John Smedley On The New Galaxies

Gamespot has part one of an extensive interview with SOE CEO John Smedley about the recent and controversial changes made to Star Wars Galaxies. From the article: "Star Wars never hit that excitement level around here. It never got--there never was a critical mass of people here that wanted to play it. So we knew we could do way better. And I guess as much out of a love for making these kinds of games, even though that sounds corny, though it's true, we wanted to make this game better." We had our own talk with Mr. Smedley not too long ago.

Comments

Changes...

[warGod3]

As a long time player of SWG, I have seen the game go from something was that fun to play to something that is being dumbed down for mass appeal. I understand that they are trying to appeal to a greater market to attract more people. They (SOE and LucasArts) have decided that the opinion of people that have played a while is irrelevant. Although numbers have not been published by SOE, it is estimated that their losses are far greater than they anticipated with the NGE.
I think part of what I read scared me. But it didn't when I read it at first, it was after a conversation I was having with someone regarding the another industry. It seems that other industries have relied heavily on their name brand to carry them through as well, and things happened and they had to tighten their belts. What will happen at SOE? I think that they are trying to target SWG to the "younger" generation.
Now we are hearing rhetoric and rumors from all sides to all extents about what is going to happen. Yes, I am still in the game, for now. Are more changes to come? Probably, but what will they bring?

How about content?

[Somatic]

Smedley: We spent quite a bit of time and money doing product research. We did a number of focus groups and talked specifically about changes that we could make to the interface, to the combat, and to the overall gameplay experience--to make it a lot more fun. We also did surveys asking the current user base what was missing, what were things we could do to make the game better.

And in all of these user surveys, apparently no one ever said they wanted content, and that's why there was none, right?

Let me spell this out for you future game developers. Randomly generated content is not content, it's crap. The brain of even the slowest human can smell the difference between hand crafted and computer generated content. It's why the Turing test hasn't been passed, it's why automated customer service menus piss people off, and it's one of the reasons SWG failed.

Creating a massive world that was 99% empty might have seemed like a good idea on the surface, I know. You'd save all that time on programming, writing, implementing... you'd create beautiful cities (and you did), players would go to them and be merry... but all the rest of the world would be random. It didn't seem like a bad idea, I know. I can follow the thought process that led to SWG's design, and on paper, I can see how it might have sounded good.

But what you've got to understand, devs, is that there is no substitute for the human hand. Technology is great when used right, but it is not a good babysitter. Random levels worked for Nethack because it was a single player game, an ASCII game, and the design was genius for its time. But random will not work in a modern MMOG.

People need to fall in love with the world they're playing in, and a computer-generated design just can't inspire that love. Only the human hand can do that. Maybe in the next 20 years a genius programmer will come along who will write the algorithm that will be able to trick the human brain into loving it that way they love something painstakingly crafted by a human... but for now, you have to do it by hand.

I'm serious, game devs. I'm trying to save you millions of dollars. Don't do it. Hire a bunch of high school aged D&D DMs if your budget is that tight. Just hire a human, k?

(oh yeah, and crippling bugs wrapped in unstable code that never get fixed are bad too)

I'll take the opposing point

[MilenCent]

I'll put aside the question of whether Star Wars Galaxies' random content was good or not. I'll even grant you that it might not have been possible for any Star Wars game to be randomly generated -- it may have been possible to get that workable engaging, but for the sake of my point I'll concede it.

I don't think that random content is inherently bad. I don't think it'll always look like a computer made it and not a human. I don't even think that always matters.

You invoke the sacred and holy name of Nethack as a talisman against the gameplay it stands for, but the plain fact is, no one's ever tried to make Nethack-style random gameplay work in a commercial product. (Diablo does not count for reasons to be revealed.) And Nethack, despite how it looks, is absolutely not outdated -- indeed, its open source nature has spawned dozens of interesting and creative patches for the game, ranging from special levels (Lethe, Heck2) to new monsters (Biodiversity) to entire new play mechanics (Color Alchemy, described below).

But it is not controversial to say something like "Nethack rulez" on Slashdot, in which the radio of Nethackers as opposed to the general population is, shall we say, higher than normal. So to avoid mere karma whoring, I'll attempt to explain how to make random content work.

You do it by randomizing more than just maps. (Re Diablo: There.) Having an infinite amount of terrain to explore is not enough to make a game interesting. Roguelikes do it by also randomizing the item definitions, restricting player knowledge of them, and making their discovery a major part of each game. Some games randomize still more, or provide mechanisms by which the basic item randomization has profound effects on the game. Examples: The presence and alignments of altars in Nethack has a profound effect upon that game, even though technically they're just part of its map generation. ADOM generates different alchemy recipes each game, which can potentially give players a potent source of resources. There is a user-created patch, Color Alchemy, that does something similar in Nethack: instead of having that game's potion mixing system be based upon type (Healing Potion + Gain Energy Potion = Extra Healing Potion), it's based on the color in the potion descriptions (Whatever Red Potions are + Whatever Yellow Potions are = Whatever Orange Potions are)!

Also, randomly placed monsters are not interesting in a game in which they...
A. ...have no "hard" way of harming the player. When overcoming player death is as simple as clicking a respawn button, no monster is really that dangerous. Rogue had *common* monsters that could do permanent strength damage, could quickly drain levels, could confuse with a glance, permanently degrade armor, etc. Nethack, of course, has the infamous cockatrice, Medusa, monsters that can curse items, monsters that can burn with a glance, thieves, and many others.

And all true Roguelikes have permanent character death. When that foe around the corner could suddenly destroy your entire character, let alone his stats and equipment in ways that are not trivial to overcome, then that random generation begins to really mean something. Show me a MMORPG like *that* and I'll be there like a shot.

B. ...are all essentially aliases for each other. This is related to point A, I think, in that many games tend to support umpteen different types of damage, but the main way in which they are differentiated from each other is in the kinds of resistances a player may have. (This is a reason I don't play Angband, despite its being a major Roguelike.) Often, many of them tend to behave in the same way as the others, have the same general types of attacks, and ultimately, due to the reluctance of the designer to have them do really bad things to characters, don't have anywhere near the personality that Nethack monsters have.

So, I think you can indeed make randomly generated content, but it can't be half-assed. And ultima