Gettysburg Unplugged: NorbSoftDev on How Scourge of War Gettysburg Thinks
By Matt Peckham
PCWorldNov 30, 2010 9:27 am PST
Gettysburg: Scourge of War isn’t a real-time strategy game for the masses. It has a perfectly functional interface, but you’d never call it “slick.” It’s more study-intensive than the comparably simplistic Total War real-time strategy games but yields commensurately higher returns. It tackles the mother of all Civil War battles with aplomb and occasionally startling historical verisimilitude, offering control of blues and grays from army scale down to regimental level. It’s wargaming without apologies, designed by hardcore history buffs for hardcore history gamers.
This spring I spoke with the Scourge of War’s creators, Jim Weaver and Norb Timpko. In part one, we talked about Gettysburg maps, canonical books, and how they got from Waterloo to Little Round Top. In part two, we discussed the design process.
GO: They ran a demonstration of artillery canon fire at Second Manassas, over by Brawner’s Farm, and one of the things they talked about was that Jackson had a sort of philosophy where if he ever caught any of the guys riding on the limbers, he’d have them shot.
So I was playing around with the artillery in Scourge of War, and I see three guys get up and sit on the limber as I’m moving it around, and was wondering if the Jackson story was apocryphal, or is that a level of detail you’re not as concerned with modeling here?
JW: Well, it would depend…on the battlefield, riding on the limbers, they could get where they were going in a hurry. But on the march, they would walk, because it reduced the load on the horses. And the Confederates, being short of horses, had to take care of them to the greatest extent that they could. The Confederate artillery teams almost exclusively had four horse hitches, whereas the Union artillery had six, which gave them a little better mobility and speed over the ground. But they had the horses to spare where the Confederates didn’t. So I can understand where Jackson would be saying, on the march, you know, get down and walk. Although he was probably less polite than that.
GO: When you’re modeling the different regiments and their how they interact, how do you model and calculate the formation physics? Does the computer see big blocky rows of guys, or discrete bodies? Are the gun ballistics modeled individually, or is it more of an abstract firepower aggregate?
JW: The formations are actually set by sprite. We have a whole file that’s nothing but laying out all the different formations and subsets that we’ve included in the game. There’s a lot we left out, of course, because we don’t want everyone to have to learn a 300 page drill manual just to play the game.
GO: I picked one up at First Manassas and thumbed through it. It’s kind of flooring. I mean, you’ve got three hundred pages devoted exclusively to complex formation drilling. Did these guys back in the day really have all that stuff down pat?
JW: The average infantry unit would drill, weather permitting, four hours a day, every day, because the drill was so complex. It had to be learned to the point of being spinal reflex, because you had to do it on the battlefield when all hell’s breaking loose around you and artillery’s going off and people are getting shot and there’s noise and smoke and chaos. When the colonel says wheel left, the feet need to do it without thinking.
That’s how you win battles, being drilled to that level of being able to do it without thinking about it, because you’ve got enough to do with loading and firing and filling in the gaps as casualties are taken. You had to know how to go from one formation to any other formation, and how to get there without ending up in a big rugby scrum in the middle of the field, which certainly happened a lot with green units.
GO: How cohesive were these formations under fire?
JW: It depended on…there’s a famous quote from I forget which southern general who basically said Confederate units were never in straight lines, that every soldier aligned on himself and marched at his own speed. So the formations were general approximations, but they were never as pretty as the drill manuals made them out to be.
The Union troops tended to be a little better, but it wasn’t like on the parade ground, because you were marching across some farmer’s field with rocks and woodchuck holes and places where an artillery shell had just landed, so the ability to keep even reasonably close formations was a trick. Keeping your formation organized and lined up allowed you to concentrate firepower down range, and that’s what…being able to put rounds on the target won the battle.
Next: Artificial intelligence and historical stupidity
GO: You have scripting dictates in some of the scenarios to establish a certain amount of historical authenticity. What kinds of choices is the A.I. considering when it’s not on script?
JW: We tell it what it ought to do, and then Norb figures out how to write the code to make it happen, and that’s part of his genius as a coder. We can tell him the troops need to do this, and he’ll translate that into lines of code.
That said, the A.I. was really designed to react to some extent unpredictably, and to react dynamically to what’s happening. That’s part of why it’s such a resource hog, because it’s doing a lot of thinking about things and checking things and looking around. As we’ve worked on it, we’ve made it smarter and smarter, although as somebody pointed out, even when it’s acting stupidly, that’s historically accurate. You can find historical examples of real generals doing some remarkably boneheaded things.
JW: Right. We designed it to read in the personality characteristics of the various generals, from the OOB [order of battle] files, so that this will tend to…if a given general was historically not very aggressive, that’s the way we want him to act as sort of his base, around which behavior will oscillate with a degree of randomness. And if somebody else was really aggressive and you put him on defend, chances are he’s still not going to stay there and he’ll go out and attack anyway.
We’re thinking to be really good at this game, you want to look at your generals and know who would be a better defender or attacker. I’m not sure how many people get into it at that level of detail, but the capability and functionality are in there.
GO: You’d think that would be the primary draw for Civil War buffs, to be able to go in and almost role-play the battle. You know, stick the game in grognard mode [which locks your visual and command ability to one person, including having to issue orders by courier dispatches].
JW: Yeah, from the player’s perspective, that’s the hardest mode to play in, that pure historical mode where everything goes out by courier. We’re still pinging back and forth about that mode, because the interface wasn’t originally designed with it in mind. It was designed for the general 200 feet up, who can see the ground from that privileged decidedly unhistorical aerial point of view. We’re still feeling our way around what else we could do to tweak the interface to make it more practical to play when you’re tethered to the level of the general’s head.
GO: It actually threw me at first. I installed the game, loaded it up, saw the difficulty settings, you know, easy, normal, veteran, grognard, and thought “well grog, of course.” So I click grog and launch a battle…bear in mind I haven’t read the manual yet…and then I’m sitting there on the back of the horse, wondering if the camera’s broken or I accidentally zoomed in or something. And then I pulled out the manual and read about the feature and thought that’s brilliant, they’ve managed to make the game accessible to general players while solving the age old “desktop general” battlefield intelligence issue, all while folded into the same game engine.
JW: Yeah, it was actually a mod from Take Command Second Manassas that a guy on the forums made and the response was pretty positive. It’s hard, yeah, but for people who really want to understand the challenges of the real Civil War command experience short of actually getting shot at, that’s where we were aiming with that mode, to make it as completely realistic as possible.
It’s one of the things Larry Tagg has been a strong advocate of, and there’s a community that’s already formed around playing in that mode.
In part four: Wargames versus real war, scenario design, and balancing fun against historical accuracy…