Elder Game

[I'm not working on the game this week, instead I'm doubling up on my other work so I can spend a full week on the game next week. But I've had one little anecdote bubbling around in my head. Indulge me, it's short.]

I had been hired at Turbine to work on the sequel to Asheron’s Call. I was an engineer through and through; I’d been coding (non-games) for nearly a decade and I was pretty good at it. I was hired as a senior engineer, and I got right to work on coding the game engine.

But the game had a troubled development cycle with lots of changes of direction. As the vision of the game changed, the team got less excited. “This isn’t a sequel to Asheron’s Call at all!” became a quiet undercurrent. It was around this time that one of the producers grabbed me at random as I was coming back from lunch one day and asked me to preview his presentation. He was working on a pitch that explained what Asheron’s Call 2 was all about, and he needed to refine it by getting feedback from people like me. I was happy to oblige, so I followed him into his office.

He then closed the door and began his impassioned pitch about why the new game direction was fun. I was a bit of a hard sell: I had specifically come to Turbine to work on the sequel to the game I loved, not to just “make a good game.” But I was willing to hear him out.

At one point, he mentioned that they were removing the Stamina bar. “In Asheron’s Call 1, you have three bars: Health, Stamina, and Mana. That’s dumb. You just end up using spells to convert stamina to mana to health to stamina to health and so on. We’re simplifying…”

I interrupted there, and said, “but that’s fun!”

He looked a little taken aback. “No it isn’t! Why is it fun?”

And… I couldn’t explain why it was fun. I was an engineer, not a game designer, and I hadn’t spent years thinking about why things are fun. I just knew that it was fun.

He figured it was because Asheron’s Call 1 was my first MMO. “People always get stuck on what their first MMO did.” But no, I corrected him: Asheron’s Call was my third MMO. I’d played a little bit of UO and a ton of EverQuest. I knew what else was out there, and I knew that Asheron’s Call 1 was fun in different ways. I just couldn’t explain why.

That was an important day for me. I don’t believe I could have changed the course of the game at that point — it made no difference what I said or didn’t say. (All I did was cause him to focus on other design elements for his presentation.) But it was important because I realized I wasn’t a game designer, despite thinking I was. I’d played tons of games, I knew all the mechanics they used. But here I was, unable to defend the simplest concept. It was frustrating.

It changed how I played games forever, and not necessarily for the better — when you study everything to see “why is this fun”, you sap some of the fun out of it — but I think it’s the critical skill that game designers must develop. Just that. The rest of the job is details, but you have to be able to articulate why something is fun… or at least why it could be fun.

Nothing is fun for everybody, and nothing is fun in every game. A lot of times I can only recognize fun by watching others: there’s plenty of genres I don’t enjoy, so I can only hypothesize about why they’re fun. And I may often be wrong. But the act of thinking about it is what’s important. If you don’t even think about it, all you can do is unconsciously mirror what you’ve seen before. You can’t put disparate elements together (let alone try brand new things) with much hope of them being fun.

I think I can now defend why Asheron’s Call 1′s three separate energy bars were fun, but I’m not going to bother doing so. It’s beside the point. The point is… screw that producer. My MMO’s going to have three bars. No — four bars. No. Five.

My plans for NPCs have been refined a bit. At first, what I envisioned sounds similar to the “Storybricks system” recently announced by Namaste. I’ve abandoned that, now, though, because it’s too convoluted given my time/resource constraints.

But there’s a lot of different things you can do with NPCs depending on your goals. I want NPCs that are memorable and that tie into the rest of the game’s mechanics elegantly. And in fact by dropping a lot of the extraneous stuff, it’s become a lot easier to see how to do that.

Every game has to figure out where NPCs fit for them. For a pure-action MMO like World of Warcraft, NPCs don’t have much role to play. But my game has lots of non-combat skills, and NPCs need to be a big part of how those work.

Examining (and Skipping) the Classic Answers

I started out by looking at some of the traditional computer game NPC systems. A classic one is schedules — giving every person a “life” beyond just waiting for you to show up and buy their crap.

Coding it? Not so bad. But it needs a whole lot of content work. (“Okay, so the town has 33 NPCs. I need to make sure there’s 33 beds, or at least that they can sleep in shifts…”) But more than that, it’s not very fun in an MMO.

In a single-player game, you can just push a button to “advance the clock two hours” when you’re waiting for a shop to open. In an MMO, you might be waiting several real-time minutes for a shop to reopen. Ugh.

I’m interested to see what the Storybricks system comes up with to avoid the annoyances of NPC schedules. But this isn’t a problem I want to try to solve for my game.

Another classic mechanic is talk trees. I do have talk trees, but they aren’t a major aspect of the game, because they are incredibly expensive to write. (They have to be interesting enough to read, or else you end up with boring talk trees like EverQuest 2 had when it first started… and that’s worse than no talk trees at all.)

In the end, I decided the best way to interact with NPCs in my game is to use verbs on them. For instance, you can use the Show Item verb to show something to an NPC and see how they react. Or you can Give Gift, or Shop, or so on. You unlock verbs by mastering skills, so eventually you will have Psychoanalyze, Entertain, Flirt, Control Mind, and others. These all affect the NPC game systems in simple ways.

Interaction Mechanics: Friendliness, Favors, Gifts

In my mind, NPCs and the skill system have a symbiotic relationship. The skill system gets a big jolt of entertainment from interacting with NPCs. And NPCs are fun specifically because of how you can use skills on them.

But in order to make NPCs fun to interact with, they need some custom mechanics.

First off, every NPC has a hidden number that represents how much they like or hate you. If you do nice things for them, this number goes up. What are “nice things”? Well, if you save the town from certain doom, it goes up for everybody in the town. But you can also give gifts to people. (You’d need to use social skills to figure out the perfect gift.)

Of course, you can’t give gifts to strangers, that’s just creepy. You need a certain amount of Friendliness before the NPC will be willing to accept a gift. And of course, when you’re friends, the NPC will sometimes offer you gifts in return.

You can increase friendliness even more by doing Favors. These are just simple quests. (They may still be very difficult in some cases, but in terms of mechanics, they’re just get-me-some-item(s) quests.) Every single NPC in the game has at least three Favor quests that you can unlock by making friends with them.

Favors, of course, don’t come with direct material rewards… they’re favors! But there are many benefits to friendship. NPCs who are skilled in certain things can help you in lots of ways, such as by teaching you new things, giving/selling raw materials or recipes, or even just buying your junk items that nobody else is interested in.

There’s also the romance angle, a la Harvest Moon: some NPCs can be “married”, wherein they appear in your house. (Of course, it’s possible for lots of players to be married to the same NPC! But who cares?)

Although Friendliness is a hidden number, the game does give you a general idea of their friendliness level at all times. I want to make it easy to “game” the mechanic. It’s not like real life, where people can be impenetrable and illogical. Game mechanics are only fun if you can see what you’re doing and how you’re affecting things.

Friendliness Rating is very simple, as you can see — that’s true of all these systems. But it can still have far-reaching implications, and when combined with a few other mechanics it can get interesting. For instance, I’ve been experimenting with “happiness factor.”

Happiness Factor

If an NPC shopkeeper is making a lot of sales and piling up a good income, they’ll be happier than if they’ve been standing around bored all day. But if a player with a good Barter skill gets an NPC to pay too much for some loot items, the NPC might get buyer’s remorse, becoming less happy for a while.

An unhappy NPC is less likely to do favors for friends, give good deals on items, and so on. However, you can increase their happiness by doing things like dancing for them, using social skills to cheer them up, or even using mental magic to force them to be happier (but watch out if the NPC realizes you’re doing that!)

Is this a fun mechanic? Eh, it’s okay, but its real purpose is to be part of the gestalt. Emergent gameplay comes from the combination of systems. Without other systems around it, this system would just be odd. But it combines with other mechanics to tie into skills, quests, combat, and crafting.

I have to admit that the main reason I like this mechanic is that it gives players ways to affect each other, indirectly — an NPC’s happiness level is global, so you can go around pissing people off or making them happy. (It’s a lot easier to make an NPC “happy” than to make them “sad”, so that griefing potential is low… but the effect is still noticeable.)

With a few more of these simple mechanics tied together, we get a nice meaty base for skills to interact with.

But Are They Memorable?

If NPCs are just sets of numbers, that can still be fun to interact with. But if the NPCs are memorable, the impact of manipulating their numbers is greatly enhanced. Ideally players will remember the names of many of them, might be willing to do favors for them, and would even care if they got kidnapped by frog men or smote dead by an evil curse.

My plan actually calls for a lot of text — probably about 60 lines per NPC — but they’re mostly terse reaction statements based on what skill you’re using on them and their current game state. So the volume of text is high, but it’s not anything like having to write compelling talk trees.

But this text gives me room to give them simple personalities. As I’ve said before, I think we should all be striving for two-dimensional personalities for our NPCs. Practically speaking, this usually translates into “quirks”. The pirate that hates rum and secretly drinks apple juice. The matronly older lady who runs a tailor shop, and secretly also runs a brothel. The troll under the bridge who likes candy. You shouldn’t always discover their quirks right away, either: they should have enough surprises that there’s some meat for players to discover via interactions.

I don’t expect the NPCs’ personalities to be so compelling that you remember them right off the bat, though. Remember that the game systems push you toward visiting the same NPCs repeatedly, so that I can take advantage of face time.

The Magic of Being There a Lot

Remember Final Fantasy 7, when Aeris died? This character’s personality was wilting, boring, and cliched. The reason we were shocked and saddened is because Aeris had been a foil in the story for hours and hours.

If the game had introduced Aeris just a few hours before she was killed off, even an incredibly interesting personality wouldn’t have made the death matter as much. I didn’t cry when Aeris died (I felt manipulated by the author, so I was pissed off more than anything else) but for a lot of gamers, this was an extremely important moment.

Yet if you asked them to write a paragraph about Aeris’s personality, I don’t think many people could. She wasn’t memorable because she was that interesting. She wasn’t memorable because of her story or the logic of the world, either. (Hell, given the laws of the Final Fantasy universe, a $100 item should have fixed her right up.) She mattered because of face time.

Now Aeris did have some personality, or otherwise the few dozen hours of face-time she got wouldn’t have been enough.

I think you can always use face-time to make someone memorable, even if they are lacking any trace of personality… but it might take a lot more time. If you watched enough Stargate SG-1, you’ll remember the guy who called out “Chevron 6 is locked!” whenever the Stargate opened. But you probably don’t even know his name. However, if he’d gotten murdered by a Villain of the Week, longtime fans would have been quite upset. Not because he was important, but because he was… well, on screen for literally hundreds of episodes of TV. But a bit of personality definitely helps speed up this process.

So my game systems increase the face time for NPCs, which helps give them importance to players, which in turn increases the importance of the game systems they’re being interacted with. That’s the plan, at least.

Keeping It Simple and Focused

All in all, the systems are very game-able — it is a game, after all. I want it to be fun. Players should be able to figure out how it works, and even if it’s a bit of a puzzle how to approach some NPCs, players should ultimately be successful, given enough time and effort. One trap that’s easy to fall into is opacity: it’s easy to make NPC systems that players can’t understand. When that happens, players aren’t having as much fun as if they understood what was going on.

I also realize that not every player will be excited about befriending NPCs. It’s not very heroic, at least in the classic “dragon murdering” vein. That’s okay; there’s enough things to do in the game that you don’t need to chat up every NPC to have fun. But even if you’re purely combat-focused, you’ll find that some combat skills also give you interesting NPC interactions, and then you may find you’re enjoying it more than you expected…

Next Week

I’m trying to maximize impact with only low-cost game systems — systems that are easy to code, balance, and create content for. Hopefully I can achieve that for NPCs (there’s some worry about the text needed for NPCs, but I think it’ll be okay… I just need to write some more of them to see if I can really sustain that efficiently). If it doesn’t work, I’ll change it until it does.

Next week I’ll talk about how I’m using this idea of “low-cost, high-impact” in the combat system. It has different bottlenecks and problems, but I’m excited about how it’s turning out.

There’s been some great replies to my last post, and I want to thank you for the ideas. I’m still sifting through them and figuring out what I can realistically make work, but I particularly liked these tidbits:

• The “rune” metaphor to explain item decay: it’s a lot more elegant than “sharpening swords” and it allows for infinite diversity of power-ups. Because sharpening a sword just makes it sharp. But a rune is maaagggicccc so it could do anything at all.
• The idea that items can be “melted down” somehow to provide temporary power-ups for other items. More directly, raw items may even be prerequisites for creating certain runes.
• The need to use different weapons for different problems (already a part of the combat design)
• Legendary items that are permanent and require no repair/allow no runes to be installed/etc. They’re very powerful but very specialized — so for instance you might find the epic sword of frog-man stabbing which is the best frog-man stabber that could ever be… it doesn’t ever need repairing. But it’s also not ver good against anything but frog-men.

And several others. I think a lot of people described what I already had pictured in my head, which means I didn’t explain what was in my head very well!

This all brings us around to the elephant in the room, though: given such finite resources, can I make this system as exciting as it should be? If I can’t do it really well, it should probably not be done: a half-assed item-decay system is far worse than no item-decay system at all. This is one of those mechanics where you need to “go big or go home.” Which I’ll talk about in another post shortly…

[This article is about my upcoming 3D fantasy MMO code-named "Project Gorgon."]

One of the many design elements I haven’t fully figured out yet is item decay. This is a tough one.

Item decay is when your weapons and armor and tools can get “used up” and disappear after some amount of use. It’s very different from “item damage”, where the items can be completely repaired. Decay is permanent, and is thus a much more complex topic.

The Case for Item Decay

Why would you have item decay in your game? There are lots of reasons to use this mechanic:

• It gives crafters more to do. If people need to buy new swords regularly, then blacksmiths have more to do. If swords don’t wear out, then blacksmiths obviously don’t need to make as many.
• It reduces twinking. A big problem in games with random loot systems is “hand-me-down weapons.” In these games, players tend to find more weapons than they personally need. They hoard these weapons and hand them down to other players. This doesn’t hurt much of anything if it’s done for an “alt” character, but when a brand new player is given a bunch of top-tier gear, it can really hurt their enjoyment of the game, and hence their retention. What seems like a nice gesture from players can really hurt the game overall. But if weapons decay, there will be fewer top-end weapons to give to others, and anyway, the twinked player won’t be permanently overpowered.
• It lets you keep your power escalation low. In most games you want players to experience the thrill of finding new equipment often. In a game with a steep level curve (like WoW for instance), you’ll get new weapons all the time as you level up. But in a game like mine where people don’t get a ton more powerful as they level up, higher-level weapons can’t get much more powerful than lower-level weapons, so what’s the point of switching weapons? Well, if your weapons keep decaying, you’re forced to switch weapons. Problem solved!

The Case Against Item Decay

The case against item decay is really simple, and really damning: human beings are ridiculously averse to loss. There’s more psychological studies about this than you can shake a library at, but for me it can be summed up in this one principle: Human beings fear loss, even more than they should. They predict that the loss of an item will hurt more than it actually does, and they go out of their way to avoid it.

In a game with item loss, there are two ways humans deal with this fear: either they refuse to use amazing items because they don’t want them to go away, or they become clinically detached from all items.

Probably the most common thing is that they just refuse to use items. You see this in any MMO that has one-use super items. I don’t know about you, but when I get a “Potion of Being Amazing For 3 Minutes”, I hoard that sucker forever. In most games, I won’t even use it to save myself from dying, because the death penalty is less painful than losing the item! (In WoW or EQ2, I wouldn’t use rare irreplaceable items unless I was in a really good group/raid and we were going to wipe… and even then, I’d have to think about it pretty hard.)

But this isn’t a big deal — who cares if people hoard their one-use quest rewards? I don’t. It gets to be a bigger deal when all items in your game can decay. Suddenly you never want to use any high-quality item ever!

Let’s unpack that idea of “high-quality items” for a moment, because it’s important. An item is not high-quality if it’s “just good enough to get the job done.” You’ll be happy to have such an item (especially if you were struggling with dangerously underpowered equipment prior to that) but you won’t think of it as particularly super-powered.

By definition, a top-tier item has to be more powerful than is strictly necessary to survive. And that means it’s never actually needed. And if it’s not needed, and is going to decay when used, that means it’s going to get hoarded.

And if you try make top-tier items mandatory in order to survive, what you’re really doing is making all lower-tier items into junk, and you’re back to square one: nobody’s excited about loot because everybody has to use the very best items already.

(I’m oversimplifying here, because it’s actually possible to create items that are “kinda crappy” without being suicidally underpowered. But I also know from experience how really hard it is to find that sweet spot for consumable items, and how hard it is to maintain that balance over a long period of game updates. It almost always falls down into items either being “junk”, “the thing we have to use”, or “the thing that’s too good to ever bother using”.)

Clinical Detachment

I said there were two scenarios that happen when you have item decay. One is that people hoard the best stuff because it’s too valuable to use. The other is even worse, though: emotional detachment from loot.

You see this in games with particularly heavy item turnover. Nobody gets too excited about even the most amazing items. “It’s nice, sure, <yawn>, but it’ll be gone soon enough.”

This means you (the designer) have a much harder time getting players excited. The game also tends to feel more like a treadmill: players can see more easily that they’re just collecting money in order to buy the same old stuff so they can collect money to buy the same old stuff so they can…

(Not that there’s not actually any more “treadmill” than any other MMO design. It’s just that item decay makes it a little easier to see.)

So once again we’re back to the same problem: we can’t have nice things, because there’s no things in the universe that are nice enough to get excited about.

If players aren’t going to be excited about loot, why bother having it? You’re better off leaving it out and focusing on something else. If you’re gonna have a complex item system, players need to be excited about items!

Gorgon Needs a Different Approach

So from a logical perspective, item decay is a great mechanic. Those problems I mentioned at the beginning are real problems, and very much worth fixing. But the psychology of loss is just too painful to me to want to do an item-decay game.

But since my main complaint is the psychology of it, maybe I can find a replacement design that fits psychology better.

And really, isn’t that what every game does? The reason your sword’s damage keeps going up and up in WoW is to force you to keep getting new weapons. It isn’t really much different than if your sword just decayed every few levels, forcing you to get a new weapon. But it feels so much less painful!

And WoW solves the problem of “hand me down items” with the trick of literally making it impossible to give great items to other people. The items become attuned to you as soon as you pick them up, and voila! No more hand me down items. It’s ham-fisted but it sure works.

But all hybrid solutions seem to have their own problems. In the WoW model, the power level has to keep escalating up and up and up forever, and at a pretty fast pace, too. If you ever get the very best set of items in the whole game, the treadmill breaks. There’s no reason to care about items anymore.

Hybrid Solution: “Permanent Plus Temporary”

I’ve been looking at different hybrid solutions for Project Gorgon. I’m calling my current plan “Permanent Plus Temporary.”

It’s a two-stage approach. Weapons, armors, and tools are permanent. So you can quest for the very best items in the game, and get excited when you find them. Hooray, you’re permanently better than before!

But by itself, the sword is kind of crap. It has to be magically sharpened (or maybe “enchanted”, or “have a new hilt put on”, or whatever — something has to be done to the item. I’m just using “sharpened” as an example).

So sharpening is not optional. An unsharp sword does so little damage that you won’t survive if you try to use it. You have to get your item sharpened, either from a professional sharpener or with a one-use do-it-yourself sharpening kit. And the sharpening “wears off” over time, requiring you to sharpen it again and again.

That solved one of the three problems I mentioned at the outset: it makes sure crafters always have enough to do. And I can fix the hand-me-down problem the same way WoW does it: for the very best items, you simply can never give them away. That’s not a particularly immersive or elegant solution, but it works.

But I still have problems with power escalation. Once you have the best item, you don’t need to look for loot anymore!

Lower Power-Escalation From Randomized Loot

In order to keep my items’ power levels from cranking through the roof, I’m using randomly-generated treasure. The idea behind randomly-generated treasure is that it uses a different reward schedule than fixed-design treasure. You have to look at every sword you get because one of them is going to be amazing. And when you find an amazing sword, there’s still a chance that you’ll find an even more amazing random item later.

So with random items, players rapidly reach a “sweet spot” of power (a little higher than the average quality level for items of that type), and then very slowly (on average) find better and better items. They have to keep looking because who knows when the randomly-better item will show up?

Lots of games have random treasure (and a few have random crafted items), but often the results are not amazing. Making a fun random loot system is a lot harder than it sounds! I remember that we struggled mightily to make Asheron’s Call 2′s treasure system the equal of AC1′s. In the end it was very cool… but it sure took a ton of time and tweaking, and it never quite had the elegance of AC1′s system.

The problem is that you can’t just graft a random treasure system onto a game; it has to be designed around it very heavily, and with a lot of subtlety. I’ll talk about random treasure another time, though, since this post seems never-ending…

(Random loot still doesn’t actually “solve” the power progression problem. It just slows the progression down a bit so it doesn’t have to keep going up and up and up every damned day.)

Better Plan Forthcoming?

I don’t feel like I’ve necessarily got the best plan yet. I’m still a ways away from actually implementing a treasure system (beyond the static loot that I have now for testing purposes), so I still have time to bat ideas around.

I’m interested to hear your thoughts on the pros and cons. There are a lot of different angles you can look at this problem from — the trick is figuring out what angles are the most important.