when Gordon is fleeing through a window and Barney shouts up to him, "Pile up some boxes to reach it!" Thus telling you what the in-game physics engine can deliver, and allowing you to build on your can-carrying experience to make a useful step.

Thing is, in say the optional Smithing tutorial in Skyrim, this inevitably unfolds as doing the exact thing, with the exact screens, models, animation, and outcome as will happen hundreds or thousands of times again over the course of play. Because the purpose of it, after all, is to instruct the player on the button presses they will be using for the rest of the game. So the excitement of first time is externally tempered by the needs of the designers not to include more detail or length than will work equally well in the middle of the game.

So can it be done? Can a game take a complex task, approach it in an engaging and detailed way that satisfies the player's need for involvement, then abstract it -- perhaps in multiple stages -- over the course of play?

There are examples of how this plays. Fast Travel, as one example; almost every RPG has a mechanism by which you walk or ride to a new destination the first time, but afterwards can snap to it with just a brief loading screen. In this case, the first time of cresting the ridge and seeing the city spread out beneath you happens; you had to work for it, spending in-game time walking, fighting off monsters on the trail, etc. And now that is done, and if you have to stop in again just to buy some more silver ingots you can just pop in and out.

And this is a choice, too. There are Skyrim players who are philosophically opposed to Fast Travel and insist on actual hikes everywhere. They are the virtual-world equivalent of backpackers, choosing to spend their time experiencing the trail and the vistas along the way over merely arriving at the destination.

A similar time-honored approach is the, "Door to before." In many games, after struggling through a difficult area you will find or open a previously inaccessible door that lets you back out without having to traverse the whole thing back again.

And as I said above, some Civilization games allow you to assign a mayor to manage the day-to-day chores. They can be consulted, inspected, over-ridden, even removed at any point you wish to go back to engaging with the original detail.

This connects to replay value; like tutorials, the first-time events are better if they are optional or otherwise can be skipped. Because for the returning player, they are also not a first time, and the big moment of getting your first horse, settling the animal and getting him used to you, setting the saddle, mounting, etc. could be rather annoying to have to sit through again.

The problem, as I see it, is not one of design. It is one of design philosophy. The paradigm of the AAA is to create one level of experience and iterate that for as many hours as it will stand. Assets are not just minimized, they are re-used from game to game and even from company to company. A deeply immersive approach to each fresh experience is essentially designing an entire mini-game around each experience you wish to highlight.

And I say "mini-game" here because the only times I've seen a first-time experience highlighted is via cutscene. No playability is added. No, in short, additional asset, design, code, play-testing is required; just time to develop one more animation (and, usually, the cinematics -- crafted camera angles, cuts, sound design, etc.)

The example that pops to mind is the deer in the early parts of Tomb Raider 2013. Lara has killed her first animal, and the game highlights this by moving to cutscene as she butchers it. After that, butchering is a quick animation and some changing numbers on the HUD.

But this doesn't -- except in that one way -- add that first-time effect I'm talking about. You don't have additional playability. There are no new choices (there are in fact fewer choices; the cutscene plays without you having to approach the deer and press the "interact" button at all). So for all the visual and aural richness added to this experience, it doesn't offer a deeper engagement.

I am reminded of the stock mathematician joke. One version; mathematician and engineer are sharing a hotel room a a conference. For some reason a small fire starts in the wastebasket, fortunately waking them up. The engineer reaches for the complimentary water glass and douses the fire. Next night, the engineer is out drinking when there is another small fire. The mathematician opens his eyes enough to confirm, mutters "yes, a solution exists," and goes back to sleep.

Wouldn't it be cool if, like Fast Travel or the Door to Before, a problem like how to swing a sword or aim a bow or craft a dagger or pick a medicinal plant had to be done, once, and then the player is given the option to mutter "yes, a solution exists," and think no more about it hereafter?

(I'm especially thinking this is in fact a game, not a tutorial. That you can totally screw it up, and have to try again, until you can finally say you've got it down and can move on.)

This could be on both ends of the detail scale. You could search the room for treasure in the time-honored way. Or click on a follower and say, "Lydia, search the room for me, will you?" Because you've already as a player had enough of that experience, and it should be your choice whether you have to spend five real-world minutes doing a rote task that ceased to be interesting many hours of gameplay ago.

But, again, where does the developer time come from? The AAA game especially has already decided that since shooting at an enemy takes ten seconds and is mildly engaging, the best way to make a game last a hundred hours is to provide 3,600 enemies. For the occasional stealth-based game, replace with 3,600 meters of corridor to sneak down.

I think the problem, in short, is in the basic framework of the AAA design and sales cycle. In the customer expectation, in the incestuous relationship with computer and graphic card manufacturers, in the advertising environment that prioritizes flash over content. Producing a problem for unique problem-solving is hard, game-ifying it in a graphically satisfying way that will look good in a ten-second advertisement spot is even harder.

For both the developer and the player, the bulk of the game experience is going to be, "yes, a solution exists," and yet having to iterate that known solution over and over and over and over and over.

*The Skyrim iron dagger experiment is one that various players have tried. The formula behind skill increases tracks to the value of the items crafted. In the early game iron daggers return the most skill points per coin spent. The faster -- fewer mouse clicks, at least -- Smithing progression means making more expensive stuff as soon as you can afford to do so.

The traditional approach is to switch to Steel as soon as you can afford it (befriending the Riverside blacksmith helps a lot) and buff your way to Dwarvish as fast as you can. Because Dwarf stuff sells for a lot and the metal is free for anyone willing to challenge the Dwemer ruins for it.

An alternate path, however, is to specialize in jewelry. And the big trick here is that there is a Transmutation spell out in the world. It only requires stealing it from some bandits in a handy iron mine, and after that you can make gold and silver from cheaper metals. Mining your own ore has another advantage here, though; you get random jewels in your ore bucket, which allow you to make more expensive jewelry, which not only covers your costs better but also buffs your skill faster.

The biggest downside to skill-buffing exploits is that Skyrim uses level tracking. Skill points in any skill counts towards levels, and many of the perks (like being allowed to work Dwarf metal) require gaining a level to spend anyhow. So it is far too easy to find yourself at level 20, facing down level-appropriate wandering monsters on the trail to your favorite iron mine, with nothing but an iron dagger in your hand and combat skills to match.

No comments:

Post a Comment