Home Store Products Research Design Strategy Support News

22 July 2009

 Products The Guns of Gettysburg Design Diary Oh, Say Can You See

Oh, Say Can You See

Over a year ago, in my 26 May 2008 diary entry, I showed the map I had made for the game, exclusive of all the regulatory apparatus that makes a map into a game board. In this diary entry, I want to talk about the process of developing that apparatus, the essential step that converts a map into a board for a game.

One of the things that really struck me in looking at maps showing troop positions at Gettysburg was the distance between the opposing armies when neither army was carrying out an attack. While Napoleonic armies often started battles at a considerable distance from each other, as the battle developed, distances tended to close quickly, and once closed, the armies tended to be locked at close ranges until one side or the other broke, ending the battle. This was not true at Gettysburg. At Gettysburg,there were long intervals when the armies remained at a distance, then an attack was made, a relatively brief period of combat ensued, and then one side or the other retreated, and the distance between the armies was re-established — until the next attack, when the process would repeat itself.

Take a look for example, at this map of the battle on July 2, carefully prepared by John Bachelder, the premiere 19th century historian of the battle. As you can see, prior to launching their attacks, the Confederates kept a distance generally between 500 and 1500 yards from the Union positions, distances they largely resumed after the attacks ended:

Gettysburg No-Mans Land

Click on the image above to open in its own window

This intermittent pattern of combat at Gettysburg had a character that reminded me of World War I (although not to the same degree) with its extended periods of relative quiet, punctuated with periods of furious, but short-lived combat. What gave World War I this character was the intense defensive fire-power the armies possessed. This enabled them to establish a killing zone (no-man’s land) to their front that forced the enemy to keep their distance. Crossing this zone was a huge event for an enemy army for which no amount of preparation ever seemed enough.

Now, Civil War armies did not have machine guns, the signature weapon of WWI that was one of the main differentiators between WWI and earlier wars. However, Civil War armies did have large numbers of artillery pieces that, by the standards of earlier periods, were extremely accurate, and they also possessed a sniper threat from marksmen who could actually use their rifled muskets with enough skill to fulfill the long-range potential of those weapons. A Civil War army in a defensive position did not project as much lethal firepower as a WWI army, but it did project enough of a threat to create a killing zone that could not be occupied by the enemy for any significant length of time.

It seemed to me that this was something that I needed to capture if I was to make a successful game about Gettysburg. However, it was not something that the game system that I used in my Napoleonic games could do. A new system would be needed.

In studying this problem, I began laying troop positions on the map I had prepared. In doing so, I was struck by the strong preference the armies had for positioning themselves on ridge lines. The reason did not require much analysis: the positions were being selected based on the fields of fire that could be projected from them. One reason that elevation was required was because even “flat” ground isn’t usually very flat. It typically has numerous shallow undulations that a medium-resolution relief map (such as one with 10 meter contours) doesn’t show. Also, there is a lot of ground clutter that can obscure sight lines; such as the fields of crops that were growing over much of the battlefield at Gettysburg. Units on “flat” ground, even if there were no ridges or woods in front of them, often could not see very far. Coster’s brigade at Gettysburg, for example, took a position on 1 July north of Gettysburg, could see only a very short distance to their front, even though if you look at a contour map you might think that they would be able to see for a thousand yards to their front. (And in fact there have been Gettysburg games published where that is exactly what they could do.) Usually, to see any distance, a unit had to seek some sort of high ground that would enable them to see over the ground clutter and look down into the undulations in the ground from above.

I felt that in order to have a decent simulation of the battle, I needed a decent simulation of fields of fire. Oddly enough, it was because I had an idea for how to model fields of fire that I decided to start on Gettysburg to start with, but the more I thought, the more (incompatible) ideas I had on the subject. And so, the game turned into a big testing ground for these different models. What all of these models had in common was that they would attempt to produce field of fire results explicitly through the design of the regulatory layer of the board, rather than relying on algorithms and lines traced through hex-grids, which is the traditional wargame approach to the subject.

Before going into the Guns of Gettysburg design process, it might be helpful to look back at how areas worked and high ground was modeled in my Napoleonic games. In Bonaparte at Marengo the area design was largely concerned with reflecting movement difficulty. (Open ground was represented with large areas and congested terrain with small areas.) The boundaries of the areas generally followed terrain, in particular the towns, the Fontanone stream, and the road network. The battlefield at Marengo was extremely flat, with a low ridge line being the only relief feature of any importance. The representation of the ridge line was not really satisfactory from a design point of view, because it didn’t reflect the advantages artillery derived from being placed on high ground. Napoleon’s Triumph handled high ground quite differently and much more satisfactorily, which it had to because of the much greater importance of hills in that battle. In the Austerlitz game, some locales were marked as hill locales, and artillery located on hills could fire with double the effect of artillery located on level ground. You can see the difference graphically between the two games below. In Marengo there was no actual area for the ridge; it was just a boundary between areas, while the Pratze Berg hill at Austerlitz is modeled with actual areas for the hill itself:

High ground in Bonaparte at Marengo
High ground in Napoleon's Triumph

Neither game, however, did anything more ambitious than to abstract fields of fire into greater artillery effectiveness. In part this was because the area design didn’t lend itself to modeling fields of fire (it wasn’t supposed to), and in part because the artillery was relatively short-ranged compared to the game scale, which made abstraction a more suitable approach to the problem. In trying to model fields of fire, The Guns of Gettysburg would have to break away from the entire map design methodology used in the previous game and go off in a new direction.

Well, I was ready and eager to give it a go. I had a lot of new ideas and was anxious to see how they worked. My first attempt at a terrain model for the game was this:

First Terrain Model for The Guns of Gettysburg

It would be very natural for you to look at that all those circles and arrows and so forth and wonder how it worked. The answer is that it didn’t work. I had this vague idea that I would include actual distances on the map and that the lines could serve as routes for movement, routes for fire, and for defensive positions. The triangles on the map represented hills, and that the circles represented obstructed ground. It all looks interesting, but I never had any rules for this that actually allowed pieces to move or fight on this board. The more I looked, the more perplexing it became how it could ever be made into a functional game board. After a while I simply decided to give up and junk the whole thing. Time to try again.

Second Terrain Model for The Guns of Gettysburg

Try not to obsess about the use of the color red here. The red butterfly shapes were construction lines and were supposed to be filled in with various little gray arrows indicating line of sight. I never finished them for the entire map: I abandoned this entire effort long before I got that far. The gray lines were supposed to be actual artwork, but were fairly bizarre looking and I never warmed up to them even as just a graphical design. (I kind of liked the way the previous version looked; it just didn’t, you know, actually work.) The little triangles on the ends of the line were supposed to represent the direction from high to low ground, with the number of arrows indicating the number of blocks that could fit on the line. Where you see little squares instead of triangles, that meant that the ground was level along that line. The odd zig-zags were supposed to indicate steep slopes.Like the previous version, this map design was completely non-functional. No rules ever existed to regulate movement or combat on it. So, with two complete failures under my belt, I threw away everything and started again:

Third Terrain Model for The Guns of Gettysburg

This version, unlike the previous version, was actually playable. There were some functional (if extremely fluid) rules that allowed you to move blocks around on it and determine what blocks had fields of fire where. The sizes and shapes of the areas on this map are based on fields of fire from the lines bounding them (the bounding lines function rather like approaches in the Napoleonic games in that blocks can occupy them during play). The little cannon symbols indicated high ground which would increase the effectiveness of defensive artillery fire, and the little triangles indicated steep slopes. The double lines (the only example of which in this illustration is in the top-left corner) indicated obstructed terrain. The dotted lines represented extended lines of sight: positions that had lines of sight to each other that were exceptions from the normal line of sight rules. I won’t attempt to explain to you in detail how this all worked, because there was no one way they worked: there were lots of different ways as I was constantly tinkering with the rules. (Not that the rules were ever written down for this design; it is not my habit to actually commit rules to paper until fairly late in the design process.)

The straight lines, by the way, are called positions in this game, although they function much like approaches in the Napoleonic games.

Anyway, even if the rules were fluid the actual model proved fairly stable, although it was subjected to many iterations of refinement. The main changes were to shift in the basic design focus in two directions: (1) simplifying the model, and (2) improving its accuracy for defensive fields of fire. You can see the result below:

Fourth Terrain Model for The Guns of Gettysburg

The key improvements of this version compared to its predecessor are: First, that the defensive artillery strength of a position, which was resolved by formula in the previous version, is printed directly onto the map in this version. (That’s why you see cannons on both sides of positions in this version and with 1, 2, or 3 cannon symbols per position; the number of cannons is the size of the defensive artillery fire bonus.) Second, the dotted lines of sight, which were mainly used for offensive fire in the previous version (and were straight lines connecting two different positions to each other) are in this version used for mainly for defensive fire (and are now curved lines connecting a position to an area).

It is too big a job to explain all the rules for movement and combat in this context, but I can explain a couple of the basic features, which are quite different from the Napoleonic games. The first is that while in the Napoleonic games, blocks could occupy the edges or interiors of areas, in this game they could only occupy the edges of the areas. (There is no concept of blocks being “in reserve” in this game.) Another basic difference is that in this game, a block projects a field of fire into the area in front of the position it occupies, and, depending on terrain, also sometimes into adjacent areas to the left and right of its front area as well. Blocks can only cross an enemy field of fire area to attack, which forces a gap between the opposing armies. During an attack, the attacker crosses the field of fire area, the winner and loser are determined, and then the loser must withdraw to the other side of the winner’s field of fire, re-establishing the gap between the opposing armies. This, in broad strokes, is how the game works to simulate the basic nature of Civil War combat described in the start of this essay. A more detailed presentation of how all of this works in the game will, however, have to wait for a future design diary entry.

Before I close this entry, I will include one more thing: the full game board, with its entire regulatory apparatus. (Play aids omitted.) You can click to open the map in its own window where you can see it more detail.

The Guns of Gettysburg Terrain Model

Gettysburg Terrain Model

Click on the image above to open in its own window