Note: If you’re not into game development this post will be pretty boring. So you may want to just skip this.

Introduction

In traditional game development there are usually two parties: The Developer and the Publisher.  There is no exact definition of the responsibilities for each party so I will only discuss our experiences with it.

The Developer

The Developer is responsible for creating the game. Their responsibility is to provide the publisher with a title that is reasonably functional and adheres to the agreed upon design. They are also responsible to address bugs and other defects identified by the publisher.

The Publisher

The publisher typically gets the lion’s share of the revenue from the game so in turn they have a greater responsibility. Their job is to make sure the game is good. If a game is not good (buggy or not fun) it is the publisher’s fault, not the developer. The publisher is supposed to provide the quality assurance and have the distance from the development process necessary to be able to stand back and look at the game objectively. They also are responsible for the marketing, support, promotion, sales and defect releases. A customer with problems contacts the publisher and the publisher puts this together as a defect report for the developer.

Stardock: The Developer but NOT the Publisher

We have developed games in the past where we were not the publisher. The Corporate Machine, Starcraft: Retribution, Business Tycoon, The Political Machine, Galactic Civilizations.  Our experience with different publishers varied greatly but in each case we can say we had a very positive experience except in the case of Strategy First who didn’t provide any real QA or feedback and didn’t pay royalties. Ubi Soft and 2K Games were both excellent with Ubi Soft providing outstanding marketing support for The Political Machine and helping considerably with QA.

Stardock: The Publisher but NOT the Developer

In my opinion, this is where we have done the best historically. A studio also doing publishing provides a level of feedback and QA not normally available to publisher-only companies.  The initial concepts for Sins of a Solar Empire and Demigod, for example, were radically different than what was eventually released. In both cases, Stardock has helped with development (i.e. source code sharing) which is not the norm but made a huge positive impact.  While Demigod had networking problems on launch, this was largely the result of the sheer volume of players versus the scoped infrastructure (i.e. it was a network engineering issue that neither Stardock or GPG had a lot of experience in).

But in terms of internal game review, both games benefited greatly from the game review process in making the games better.

Stardock: The Developer AND Publisher

Ironically, Stardock has a much less even record in this regard. With the benefit of hindsight, it isn’t surprising. A lot of benefit of the game review process comes from not having an emotional investment into the game “as-is”.  Otherwise you get this:

It’s just a little slimy. It’s still good! It’s STILL GOOD!

Literally one year ago, Stardock finally did something about this problem by explicitly breaking the groups up. This required quite a few new hires that was a pretty big investment for a small company that didn’t anticipate having any new games out for a year. But it was important.  Fallen Enchantress is the first title to go through this new process.

Example: Fallen Enchantress 0.42

WARNING: SPOILERS! TURN BACK NOW IF YE DON’T WANT TO LEARN THINGS YOU MAY NOT WANT TO KNOW…

.

.

.

.

.

.

.

You’ve been warned…

.

.

.

.

.

This is a typical UI comment that’ll get sent in and logged. Nice to have means “If there’s time”. The other request is more pressing.

Nothing new here. Just an example of a missing asset.

This is one of those XML / programming things. The game sets the minimum distance between players but there is no XML to define typical distance. I was on a large map recently where all players were bunched up in a corner.  You don’t want to make it so that players know that no one else will be near them but at the same time, you want to generally have some distance between players if you can so that they can build up. You can always add lots of players (I think our limit is 32 players) if you want a brawl.

Some UI requests.

Information UI request.

Another UI request.

The process you see here is the same with titles we publish. But it wasn’t really present before on games internally developed because the Executive Producer was also the lead designer (i.e. me) which is an inherent conflict of interest.  I’m much better at complaining about other people’s designs than my own.

All this has happened before. All this will happen again.

I first started moderating “online communities”back in the mid 80s. Back then, they were called BBSes and the title “Sub-Op”.  Without moderation, Internet communities ultimately die off for the same reason: They become toxic. 

What causes them to become toxic is predictable and preventable. Identify and remove people who are toxic before toxic cliques form. It’s easier said than done because every time I’ve seen it happen, the people causing the harm have no idea that they are the problem. That’s because, in the beginning, they weren’t.

In the beginning of a successful online community, you have a core group of prolific, high quality contributors to the forum. Their contributions help the community grow and thrive.  In time, a certain percentage of these users will believe that they are entitled to judge others and abuse them if they are found wanting.

Eventually, the entitled group will cross some line that the owners of the site object to. This confrontation leads to the debate on what rights the site owner has over the community. And this leads to the following flow chart:

I’ve seen all three paths over the years.

The end result is the same, it’s just a matter of how long it takes to get to the end point. Of course, the other option is for the site owner to get fed up and shut down the site.

WinCustomize Example

I took the hands off approach for WinCustomize.com at one point as some users may recall. I subscribed to the view that the community had evolved beyond the site owners and could police itself. That merely led to many of the “old guard” bullying those they disagreed with.

Many of the problem people were friends and I would message them to lay off with the response being “That guy deserved it”.  That is usually how you can tell you’re going to have problems with someone because nobody deserves to be griefed.  (Anyone remember the “Community” er “The Council” and the first GUI Olympics? That still stands out as the most blatant example of entitled).

Neowin Example

Neowin.net took the path of making their Entitled group into moderators. And for awhile, it was good. But pretty soon, the same problem came up except now the Entitled group had the power to ban users. The site owner had to step in and it was doubly painful because the moderators had built a considerable following who departed with them. That one I watched from the sidelines.

JoeUser Example

JoeUser.com took the third path a few years ago as some reading this may recall.  In a blogging community it is particularly tough to lose people whose posts are the primary source of traffic. I won’t mention the names but anyone who’s been on here for awhile know the people I speak of.

In every case, the newly banned go to or create some splinter site. I’ve never seen one thrive (they often stay around in a kind of near death for years to come).

I wish I could say I knew what specifically causes people who start out being outstanding contributors to a community to becoming toxic users. They are usually very intelligent people. The only thing I’ve seen them have in common is the belief that some people deserve to be griefed and that they are entitled to make the determination on who deserves the griefing.

As we close in on the beta of the second game in the Elemental universe (Elemental: Fallen Enchantress) we’re doing a lot of “play” testing.  With that in mind, here are a couple of videos of me playing with some commentary.

As has been previously mentioned, everyone who bought Elemental: War of Magic in 2010 will get Fallen Enchantress for free and get the beta added to their account this month.

That’s right: 145 CPUs.

In mid 2010, we released Elemental: War of Magic. This was our first attempt at: [A] a fantasy strategy game and [B] a game that involved terrain. This produced…mixed results in that when released it was buggy (like on nVidia 7900 cards you get a white tactical battle screen which caused a particular PC Gamer UK reviewer to claim the game was totally broken) AND even once fixed it was…meh. v1.4 is decent but it’s no Master of Magic.

So the question is, what do you think we should do to describe Fallen Enchantress?

It exists in the same world but it’s a very different game. It’s not an expansion since there’s not much of War of Magic specific stuff in it.  But I hasten to call it a sequel because sequels imply a continuation of an existing series of gameplay mechanics none of which exist.

So what do you suggest we do to describe it to new players?

This is a very good study for people who design websites:

http://shine.yahoo.com/power-your-future/eye-tracking-study-shows-people-really-look-facebook-223900573.html

In “real life” I’m not particularly political. I don’t donate to candidates or political causes, I don’t put signs on my lawn. I’m interested in the discussion of politics in the same way that other people are interested in the discussion of sports.

If you’re come here from a link claiming I’m some sort of political lunatic, a few things to bear in mind:

When JoeUser.com started in 2001, I posted intentionally provocative articles. The idea was to get a discussion going.

As time went on and I became better known, this started to become a problem as people not familiar with JoeUser.com started to link to cherry picked posts and use them as forum fodder for my work as a software developer or business owner.

The analogy I would use is imagine if one of the writers from Cracked.com later became really well known years later and people, not realizing the nature of Cracked.com started using his articles as representative of what that person’s overall views are.

Now, I’m using an extreme example and for the record, I consider myself politically conservative. But hopefully most people reading this are intelligent enough to make the distinction between an article written on a discussion forum for the sake of getting a discussion forum versus a personal blog (JoeUser.com is a community blog site, it’s not my online diary).

Over the years, I’ve had to make private some of the earlier year posts because I got tired of having to explain how much of what I have written is tongue-in-cheek.  Such as…

• No, I don’t support converting homeless people into food for the hungry even if I suggested it in a post back in 2002.
• No, I don’t actually harvest a kidney from each of my children so I can drink poisons safely.
• No, Social Services has never actually visited my house.
• No, I don’t think Obama is Hitler.
• No, I don’t think Bush is Hitler.
• No, I don’t think Hitler is Hitler.
• No, I don’t think liberalism is a disease.
• Etc.

The reality is, there are people out there that are always looking for an excuse to be offended or outraged.  If you post enough things over enough time, you can always cherry pick something that you can use to pain very broad strokes about them.

Greetings!

I just wanted to give all skinners fair warning that we WILL be having a GUI Championships competition in Fall of 2012.

The categories will include:

* WindowBlinds skins

* IconPackager themes

* DeskScapes animated backgrounds

* Static Desktop backgrounds (traditional windows wallpapers)

* And Possibly, depending on when Windows 8 ships, skins for a new, unannounced, Object Desktop program

Tom Chick, owner of QuarterToThree.com just finished a controlled clean up of his forums where several prominent users were shown the door.

Moderation is a very delicate and thankless art.  A community can easily be taken over by cliques of griefers without careful moderation.

I think JoeUser.com users can remember a certain user by the acronym of LW who was a very prolific and interesting poster but yet couldn’t keep herself from griefing those she disliked.

Over on Neowin.net, we’ve struggled with this issue for years on how to have a community that is interesting and fun but can keep the bi-polar crowd from getting out of control (as someone who tends to be a bit bi-polar, I understand the issue but you have to just shut the heck up when you’re in one of the extreme moods).

AI is now set to “Challenging”.

Challenging is a difficulty level that I’m trying to create where the AI doesn’t cheat but instead soaks up a bit more CPU.  The goal is to have something that in a 1 on 1 game will consistently provide a challenge.

In a game with several players set to challenging, winning becomes something that involves serious skill since the AI will make it a point to work together.

I’ve been working on a number of game mechanic tools that let players work together in more tangible ways.  When we worked on Sins of a Solar Empire: Diplomacy, we (Stardock and Ironclad) came up with the concept of treaties. I wish GalCiv had had this tool (hell, I wish WOM had it). FE has a number of treat options that will allow players to help friends and diss foes.

But for today’s tests, the goals are to see how well it plays ONE vs. ONE. 

These aren’t fair fights. I’ve gotten emails from people who are concerned that the game will be  “too hard”.  Challenging isn’t the default setting and there are lower levels of difficulty that provide new players a reasonable challenge without getting creamed (i.e. different algorithms get used).

In the XML, there’s a new AI feature called AIintelligence that at Normal is set to 1 and at easier levels is less than 1. I use this number to determine what functions get used. Up to 1.5, the AI plays “fair”.  After 1.5, I’ve written functions that let the AI do things that are less fair (like getting better start up conditions, better treasure, etc.).

For you old time 4X players, you can probably imagine all the kinds of things we can do with the AI here that aren’t normally doable in a typical strategy game.   Poor Jon Shafer, his Civ V AI stuff didn’t have magic available. We do.

Anyway, let’s start with game 47…

Game 47: Challenging AI

A few caveats:

(1) A lot of new art assets aren’t in my build. That’s why there’s so much brown.  So no “blasted lands”

(2) There’s only 1 tactical arena map. I see the same tactical battle over…and over…and over…please kill me.

(3) I don’t play fair. I am playing to kill the AI efficiently, not to “play” the game.

My starting position.

Exploring the wilds

AI stupidly sits by while I attack his city.

Victory in 60 turns.

Score: 21

Not a good showing.

Sovereign should have ducked into town. Why didn’t he?

Coding coding coding…

Game 48:

My start is pretty decent.

The AI is on a peninsula and has no problem sending forces home before me to reinforce. He’s bottled up.

..Fixed a bug with the way the AI determines what item it should equip…

This fix actually just made a huge HUGE difference to the way the AI champions and the AI sovereign function.

Relias proceeds to kick my ass.

Badly.

Now this is pretty ballsy, the AI has calculated that Relias can take us all out by himself. I don’t see how this is possible.

Relias tries to take us all on.

A bad move. I need to look at the calculations here as to why he’s running around on his own like that.

I try to attack him outside the city a bit later and the story is different.

Now, noticing the AI has a good lead here and should be sending more units after me.

Looking at a bug. 

It’s often difficult for me to tell whether a Unit is a Unit or whether a Unit is really an army. There are functions that tell this but because of all the performance optimization, there are some glitches in this area that I’m having to fix to make sure I have “live” data and not cached data.  You don’t want units to be recalculating their stats very often as it adds up very fast.

Ok fixed.

oh.

Hmm.

It worked.

Oh f..ck.

Defeat in 63 turns.

Ok. enough for tonight.