Friday, January 28, 2011

The Opt-Out

This is a rant. Maybe an unfair one, who knows.

So, I made a purchase from Etsy for Christmas gifts. I kinda like the place, and I have friends that do quite a bit of business through them. Today, I get an email from Etsy.

"We recently launched a new feature, Circles, that lets you connect with other people on Etsy. When you add someone to your Etsy circle, you can follow along with their favorites in your activity feed. It's illuminating!

Right now it's hard to find people you know on Etsy, and that's sad. Well, we're changing that. We're making it easy to connect your email address book to Etsy, so we can find people you know who are also members.

(If you don't want people you know to be able to find you, you will be able easily to opt out through your account privacy settings.)"

Wow, that is not at all something I'm even vaguely interested in. Okay, lost of good sites have features I don't care about. But read carefully: this is not a feature I don't care about. Turns out, I care deeply!

Let's review the severe problems in this message.

The only obvious issue is that they have decided to make my email address a public part of my account. This only becomes a real issue because they also automatically opt you into sharing your purchase history. Well, brilliant.

I can see how many people would find this a good new feature. For example, my friends that do a lot of sales through Etsy. But me? I find this an extreme breach of privacy. To the point where I am now significantly less likely to use Etsy.

I logged in and opted out of all that shit. Turns out there's no "delete account" option, by the way.

"Easy to opt out" is an oxymoron. "Opt out" is a poison term, it already means you raised a barrier specifically to make it more likely people will participate. Nothing can be "easy to opt out", because that's called "opt in". "Opt in" is the only responsible, ethical way to reveal client information.

The only points Etsy gets here is that they sent this email before enabling the new feature - it doesn't go live until next month. I don't think Etsy was trying to be sleazy... But linking personal email addresses to their purchase history in public? Yeah, that's not shady!

I'd stop using Amazon in a heartbeat if they did that, and I practically live through them.

Of course, I'm pretty touchy on privacy. I'm the guy who refuses to get a Facebook account. And I'm not really Etsy's target audience, or even a good customer. Still, it's a shame they took a sleazy route. I guess most people don't mind an utter and almost criminal lack of privacy. At least, I hope so, or Etsy's in trouble.

It's possible this is not sleazy, and they've just been really horrible about explaining it. However, that seems vanishingly unlikely given the reckless disregard for privacy shown by nearly every company these days.

Grrrr.

Terrain Play

I've talked about this before, but I'd like to talk about it again. Constructive terrain games. That is, games where you build a base/s. Whether it's Starcraft or Evil Genius or Civilization.

In thinking about these games, I've tried very hard to come up with the methods used to make them interesting and challenging. The challenges I'm most interested in are those related to building the base, rather than those related to putting the base in the right spot on the map. So I'm not going to talk about map-tile resource allocations or complex topography much. I'm going to talk about the parts where the player twists himself up in a knot of his own making.

There are three kinds of challenges that relate to this. Most of these challenges are expressed via semipermanent, fixed structures that you can place.

The first is the arbitrary challenge/facility. This is a situation where the location on the map doesn't matter. For example, placing a research facility: no matter where you place it, the research happens at the same rate and is just as applicable.

Arbitrary challenges are often simple treadmills to extend the gameplay. Plunk this down to steadily unlock stronger soldiers, if you have the money for it. However, they can also serve as a gating mechanism, unlocking additional facilities over time. This forces the player to build up their core base with a complexifying crowd of fundamental buildings, making the placement of advanced buildings somewhat tricky. Advanced players will often leave empty zones in their base as they build, knowing that they'll want to put advanced structures in safe places well inside their core base zone.

Arbitrary challenges are usually made a bit trickier simply by resource allocation. An arbitrary challenge facility will siphon away valuable resources, and your strategy requires you to balance the upgrades/abilities it offers against your more direct unit/facility purchasing.

They can also be made trickier by having very awkward shapes. This is especially useful if you're playing a contiguous base game, where every facility must be butting up against another facility, door-to-door. While any facility can have an awkward shape, this is the best type to make awkward, as the other types should be reasonably easy to place anywhere.

Speaking of the other types, in addition to arbitrary challenges, there are ZOC challenges.

Zone of control challenge facilities are those that have a radius of effect, or whose effect diminishes as range increases. For example, you put the lumberjack shack next to the forest so the round trip is tiny. You put the turret on a path that the enemies are likely to use.

A subset of ZOC are those that are required in order to place your own facilities, such as Protoss pylons or Dune's concrete. I'll call these "extend ZOC". These are worth mentioning as their own distinct type simply because if they exist, they are critically important parts of your strategy.

ZOC challenges sometimes relate to fixed map resources, in which case the point is to build a ZOC facility very near the resource. Combined with complex topology and extend ZOCs, this can actually be quite a challenge. Protecting a particularly remote facility may also be quite a challenge.

Other ZOC challenges relate to non-fixed map elements, such as traveling enemy soldiers or random resource spawns. Knowing the paths/likely positions of these elements is critical, and there are often "layers" of these kinds of ZOC challenges. For example, this turret protects against ground units, that one against air units. You place them assuming the ground units will go around the mountain and the air units over it: the ZOC has very different weights because they are addressing different kinds of non-fixed map elements.

The third ZOC challenge type is when the ZOC relates to your own base. Extend ZOC facilities are one example, but there are often other buildings such as police stations, repair droid ports, energy broadcast stations, sewage treatment facilities, and so on. Games with a lot of my-base-ZOC challenges are typically not heavy combat, but are instead heavy on building.

After arbitrary and ZOC challenges there is a third type: pathing challenges.

Pathing challenges are when the point is to direct units or resources along specific paths. For example, walls of sandbags to force the zombies to come along a torturous maze. Or electrical lines to carry power to your distant buildings. Or roads.

Pathing challenges sometimes intertwine with arbitrary and ZOC challenges. For example, any contiguous base game makes the whole base into a pathing challenge, where you may have to be quite clever with your base design and where you extend to. Pathing challenges may also interact with ZOC challenges, especially if you are putting down a path for enemy units: the ZOC facilities will cover the path and bombard the enemy.

However, the deepest kind of pathing challenges are those that change over time, or butt heads with changing situations.

For example, in Evil Genius, you have a contiguous base. Your minions walk through the base in a strictly simulated manner. The pathing you provide in the early game will rapidly become unwieldy as you struggle with larger numbers of minions and desperately trying to keep travel times short rather than putting that new facility a mile away where there's actually room.

Oldschool RTS players will probably remember difficulties with base placement in old games: you would build a base, then realize you had cut the travel path, and units couldn't actually get out. Another example.

This is relatively rare with things other than travel paths, though. Electrical lines can always carry infinite energy, water pipes can always carry infinite sewage, and so on. Still, that's just convention, and it can be bucked.

Another fun complexity to add to pathing challenges is to have multiple kinds of paths that have to coexist. For example, roads and railways. Or even just local roads and highways. Underpasses and overpasses become common, and as the situation changes, the roads inevitably become too limited and narrow. This is especially common in train games.

There are plenty of other ways to make pathing challenges interact with each other if you want.

The point is, if the game wants to let the player twist himself into knots, the game has to allow the player to build his own base, and make that the fundamental restriction on his future expansion.

The core elements of that are:

A) Gate facilities, such that you have to build certain buildings before you can build others.

B) Use ZOC (and possibly travel time on travel paths) to force the player to consolidate.

C) Use travel paths with load limits and steadily increase the load. Optionally use multiple kinds of travel paths that can't perfectly coexist.

...

There's a subcategory of games where you build your base for another kind of objective entirely. For example, the self-styled bases of Minecraft players. Technically, they use these same kinds of challenges, but the weights are very odd and often selectively imposed. Not every base has to have water, for example, so the water pathing challenge is often simply ignored.

These "erratic challenge" systems are another topic for another day!

Monday, January 24, 2011

Shadow Puppets

I came up with this idea while trying to invent a hacker game setting, and I thought it was an interesting thought experiment. So here's the deal:

Imagine that in the future, we have software that can analyze audio and video (and writing and pictures). By analyzing the media, the software can program a virtual entity with a personality and appearance very similar to the writer/actor/whoever is in the media you passed it.

The virtual entity is quite advanced. Assuming you have a good computer and the most recent pseudo-AI drivers, the virtual entity is almost indistinguishable from the original, at least in the context of media. The gaps in the seed media are overcome by careful guessing. That mean the virtual entity can act very reasonably even in situations not covered by the seed media.

In the game world I designed, problems arose when people began creating virtual entities of politicians and releasing very convincing fake videos of them saying or doing things they never did or said. It came to a head with a bevy of fake-porn cases in which the real versions of the virtual porn stars launched a whole lot of lawsuits for everything from character defamation to abuse.

The courts came up with a rather novel solution. Which, I guess, is the most fantastical part of this setting. It's called the "virtual volition" rule.

Basically, the courts ruled that everyone had the right to personal privacy. The programs were considered to breach that privacy by simulating that person's behavior and body in a way they did not approve. This was considered fundamentally different than drawing cartoons or even photoshopping lewd pictures because of the extremely high fidelity and accuracy of the virtual entities.

The misused virtual entities were labeled "shadow puppets" - programs that looked so much like specific real world individuals that they violated the privacy of those real world individuals. Shadow puppets became illegal. Later, the definition expanded to include virtual entities in breach of copyright, such as one fed on Mickey Mouse cartoons that looks and acts like Mickey Mouse.

Of course, it's impossible to ban a piece of software like this. Legal, open source software could be fed with publicly available media to produce a high-fidelity shadow puppet. Still, don't get caught: it's considered abuse or assault on the person being simulated.

Legal virtual entities abounded. Many simply were unrealistic, or a mishmash of media that didn't attempt to simulate any one person. Others were specific people or cartoons or whatever, but had very strict limits and would refuse to "fill in" the gaps and act beyond the "volition lock". IE, you couldn't convince your official Steven Fry virtual entity to strip down and sing show tunes.

In the game world, it had tons of ramifications and made the world a very complex place.

Anyway, there's not really any point to this essay. It's just a random interesting thing I thought I'd mention.

Sunday, January 23, 2011

Connectivity

I was watching this archived webcast today. It gets contentious about halfway in, but I like that, it makes me want to write about the subject. This is a borderline rant, I guess, but I hope you enjoy it.

A lot of the conversations we have about the nature of technology stresses, well, stress. Just as an easy example, the author of that webcast specifically calls it "panic architecture" which, like many other terms, is coined to make a point rather than be useful.

Assuming you don't want to listen to the hour of webcast, I'll cover the bits I touch on. In this case, "panic architecture" is used to refer to the theoretical stimulus overload from tweets and emails and IMs all popping up at you all the time.

I can't help but turn down this kind of shallow analysis. Not only is there no real evidence to support it, but there's also no sign that these clumsy first-generation implementations are the only ones we're going to see.

We've developed tools to let us share with everyone, everywhere, instantly. We're now using tools that let us "hear" without having to be listening.

Every kind of "panic architecture" interaction is actually asynchronous. That is, the computer tells you that there's a new message, and you can get around to reading it whenever you like. Some pieces of software are worse than others - Twitter is only barely asynchronous, since if you wait more than an hour or so, you'll have a backlog too deep to sift through. So you lose a lot of data down the tubes if you wait.

I guess you could say that's a high-pressure environment, but not really. The only reason it's high-pressure is because you've tricked yourself into thinking it's important to view everything, and view it fast. Trying to make sure you don't miss any tweets is sort of like trying to make absolutely sure you see every single person who passes by the Starbucks. It's not a sensible reaction.

If we feel stressed by these "always on" tools, it's because we've tricked ourselves into viewing them in the wrong way. Part of that might be pure neurochemistry, but I think it's mostly a cultural maladaptation which will fade in time as we get used to the technology and learn, as a populace, that it's okay to let the river flow by without trying to drink it all.

Also, there is another generation of tools past these "hear without listening" tools.

Phone calls extend our ears so we can talk over long distances. But that's only this very instant. So we use emails, RSS and the like to extend our memory: we can think about something long after it has been said.

And now we're going to get a new breed of tool: one which extends our thinking about something across vast distances. Already, we're seeing it. We don't get all our emails: we get only the emails our outboard brain doesn't discard as spam. Many of us have dozens of rules about which emails are further flagged and sorted. This lets us determine whether something needs to be read soon, or whether it can be put off until we feel vaguely curious.

In the future, more intelligent tools will have a system for cuing things up. If it's important, it tells you immediately. If it's interesting but not important, it'll bother you with a quick line and link every few days. If it's something good for you that you can't seem to sit down and get through, it'll keep pestering you.

Essentially, this new breed of tool is a computerized version of the thing in our brains that says "hey, the roof is leaking, maybe we should fix that before it rains again."

Right now, we often try to deal with a true deluge of information right against our brains. I agree that this is not really a very good idea. At the very least, it sets a very low upper limit to the amount of things we can consider. By creating smart tools to deal with that deluge for us, we can consider terabytes of information every hour.

I need to stress that this isn't some fantasy tale of strong AI. We already do this to a great extent by using RSS feeds - we sign up for the news outlets we like, whether they're brand names or individuals with blogs. We use them as our roof-is-leaking engines. I'm simply pointing forward and saying "there's gonna be more of this." For example: Priority Inbox

It is true that humans sometimes feel lost and confused by advances. But this is nothing new. People felt lost and confused when the king changed, or weirdos with funny accents started settling in their alluvial plains. Change makes people feel stressed out, regardless of whether it's technology or culture or any other source.

So, yes, a lot of people feel stressed out by the "constant bombardment" of our "always on" culture. Except that it's not constant bombardment, it's not always on, and the tools for dealing with it gracefully are evolving as we speak. As always, we adapt to our situations and find out the best way to take advantage of the opportunities we are offered.

To me, the real "panic architecture" is the way soft sciences and media outlets continuously scream about dangers that never arrive, dangers that were never even dangers. It interferes with adult's acceptance of change when you scream that every change is a disaster. Kids don't give a shit - a kid will use Twitter and Facebook even if their parents think it is pure poison. But adults have to work to learn these tools, to integrate them into their life even a little bit, and panicking about them just makes the task that much more difficult.

I don't, by the way, think that the webcast I mentioned at the beginning is an example of that kind of panicky statement. I don't know enough about the author to say anything about her, and at worst, the webcast itself is kind of ambivalent on the subject. It's just a handy turn of phrase to steal her words and make another point with them.

There are actually a lot of other things I want to argue about that were covered in that webcast, but this is long enough already.

(By the way, I'm going to coin a term in opposition to the normal way of coining a term. "Roof-is-leaking engines", or RILEs: that's a term that could be useful, but it isn't very good at making a point. )

Tuesday, January 18, 2011

Tool Using Play

After a brief conversation about why some people don't consider Minecraft a game, I'd like to talk about it. Not Minecraft - about the use of play as a tool.

There are a fair number of games that feature content creation as a major play element. There have been some pretty much since we figured out how to draw space invaders to a computer screen. Let's talk about the way games integrate content creation.

In the "classic" style, games integrate content creation as a stand-alone editor and/or importer. For example, Pinball Construction Set in 1985 (one of my favorites as a kid) had a level editor you could play with. It's become pretty typical these days.

Many modern games feature an edit mode switch in the game proper. For example, Little Big Planet and Second Life both feature an "edit mode" where you can create and alter objects from wherever you're standing in the level.

But the method steadily gaining in popularity is when the edit mode is tightly integrated with the gameplay: as you play the game, you edit the game. An obvious example is Minecraft, which is basically impossible to play without editing the level. In such cases, you initially use the editor to survive and explore a bit, but once you get good enough, you begin to use the editor to do increasingly self-driven things, such as building a giant garden or the USS Enterprise.

Another example of integrated content creation is when the act of playing the game ranks content and makes other players more or less likely to see it. Later, I foresee a fair number of games where you'll run into random mutations of content and rank it simply by playing, allowing a fairly large number of players to participate in an evolving ecosystem of content. So you don't absolutely have to be creating content yourself: ranking and sorting content is a valuable form of content creation in and of itself. Well, assuming you have the requisite foundation of content to rank and sort.

In my mind, this final form of content creation is the future. Integrated content creation. It has the highest "pull" - you'll get more content per player (on average) and you'll get better sorting and distribution if you use integrated creation and ranking. Much more efficient than mode-swap content creation, and almost infinitely more efficient than level-editor content creation. (To be fair, I think most games will use all three to some extent: they have different specialties.)

Since content creation / player-sorted content is such a powerful tool, I expect to see more of it in future games. Therefore, how to do it best is also an important topic.

I'm not as expert as I would like to be on this matter, but here are three areas I feel should be on everyone's list:

Base content creation: how does your player base create content as they play? The more restrictive the framework, the lower the barrier to entry... but the lower the maximum quality of the content will be. Minecraft uses a rigidly square grid and almost no moving parts, allowing players a fairly easy time of it, but also crippling some high-end content. For example, no NPCs, no elevators, no swinging pendulum traps...

Distribution and sorting: how do your players share content? The more helpful your distribution system (IE marketplace), the happier your moderately skilled player base will be... but it will also be very crowded and full of crap. A sorting mechanism driven by player preferences can be very handy, but can also brutally suppress starting players who are still getting used to the editing system.

Content cooperation: how can your content be used in relationship to other content? For example, create a shirt, and you can put that shirt on any humanoid frame. That's cooperative content. The more ways content can interact with other content and/or rules, the deeper the system will be. If you create an axle with wheels, can you only use it with cars? Or can you staple a hundred of them to the bottom of a house? If you create a shirt, can it be torn up, made wet, transformed by the environment, mutated, tailored, or painted? Obviously, the more kinds of interactions content can have, the more potential to draw players in.

The primary drawbacks are A) the amount of time it takes to code those interactions and B) accidentally forcing players to wade through details when they just want a quick result. IE, if they're making a skyscraper and you force them to choose every door, every carpet... that's a big barrier.

Anyway, that's content creation as I see it. You?

Thursday, January 13, 2011

Skyrim

I posted a very dismissive comment on Skyrim, and I've gotten a fair number of people telling me not to be so dismissive. So, I'm going to talk about it a bit. By "talk", I mean "rant". (Edit: I really mean rant. This is a way over-the-top rant. I only clarify this because now there's some outside links leading here. This is not a measured discussion of merit. It is a rant.)

First, I'm sure Skyrim is going to be an extremely popular game with very high ratings. It'll certainly get game of the year. What I'm saying is that you will enjoy Skyrim if you're even vaguely the sort of person who likes things like Oblivion or Fallout 3. Hell, even I might enjoy it, maybe. It's not a gamble, it's not a mystery, there's not even any need to see gameplay videos. It's a well-funded game from an experienced studio that has put out many of the most successful games in this genre.

But let me tell you why I don't like it.

There's a stench crawling up from below, and I smell it in every press release and article. The stench of... streamlining.

Let me start with the most obvious example: they've taken out the the "restrictive class system" and allow you to gain points in skills just as you use them. Of course, the earlier games had a class system... and still improved your skills as you use them. So we can ignore the "improve your skills as you use them" chaff. All they're doing is explicitly removing classes.

At first glance, you shrug and say "why not? Sounds fine. A lot of successful games like Fallout 3 don't have classes. They're an outdated design philosophy."

So, let me ask a simple question: what was your favorite part of Oblivion and Morrowind?

So far, most of the people I've asked have said either "the first ten hours" or "character customization". Which are basically the same answer.

I played Oblivion for at least two hundred hours. The vast majority of those hours were the first ten hours over and over. This was possible for several reasons. One: character design is fun and has ramifications. So you can simply change your design and be playing a different game, right from the start. Two: you can easily strike off into the world in any way you like, so you're not stapled to the same starting narrative. We'll leave that bit alone for now, except to say that you'll probably be stapled to a starting narrative in Skyrim.

Character design - choosing not just your visual look but also your stats - is a core part of RPG gameplay. The "class" was simply an extension of this, allowing you to radically alter your growth curve and quickly develop different ways to play the game. If I want to play a warrior, I walk out the door as a warrior and do warrior things right away. If I want to play as a thief-mage, I roll it up and run with it.

On the other hand, in Skyrim that's not what you do. At least according to the press releases, you start with potential in everything. Like Fable, perhaps.

It is still possible to become a warrior, or a mage, or whatever. But not until you're ten hours into the game. The first ten hours are almost always nearly identical. You can't skip them: you need to grind character skills during them.

Oh, yes, those restrictive straightjacket classes. Oh how I hated the way they let me play the game any way I wanted. I really hated being allowed to be competent right from the starting gate. It's much better to ditch replay value in favor of locking characters into a sluggish, long-term path that you won't know whether you like or not until ten hours in.

...

I guess that might seem a bit overstating it, but it's not. That's the purpose classes serve, and the damage done by ditching them.

There are a million things I react poorly to. For example, they say that there are X spells - a specific number, something like thirty.

Whut.

Um, why? Where did spell creation go? I really hated being able to create my own fun custom spells! I especially hated being able to enchant stuff, I hope that's gone, too.

They mention there is a focus on blunting and preventing exploits.

Um, why? A single player game doesn't have to be balanced. The exploits are a big part of why all the other Elder Scrolls games were fun. The only reason I can think of to aggressively balance the game is that there will be some kind of ranked multiplayer situation where game balance suddenly matters. I actually hope that they just did it out of a random brain fart instead.

They mention they dropped mysticism. Because who needs a school of magic about magic? Well, obviously, not our avatar. Since he can't roll his own spells or anything.

Three fewer skills? Stinks the same way. Let's guess: is alchemy one of these missing skills, or is it just blunted to the point where you might as well just buy healing potions from the shop?

My complaints could go on basically forever. I have loads of more tenuous concerns. For example, this being a cross-platform RPG, it will almost certainly have "the Console Blight". This is a disease where custom content is verboten, and PC installs come with a gig of malware to keep you from accidentally having any fun. Sure the earlier games didn't have it so bad, but that was then. There's a universal trend towards the console blight, I hardly expect this game to ignore it.

The menu system is mentioned as being more "in game". Like with Fable III and its hilariously, game-breakingly bad "in game" menus? They mention the "horrible" menus from Oblivion. Which menus were those? The ones that let me see the things I needed to see within two clicks using distinct icons? The ones that didn't make my character seem like he was having a psychotic hallucination? The ones that didn't take a second to scroll between submenus? The ones that actually allowed for complex gameplay?

Ugh!

So, yes, Skyrim. Game of the year, guaranteed. You'll love it. Maybe I'll even like it. But it throws away a lot of good stuff in the name of "streamlining".

Because players are retards! We hate designing our characters, having fun with exploits, and playing with complexity. I mean, sure, those'r universally considered the best parts of your earlier games, but those people were all deluded.

... it's been a while since I've done quite so much ranting on this blog. I'm probably the only person on the internet who dislikes what I see about this game. 50% marketing speak and 50% lowest-common-denominator game design. Ugh.

Monday, January 10, 2011

Going Home

Some friends and I had a bit of a talk about home bases in video games. There are many examples. The Normandies in the Mass Effects. The Suikoden games often had castles or similar. The Overlord games had castles. Mario Galaxy had the space base thing. Many more.

In talking about them, we found we all had completely different opinions. So I started thinking, and I came to some conclusions.

First, there are four distinct functions of a home base. These aren't exclusive: you can mix and match. The four types are hub worlds, enhanced menus, art galleries, and gardens.

Hub worlds are simply homes where the point is to have a central place from which you can go anywhere. Hub worlds are usually a "safe" point in the game world, allowing you to recharge, recover, and save. Their other function is to shrink the size of the world by allowing you to reach various points very quickly.

Enhanced menus are homes which could be replaced by a simple menu system, but are instead spread into various rooms. For example, you might have a travel room, an equipment room, a few shops, and so on. These are getting more common, and the most egregious example is the newest Fable game. A less nasty example is the Overlord's castle.

Art galleries are homes where you can make all sorts of aesthetic changes, such as doing interior decorating or displaying trophies. These are becoming extremely common, probably because they are both easy and effective. Many of my friends - and many gamers in general - love the idea of expressing themselves by redecorating their homes.

Gardens are homes which grow over the course of the game. These homes reflect your progress through the storyline(s), but also frequently allow you to access optional content such as unlocking NPC personal quests, growing/crafting rare items, participating in challenges, and so on. The core idea behind a garden isn't that it allows you to do these things, but that it steadily grows in its capabilities to allow you to do these things.

The Normandy in ME2 is a weak but easy example: there are many subquests to do, and as the game progresses, you get more. A better example would be the starship Calnus from Star Ocean: the Last Hope. The Calnus allows you to change who is rooming with who, wander around randomly and talk to your crew, forge items, and many other things. The potential for all these grew over the course of the game. That said, the Calnus was an extremely high-pressure implementation, since it is very definitely possible to end up missing a huge amount of content if you don't have a walkthrough.

Anyway, those are four things a home base generally can do: hub world, enhanced menu, art galleries, and gardens.

If you want your home to have an emotional kick to it, you need to make sure to allow for both art gallery and garden elements. A key is to keep the player coming back to the home for the sake of interacting with the home. This can be pushed a bit by enticing the player to come back for other reasons (such as needing to shop at the store or open a new gate), but those enticements are not sufficient in and of themselves. The player needs to want to interact with the home.

Another factor to consider is the architecture of the home. Architecture is important for several reasons. One reason is that people remember rooms and places really well, so you need to have unique and varied rooms and places connected in interesting but intuitive ways. Another reason is because cunning architectural design can radically enhance the art gallery aspect of the game - for example, if you change the flags in the main hall, that's fine. But if you then notice that you can see down into the main hall from the top of the tower, that change becomes a bit more interesting.

Anyhow, just some thoughts on creating home bases. What are your favorite home bases? Your thoughts?