Doomed: How id Lost Its Crown 491
bonch writes "Steve Bowler, lead animator for Midway Games, has written an article for Next Generation called Doomed: How id Lost Its Crown. He talks about id no longer being the king of the hill in the FPS genre, losing the multiplayer gaming wars to Counter-strike and the engine licensing wars to competitors like Unreal 3.0, and focusing too much on rendering realistic environments at the expense of modern gameplay features. From the article: 'It's hard to stomach having to shoot a zombie in the head the same number of times as in the body (six rounds from a pistol, thanks for asking) to dispatch it, when you can shoot a light fixture and watch how realistically light dances around the room.'"
Seriously- (Score:5, Insightful)
Counterstrike runs on crap hardware, and basically, a crap internet connection. You'll get called a lagger, a newbie, and a lamer, but it will work, and you can play, and have fun.
Gameplay is extremely important, but so too is availability.
doop (Score:3, Insightful)
Doom 3 was a great game, imo, however people's complaints about the whole flashlight mechanism were justified, and I can see how it would detract from the entertainment value. Id's goal was to make a scary game, and if you played the game with the swapped-in flashlight as they intended, it was indeed scary. The lighting was better than in any game I'd played at that point and created an unparalleled atmosphere of creepiness.
That being said, the idea that in "the mysterious future" you wouldn't be able to hold both a flashlight and a gun hurt the game's credibility. And going for the cheap scare so many times did tend to get old.
They were also determined to make D3 a single-player game in a field now dominated by multiplayer and massively-multiplayer games. I would have thought that they'd have realized this better than anyone, given that they practically created the market for multiplayer FPS gaming, but they chose to make Doom 3 a single player game, and between that and the whole flashlight deal, many people decided the game was a dud, and thus its fate was sealed.
I still thought it was a great game though!
It's the multiplayer, stupid. (Score:2, Insightful)
Everyone plays ut2k4, hl2, CS, whatever because it's fun either sneaking around and sniping people, or jumping around flinging rockets. Doom III kinda mixed them, and failed to create a fun multiplayer experience.
I'm still looking forward to Quake 4, however.
I don't recall... (Score:3, Insightful)
Id hasn't really been a player on the FPS game market in a while. Their recent games (Quake 3, Doom 3) have basically been technology demos. They sell well because we nerds think it is cool, but the actual games leave much to be desired.
We know that Id makes its money from licensing its engines to people. Half-life made Id some money. Keep that in mind. I'm not sure if the Source engine takes anything from one of Id's engines.
Re:Seriously- (Score:5, Insightful)
The sequence to Doom (someone else said this once).
1. Move around corner
2. Light turns off
3. Loud noise
4. Option - Load Saved Game
5. Wash - Rinse - Repeat.
The problem I have with Doom 3 (Score:3, Insightful)
Radiosity is the property of multiple light reflections. When a light shines on a surface it reflects, of course. However that light can then further reflect off another surface and so on. That's what leads to soft shadows, and is the reason why when you turn on a flashlight, the whole room is slightly illuminated, not just what oyu are pointing at.
Doom 3 doesn't do this, a light hits a surface and will reflect to the screen, but there's no multple levels of reflections. The net effect is hard shadows, corners that are always dark. You can't get a good brightly lit scene.
Now I don't fault them on this, doing radiosity in realtime isn't feasable at this point on most cards. However other games can deal with this, the don't do all their lighting in realtime. Some is done in realtime, some is a precomputed light map. That allows for a global illumination, but one that doesn't have to happen in realtime.
That is my big problem with the engine. Sure it's more accurate than the UT2004 engine, technicly speaking, but it doesn't look as good. UT is "faking" the lighting and shadows, but they look good, and you can have a nice brightly lit outdoor map, or a dark indoor map, and they both work. You can have a light source that casts light on to all surfaces, even those it doesn't directly hit, since it's calculated before hand.
Personally, I'd rather have a game engine that looks good rather than one that is more accurate.
The next gen of good FPS's will be like Morrowind (Score:5, Insightful)
The thing is, non linear games where your actions determine your standing in the game, as well as its path and outcome, are the wave of the future. Especially games with thousands of mini adventures on the side. Also, in Morrowind you interacted with practically *everything*.
If Morrowind were not done years ago and were done today through the Doom 3 *or* Unreal 2 Engine (either of which would imply far fewer bugs than Bethesda's own "engine"), it would eclipse all other games in popularity for 2 years. I say that because Morrowind appears to be almost the single player's equivalent of Starcraft in popularity and longevity.
The lesson: forget the graphics arms race, achieve Doom 3 or UT2004 level graphics and leave it at that, and concentrate on a deep, complex, non linear, "easy to get into it quick" story lines, and endless paths of quest resolution. Give FPS players a world to explore, tweak the outcomes, and generally have fun in.
ID somehow appears to be furthest behind in pursuing this goal, even though Doom 3 is no more linear than HL2 or Unreal 2.
Re:Light? (Score:4, Insightful)
This [glenmurphy.com] is an ingenious mod. Seriously - in my view, the main reason Doom 3 was such a poor game was the fact that you could see *nothing* without the flashlight. Who cares about the super new graphics engine if you're barely given an opportunity to take a look at the environment?
Re:Of course you have to keep shooting (Score:4, Insightful)
Re:Seriously- (Score:4, Insightful)
Other than new graphics, this game had nothing going for it. A total dud and that's that. Only a total fanboy could like it.
Lets face it, Unreal and Half Life kick Doom's ass all over the place at this point. Although I admit that HL2 was disappointing because of Steam (I love waiting 5 minutes to play my game because of that fucking piece of shit). I will never by another half life that has Steam in it. The only online FPS I really like Is UT2k4. Everything else is pretty much ass these days.
Finally, are there still people playing Counterstrike? rofl
Re:Of course you have to keep shooting (Score:1, Insightful)
Re:The next gen of good FPS's will be like Morrowi (Score:2, Insightful)
Re:The next gen of good FPS's will be like Morrowi (Score:4, Insightful)
Far more people are used to "On-The-Rails" RPGs, where you must do this, then this then this, etc. Playing a game like Morrowind requires a mentality shift. The game is not in control of the experience, you are. I firmly believe that is the better type of game, but opinions differ.
The problem I see is that for too many gamers, the goals must be rammed down their throats before they can do anything. To suddenly say "Hey, we've created this entire world, have fun!" is too much for them to deal with. Not because they are stupid, but because it is so vastly alien to them.
Personally, I will be far less likely to buy an "On-The Rails" game because it is too frustrating. I have this great world around me, but it is very much like being on a train. I can see all of the potential, but everything is predestined. No getting off to enjoy the scenery and explore the world rushing past my window. I must fight this guy, I must go here...I get pissed. If I wanted to be led around, I would have put in a DVD movie instead. I want to go off on my own and do my own thing.
With Morrowind, I can slip in and out of the main plot at will, or ignore it completely. There is no forced-anything. From the moment I walk out into Seyda Neen, the options are virtually limitless. I can go anywhere, do anything and be any kind of character that I want to.
Choice is good, replayability is fantastic and having the opportunity to simply walk away for a while and pick up right back where I was in my "other life" is priceless. I've done marathon sessions before, but only because I've had a night where I had nothing better to do. You get sucked so far into the game and your character, but unlike Everquest and World Of Warcraft, myself (and other Morrowind fans that I know) find it very easy to put the controller down and not let it consume our real lives.
Being able to not just live one adventure, but continue on a lifetime of them, without it getting in the way of everything else (or costing money every month) just can't be beat.
Don't care anymore (Score:4, Insightful)
I played D3 at a friends house for about an hour. Big screen, dark room, was fun. But the fact is, I'm not the gamer I was back in the early nineties; zombies just don't do it for me anymore. Doom was amazing because it was a technical tour de force; I still remember walking into some of those rooms and going 'that is so *cool*!' Frankly, the whole zombie/monster story was pretty old even then, but Doom was such a good game, I was happy to play the entire thing. Hell, I did the same for D2.
Quake was pretty good, but seemed like pretty much the same thing with a slightly mideval twist to it. By the time Q2 came around, it seemed like I was playing the "same-old-thing", even though, id never disappointed in the graphics level.
But in the intervening years I'd gotten married, had kids, played a lot of other games, and given the time I now have to play, I'm looking for something different and original. Id seems to think that they can coast on demonic bitmaps and licensing forever.
Re:The problem I have with Doom 3 (Score:1, Insightful)
The reason Doom 3 didn't use it was because of 2 big reasons:
a) realtime radiosity is very hard to render (in realtime) and Carmac insisted that every light should be fully dynamic and cast fully dynamic (realtime) stencil shadows. That was (as we know now) a mistake. In fact even Carmack _almost_ amits it in his QuakeCon 2004 keynote. The reason for this is, of course: Why would you want all lighting to be done in realtime? 90% of lights in Doom 3 couldn't be destroyed - meaning they casted (the same) shadows every frame. So every frame rendered the engine had to re-render every shadow. Caching helped some things move a little faster here, but for the most part it was unneccessary and it costed a lot of CPU (read point b).
b) Carmack made, what I believe to be, the fatal mistake of trying to support older cards by transfering stencil shadows to the CPU. For a game of this technology, I think this was a very bad move, because you couldn't even play the game on a GF3 because of the stuttering, so why even bother? Carmack often said he would rather see people lowering the res than lowering detail - he absolutely wanted everyone to display those damn shadows (and TBH the game would have sucked wihtout them). Now, realtime radiosity is one of the things I really want to see used in games, but I imagine you'd need at least 4-8 light bounces to even start speaking about radiosity and probably 16 bounces (I speak this from the top of my head, maybe you'd need even more) to start seeing good results. Now if you put 16 extra light passes on the CPU, then you're really pushing it. Of course it would have been much, much faster if he would have put the shadows on the GPU and used a lot of static lights with precompiled radiosity, then IMHO the game would really show in its full glory.
To sum it up; Carmack was ahead of his time, but was too worried about old hardware. For example, even Unreal Engine 3 doesn't use all dynamic lights and it still looks great. I really believe Carmack is the guy to pull of some amazing graphics but he should start using some advanced shaders and forget about legacy hardware. But then again, id believed they can ship D3 in 2002/03(?) or something, when those restrictions would have mattered.
If anything id can take a note or two from the Epic guys and the work they're doing on Gears of War - old skool FPSs just don't matter much today. No one wants to shoot, reload, shoot,... for the whole game. People expect more than this from a game with such technology. Last year, Carmack didn't want to show off with his new technology at QCon, but I think he promised to give a peak this year. This should be interesting, but AFAIK he's still sticking with the all-dynamic-shadows thing, but he's adding soft shadows on top of that. IMHO yet another mistake.
btw, Carmack if you're reading this please stop caring about 3 year old hardware and just give us some eye candy. If I buy a GF7/8 I at least want to see some cool shaders in action. Not the same thing my neighboor with a R9200 is seeing, just faster.
Wrong (Score:2, Insightful)
Nope. CS has been by far the biggest multiplayer FPS. Unreal Tournament comes in a distant second.
See for yourself. [wikipedia.org]
And from talking to other gamers, it seems that most of the Quake 3 players have moved on to either Painkiller, RTCW. And don't forget Call of Duty, Battlefield (1942,vietnam,2), and Day of Defeat...I think some one is a little behind on their FPS's
Doom 3 is more a demo then a standalone game (Score:3, Insightful)
What was good about Quake 2 was its multiplayer mode. While Quake 1 allowed multiplayer as well, the initial DOS version required either external tools for IP networking, or a nullmodem cable or modem connection for multiplayer modes.
Then came Quake 3, which never got a playable single player mode, rather, it concentrated on multiplayer mode almost exclusively. The engine however was capable of single player mode quite well as shown by for example Return to Castle Wolfenstein.
By the time of Quake 3, it was clear that ID could create a good engine, but needed third parties for creating good content, and one can argue that Quake 3 served more as a demonstration of the graphics capabilities of the Engine then anything else.
In the years to follow quite a few good games were build on this engine, including what I still consider one of the best multiplayer games so far, Enemy Territory. This resulted from finding a very good balance between complexity of gameplay (relatively simple) and realism (amazingly good for its time). You can get inmersed in the games without having to learn too much, and can quickly learn enough to have an enjoyable gaming experience.
It seems to me the mistakes with Doom 3 are in 2 distinct areas. First of all, the balance between gameplay and realism is not right (as the article suggests also). Second, and imho even more important, ID can't create proper content, and rehashing the same old content in a new engine is just boring. They saw this when making Quake 3, and didn't even try, but failed to remember this for Doom 3. They were making a demo for the engine and confused it with making a complete game with entertaining content.
To me this is quite evident from the fact that old (Doom 2) based games like terminal velocity and a game like Duke Nukem 3d are a lot more fun to play then anything ID ever made except maybe for the original Doom, Quake and Wolfenstein.
Re:Seriously- (Score:3, Insightful)
Re:You know.... (Score:4, Insightful)
The cheap scare (Score:3, Insightful)
I think the big problem with most FPS games these days is that the story that goes into them always feels like a lame excuse to kill crap. You run into each bigger and badder boss of some level and you get some new uber weapon to kill him with. There's a story, but it's only to give some sense of logic to all these things you have to shoot.
Frankly, until the technology's evolution rate slows down, we'll have to deal with this crap. I mean, can somebody tell me what the plot is of far cry? I have no idea, I just know it has really realistic looking water. When the technology evolves to become more of a story telling medium than an R&D lab for rendering techniques, we may have something. Doom seems to forshadow that a bit, having effectivley ignored the multiplayer element in favor of atmospherics, etc, but it still seems too interested in graphics rendering navel gazing. In the meantime, the FPS genre will make up for it's utter lack of creativity by networking us so we can kill eachother and drool over the special effects wizardry.
Frankly the only game that I've seen recently that I thought was genuinely innovative was PlanetSide. I've been playing it for two years now and it's still way better than anything else out there. It's not that pretty and it's a massive resource hog, but it really is a good demonstration of where this goes. It falls a bit short, but it at least gives you a grander sense of some point to the fighting.
On most games, you fight a round and you kill, capture flags, etc, then the round ends and you start over. PlanetSide does get to feeling like a hampster wheel after a while because there is no win condition, but there's at least a larger sense of the battle always going on and that your contribution to it does have an influence. I think what comes after it should be really interesting, but we'll have to wait and see.
Not accurate to how people use game engines. (Score:5, Insightful)
Precomputed light maps do indeed have to do with the choice of engine, because the engine takes care of computing the lightmaps for you. Halflife 2 for instance, supports normal mapped radiosity calculations, in which the diffuse lighting components are added along different vectors during compilation, and then dotted with the normal map during the rendering. "Level designers" don't store them in textures, the compile tools that are associated with the engine do, and the engine takes care of displaying them appropriately.
Having precomputed lightmaps in the doom 3 engine would break all the internal consistency of the lighting. Mobile lights in engines based on precomputed lighting are treated differently from static lights. Doom 3 doesn't have this distinction.
The doom 3 approach allows lights to be much more dynamic, but when a light is that dynamic, you can't have precomputed light maps. You wouldn't have any way of updating them to reflect changing light conditions. Every time an imp warps in, all the lights dim. This couldn't be done realistically with precomputed light maps.
Adding precomputed light maps would require redoing all the internal assumptions about lights in the engine, and you would be basically writing your own.
Re:Seriously- (Score:2, Insightful)
Re:Dupes: How Slashdot Lost Its Crown (Score:4, Insightful)
Re:Wrong (Score:3, Insightful)
Re:It's the multiplayer, stupid. (Score:3, Insightful)