quote: Original post by Erebus in order to better clarify i will give an example. there is a current post in this forum discussing skill systems in RPG games. the argument about it not being the programmers fault was used in this discussion, specifically by InnocuousFox and agreed with by dwarfsoft. InnocuousFox makes reference to a "nebulous concept known as ''the Real World''", and dwarfsoft mentions technological limitations. in the case of a skill system, i don''t really see what kinds of obstacles could stop a game programmer from making their ideal system. it all seems to have to do with the programmers ability to develop a system and their ability to design the algorithms to implement it.

My original statement was:

quote: Perhaps there are forces beyond your awareness (posing as a nebulous concept known as "the Real World") that make it more difficult than you believe.

This was in reference to the comment made by DrMol to the effect that ''all games suck'' and he could do so much better just because he is armed with the knowledge (opinion) that ''all games suck.''

Your commment saying "i don''t really see what kinds of obstacles could stop a game programmer from making their ideal system." is similar but without the attitude. (Thank you, btw.) All I was saying with my above comment is that there ARE many things that stand in the way of getting exactly what you want. Many of them are political and make little sense. Others are technical in that there are only so many clock cycles and CD bytes to go around. Others are financial... e.g. "we COULD do that really cool thing, but the publisher only gave us so much money to spend on labor."

Now, on most of these, management can be blamed to some extent if there is a choice between a cosmetic feature and a game play feature. However, management also cannot be held responsible for the presures of economy. There is no way that anyone can GUARANTEE that a title will sell the extra X number of copies needed to cover the cost of development of an extra feature. Therefore, it gets to a point where adding a feature or refining it to death (diminishing returns) gets to be a profit risk. Tkae a few too many of those and the company ceases to exist.



Dave Mark
Intrinsic Algorithm Development
"Reducing the world to mathematical equations!"

If games could be made with the expression of some of the better movies, they would evolve to a higher level of quality.

Take The Matrix for example. When I looked at the cover on the DVD, I thought, well, ok a bunch of leather-clad people with guns running around.

But when I saw the movie, it had so much more character than that. The people were authentic, the villians complex (for a machine, anyway). The whole movie just brimmed with style, attitude and mood. And that mood was consistent.

When devlopment teams can get together and marry all the attributes into a well tuned whole, I will have a new found interest in games. Until then, my interest is waning because they are all so much of that same thing anymore.

It''s like seeing "Aliens" (a movie I really liked, and wanted to watch again and again), and then seeing "Creature", an incredibly cheap knockoff of "Alien".

No One Lives Forever and Half-Life were, for me, two fine examples of FPS games that combined the typical challenges of a FPS with a storyline and mood that stayed consistent and compelling throughout the game, without insulting my intelligence or interest in being entertained.

Return to Castle Wolfenstein, however, didn''t hold my interest for long. It had nice graphics, but for me was too much of what I''ve done already. I don''t want to play a game just to say to myself, "Ok, let''s see they did with the NEXT level in this game...". I''m at the point where I want to be entertained by where I am at the current point and not finish the game just to find out what they put in it, or what the climactic encounter is.

Thank you InnocuousFox. I appreciate your responce and the lack of hostility. That''s what I wanted to know. Most of the answers you gave me involved economics, etc. which were things I already knew. However, you did mention clock cycles. This isn''t something I know a lot about. The whole point of this post was to understand what you were talking about, not to attack your opinion. I hope you realize that and thank you for clearing it up.

Someplace on Gamasutra (I think the article is linked from here as well) are a couple of articles regarding the state of the AI industry in games. There were a couple of surveys as to what percentage of processor time is devoted to AI and things like that rather than display, etc. The same can be said for things like complex skill systems et al. The more complex you get with something, the more it will tie up the system. At that point, you are trying to sell the concept of something that is invisible (skill systems leading to immersion) against something that is blantantly out front (graphics and sound leading to immersion). Unless the game designer has a good focus on game play rather than eye candy, you will tend to lose this battle.

Dave Mark
Intrinsic Algorithm Development
"Reducing the world to mathematical equations!"

After reading this, I'm even more convinced that your better games are going to be first implemented in MMOGs. Why you ask?

quote: Time and Budget.

How much time do you have to get it to a state of 'working' with an MMOG? About 18-24 months for a beta state, maybe a little less. Budget? After the game goes live you start having income, which means that you can devote resources that won't actually cost you anything other than profits, though many will argue that the investment in making the game better will net you more profits. 18-24 months isn't much time, but any MMOG team would hopefully continue developing the game after release, making it better, thus increasing the actual development time out indefinitely.

quote: processor time is devoted to AI

With an MMOG, how much CPU time do you have total in any minute? A whole lot more than a regular computer game does. You can build entire servers that have one task: AI. With this, you gain much more freedom to develop the AI to a more reasonable state. You don't have to make NPCs that will either fight to the death or do the fight/run/recover/return/fight... ad nauseum, you can make NPCs that know when they are outclassed and react accordingly. If you've taken 50% of your life in damage and done barely 10% to your opponent are you going to stick around? I wouldn't be and I wouldn't expect an NPC to either.

So let's see, more time, more resources, time to do things right after the game has gone gold, time to add those features that couldn't be implemented under the original release restraints, more CPU time to handle complex AI, larger databases to handle NPC AI rules, etc...

There isn't a single detriment to making an MMOG as opposed to a normal computer game, except for the initial investment, but if you look at the post-mortem for DAoC on gamasutra, you'll see that you don't need to spend $8Million to make an MMOG, you can do it with $2-$3 Million... if they were honest about their costs. I'm guessing that the cost of the servers was not included in that though, I could be wrong though. You could probably run a couple of zones on one dual processor blade server with a Gig or two of RAM, so you're talking about a server that will probably cost around $3-5k. Figure with today's servers being much more powerful than the ones that EQ started with and you can make some pretty darn big zones and still not see a serious performance hit in any zone until a lot of people get in it. So figure around 80 zones, 2 serious servers for handling the character database (mirrored dbs for better uptime) for each shard, a couple of serious servers for the more complex AI which would only be needed for certain situations and you're probably looking at around $300,000 in server costs (40x$5000, plus another $100,000 for 4 quad servers), if you use a Linux distro and MySQL as your server. If you change over to Oracle and any other OS, you've probably just tripled your costs.

So figure you spend $3.5Million on game development and the first couple of shards that will support around 4xserver load (hopefully around 3000 or so online, supporting 12000 users total) users, if you want to support 100,000 users you'll have to add another 6 shards, or $1.8Million. It's probably possible to do this with fewer shards and supporting more users per shard or with fewer initial zones. If you support 100,000 accts, or 25,000 online at any time, you can figure that your bandwidth costs will be 2kbps(25000 online users), or around a single T3(OC3 IIRC, or 54Mbps) per month. Toss in everything else (employees, location costs)and with 100,000 users you should be able to bring back $500,000 per month in return, or recoup your initial investment of around $5.5Million, including servers in one year and that's not even including the fact that these games are sold for $40 each (not that the publisher or the development company gets that per copy).

Sorry I got a bit heavily into the economics of MMOGs a bit in-depth there, just providing supporting evidence. Obviously though, as you can see, with a userbase of 100,000 any MMOG becomes successful very quickly. Considering the fact that you're probably putting around $200,000 into your programmers per month (maybe a dozen programmers), you should have little problem further developing the game and getting more people into the game regularly, which means that you will have more resources coming in every month and you can invest more into fixing all the stuff that was 'half-assed' in the initial release of the game and implementing those features that you wanted but couldn't put the time into before release.

[edited by - solinear on March 14, 2002 12:42:01 PM]