Grognards and Graphics:

Coming out of the 2D Closet

But then I started to buy computer games on my own, and more importantly I began recently to review them.  This means I have the fortunate opportunity to see a broad variety of games and judge not just the quality of a game on its own merits, but to look at it more closely against its peers.  My secret?   Come close, I'll whisper it to you:

I like games that look good too.

There.  I've said it. I'm outed -  I confess. I've always been an ultra-realist when it comes to games, someone who generally bashed any but the narrowest rules rationalization in favor of more numbers, more detail, more ACCURACY in simulation.  But I've discovered that I prefer games that look good, too.  In the war gaming community, it's tantamount to a 300-lb NFL lineman realizing that deep underneath his filthy, bloody uniform and sweat-soaked pads, he's really wearing a dainty pink camisole.

Why am I so embarrassed about my predilection for attractive games?  For some reason, computer wargamers seem to exhibit conservatism to an extraordinary degree as regards - of all things - graphics in their games.  Go to any wargames newsgroup on USENET, for example.  Among the posts detailing the orders of battle for the Latvian Army, you'll see discussions of various games.  One phrase or idea crops up over and over again "of course, I don't really care about the graphics, it's the gameplay that counts."  Like a mantra, this phrase is repeated over and over.  Does anyone actually mean it as much as they say it?  Or is it more a case of credibility gained by parroting the popular line?  (Or 'drinking the group kool-aide' as a particularly tactless friend likes to say.)

We're nearing the end of the millennium, technology and innovation surround us and somehow people persist in a certain special reluctance to embrace the new, the different even when it is something that is clearly better.   Sure, it's a trait particularly found in the older generations, but the occurrence is essentially universal - across ages, cultures, and genders.  It could be UNIX weenies in GUI denial.  It can be your parents not really wanting to try DVD because LP's sound "just fine for us, thank you."  It can even be the 14-year old who doesn't want to switch from Quake to Quake3 "because it just doesn't feel right!"

I really do find that ironic.  Irony gives way to amusement when (reading the same newsgroups) one comes away with a clear feeling of disdain by "wargamers" for the unwashed masses of other computer gamers - such as the vapid wargamer-wannabe RTS fans, or (worst of all) the mindless reflex junkies playing first-person shooters.  (Actually, I play both of these types of games and enjoy them quite a bit.  But I'm speaking here as a baseline wargamer.  Oops, my slip is showing.)

Perhaps it's nothing deeper than a faux elitism, crystallized around the most obvious difference between these games.  There's no question, there are some beautiful games out there and almost none of them are "wargames" in the generally-accepted definition.  But I come here not to praise prejudice but to kill it.  First let's dispense with the mantra.   Only the deeply esthetically confused or the deliberately aberrant would choose something uglier over something more attractive, all else being equal.  Taken to extremes, grognards (French for "grumblers"; formerly an appellation for Napoleon's Old Guard, but a title which -in a burst of appropriately circular logic- has been adopted by wargamers to mean the most serious members of their hobby) would suggest that, given the choice between a game where the combatants are ultra-accurately represented by featureless colored blocks or playing Steel Panthers (fun, if not the most accurate simulation), they'd seriously take the blocks?

Maybe it really isn't such a stretch for them.  After all, the older grognards came from the era of counter-based board wargames, as did I.   Sometimes we did use feature-sparse blocks to represent our troops (Gettysburg, anyone?).  But this is the COMPUTER era.  In the 80's computers weren't fast enough to BOTH crunch the necessary numbers for a good simulation, and push the polygons for a good-looking game.  But it's almost 2000, and computers whose power rivals minicomputers of only a few years ago sit on most desktops.  More and more games are coming out that I have the opportunity to review that are both attractive AND good simulations, if not simply darn fun games.  To put it bluntly, the future is coming, and it looks good.  There is absolutely no reason wargamers need to be satisfied with ugly games any longer.

In this vein, and recognizing that wargamers have generally been paying more attention to their Orders of Battle than developments in the graphics field, we at Strategy Gaming Online felt it might be a useful service to offer here a brief synopsis of the recent history of the graphics market.  We want to try to "catch you up" in a field that probably hasn't seemed relevant to your needs - until now.     

A brief history of graphics and computers, anecdotal, subjective, and with no particular scholarly rigor.

Of course the development of computer graphics paralleled the increase in computing power available in the CPU's to process and display such graphics.   At the start (just after the Cretaceous) computers were black and white (1 bit graphics).  Either a pixel was on, or it was off.  Later, right about when the Ice Age was ending, certain advanced systems began to display 4 colors, black and white plus possibly the two ugliest colors viewable by the human eye, cyan and magenta (2 bit color, also known as CGA).  This is known as the color DEPTH.  Simultaneously, computers were beginning to adhere to defined "resolutions", that is, the number of pixels shown on the screen.  This was normally stated as "pixels wide by pixels high" so the CGA standard was defined by the developer (IBM) in 1981 as 4-color,  320x200. 

Finally, just as man developed fire, 4-bit color came along and we started to have colors that resembled colors found in nature.  Likewise, resolutions were slowly improving, and the EGA standard was 16-color, 640x350.

In 1987, a major step forward in graphics standards took place with the promulgation of the VGA standard by IBM.  This became (and still is) the bottom-line compatibility standard for PC's today. After IBM lost its pole position in the computer market, the whole standards thing went out the window.   UVGA, SVGA, etc are a lot of terms you may see in the market, but they really don't mean much beyond what the manufacturer wants them to mean.

Since that time, in terms of the 2d performance of your computer (which is all we're talking about so far) the modern PC has a host of resolutions and color depths to choose from, with 800x600 16 bit color being a pretty common standard, the bleeding edge being 1600x1200, 64 bit color (and even higher).

To understand what this development meant for processor load caused by graphics, we can do a little math: for CGA, you have every pixel (320x200) defined by two bits.  So, 320 x 200 x 2 = 128,000 bits per screen redraw.  Of course, a BYTE is what is what we usually measure memory usage in, and that's 8 bits, meaning 128,000 bits = 16k.  The table below shows the incredible increase in processor load caused by the higher resolutions (note, this is the amount of data required to be crunched by your CPU for every single refresh of your screen):

Functionally, once you get about 1024x768, typical (<20") monitors can show the resolutions, but the displayed objects get pretty darn small.  Where to go from here?

The next frontier was hardware acceleration.  Really, it was the logical step after the development of GUI's in the late 80's/early 90's.  Since suddenly everything the user DID involved a graphics action, the CPU's burden increased tremendously.  So manufacturers began tinkering with ways to offload the typical windows graphics tasks from the CPU onto specialized graphics-card hardware (box drawing for example). 

Once this technique had been perfected, and games started becoming more and more demanding graphically, graphics card companies realised that this load-sharing would work for other graphics intensive processes as well.  Rather than asking a CPU (busy with everything else) to draw, shade, rotate, draw, and shade again a spinning cube (something that involves a great deal of intensive math operations to render onto a 2D computer screen realistically), they developed hardware that could take a simple command from the cpu - box with corners here, here and here, spin it at X speed - and do it (almost) by itself.  That, in essence, is 3d-acceleration.

Of course, this got the game developers drooling with possibilities and in certain segments of the entertainment industry (read: first person shooters) games are now written that REQUIRE 3d-acceleration.

The first steps in the 3d-acceleration field were small ones, and dominated by a single firm, 3dfx.  It's not a stretch to say that id Software's release of Quake, and 3dfx's similarly-timed release of the Voodoo graphics card started the entire 3d-accelerated market.  The Voodoo was what is called a 3d accelerator-only card.   It did not have 2d functions, and required installation adjacent to a normal 2d graphics card that everyone already had.  The Voodoo's descendant the Voodoo2 didn't have it quite so easy.  The Voodoo2 was a great card but other companies had realized that 3dfx was almost alone in the 3d-accel marketplace and decided that they too wanted a piece of the pie.  The NVidia TNT was probably the main competitor for 3dfx at that time, and they roughly split the market. 

(Note here we're talking about the 3d-accel chipsets.  These are usually not the same as the manufacturer of the graphics board, ergo the "Creative Labs" Voodoo3, or the "Creative Labs" TNT2".)

Now, we're in the 3rd generation of the technology, and there are a host of 3d-acceleration cards (all the competitive cards now are standalones that combine 2d and 3d operations) taking advantage of even faster pipelines from the CPU such as the AGP 2x and even 4x technology.  Currently the main contenders for top guns in this field are the 3dfx Voodoo3 series cards, the NVidia TNT2 products, the Matrox G400 series, and the NVidia Ge256.  3dfx, the former leader, is besieged by their design-level  acceptance of the adequacy of 16-bit color (supposedly in favor of raw speed, but this hasn't really been clearly demonstrated to me at least) in 3d-accelerated software, hobbling them just as all of their competition has moved to 32-bit color.   NVidia and Matrox specifically have released recent products that quite frankly blow away any other card in the market in terms of performance (at a reasonable price, as well).

Yeah, but I play WARGAMES.  How important is 3D to me?

It's a worthwhile question and really the reason for this whole article.  My goal when playing a wargame is to be presented as closely as possible with the decisions faced by a real(istic) participant of the battle I'm simulating.  This could be a panzer commander on the steppes and cornfields of Kharkov; or it could be Archduke Charles at Wagram, a Federation Command Cruiser captain in the Mutaran Nebula. 

Personally, I'm not just playing wargames to see if I can crunch numbers and estimate probabilities from 'to hit' tables faster than my opponent.   I want to be there - I want to feel the desperation when my attack to cutoff the salient is rebuffed with heavy losses.  I want to feel the cold sweat down my proverbial back when an opponent sneaks through my pickets to attack from a totally unexpected angle.  I'm an adrenaline junkie as much as any Xtreme athlete, just more cerebral.  I don't think I'm unique, either.  Perhaps I'm guilty of overgeneralization, but I think this is pretty common.

Now, computers are beginning to achieve levels that we only dreamt of 20 years ago.  The desktop machine in some l33t h4x0r's grungy room is several HUNDRED times more powerful than even the university-based minicomputers of 1980.   In fact, a quick search of tech-news will reveal that processor speeds are now widely felt to be accelerating faster than the software designers can write for (of course, blatant code bloat seems to also be doing ITS part to mitigate the speed increases as well...but that's a whole 'nother subject).  To suggest that wargamers must accept an either/or proposition based on the power of their machines is no longer credible.

There is no question that games are starting to come out with sophisticated engines AND believable interfaces.  And believability is what it's all about after all. 

Take Big Time/Battlefront's Combat Mission for example.  This is a perfect example of where wargaming is going.  Certainly, there will always be the hex-based, counter-mimicking wargames - there's an almost antique appeal to these that I find hard to shake.  But Combat Mission takes that paradigm (the proverbial "1000 foot general" floating above the battlefield and controlling every unit's minute by minute actions) and beats it severely with the reality stick.  Certainly, intense effort has gone into modeling the combat results - equaling any board wargame for example.  But more interestingly for this discussion is the 3d-modelled world.  If you are viewing from the point of view of an infantry squad, there's not a random factor somewhere determining if you can see the enemy tank through the trees.  If you can see it, you can see it.  Or, in the case of Combat Mission, you can hear it in true 3d.  If it's off to the left, you HEAR it off to the left.  (Has it really taken 20 years to get this critical facet of combat put into a game?)  Gone are abstractions - no longer is your squad "down to 3 firepower factors from 6".  Now you can look around and see that your LMG team is dead and a few of the guys are pretty badly wounded.  Combat Mission certainly has rules for visibility and command control, but by and large it lets REALITY impact you; rather than having a limited number of command points to issue (for example) you simply have to deal with the situation in real time.  Instead of a rule simulating limited control, you have (as in real life) information overload and task saturation.  Which of these alternatives is *more* realistic?  Which would you rather play?

Combat Mission is only one example; the recently reviewed Hidden and Dangerous, Homeworld, Panzer General 3D, Myth, and a host of others are accepting that we no longer have to simulate a 3d world in two dimensions.  I haven't even mentioned the simulator-descended products that are in the works from such companies as Jane's - their concept of a persistent online war, with tactical elements represented by worldwide players and those players' strategic goals set by OTHER worldwide strategic players - which could prove to be the most significant advance in the history of simulation and wargaming since dice.

It's almost a form of Pascal's wager (or then again I may be staying up too late writing again when I start using Dennis Miller-isms).  You can either drop the couple hundred bucks for a kick butt 3d-accel card (such as the card in our feature review, the Matrox G400) and play any game that suits your fancy; or you can decline to participate.  You certainly save money (but then who ever got involved in computers to SAVE money? Ha ha ha.) but your options are limited to what companies feel charitable enough to build down to your machine specs.

Suffice to say that in the very near future, there are going to be two kinds of wargamers out there.  The first, sticking to conventional wisdom and happy with their narrow realm, will remain oblivious and unconcerned with anything remotely having to do with graphics.  And they will probably be happy.  The second will be the host of gamers that see that 3d-acceleration is no longer the twitch-gamer luxury it was, and recognize more wargames will require 3d-accel as the minimum hardware standard to participate in the most intense, realistic wargames the technology can put out.  3D Acceleration is no longer limited to reflex games - it has become an integral component of games that seek to simulate reality. The future of wargaming is coming, and that future looks shaded, bump-mapped, multi-textured, and real.  Will you be able to see it?

If you like to comment on this feature, please post a message at the forum.
Authored by Steve Lieb