, , , , , ,

I’ve written several posts on the Gettysburg Campaign. The first is a combined book/movie/game review around The Killer Angels. I followed that up with some more thoughts on the Scourge of War command/AI system. This is the third I’ve written on the subject, discussing to tangentially-related topics. First the larger campaign in the context of a book and gaming. Second, a discussion of the level of necessary modeling for a proper Civil War battle game.

I expect a series of thoughts as I work my way through the rest of the book…

Having begun contemplating the Battle of Gettysburg, I got my dander up (as Harry Heth may or may not have said). I started in on the book The Gettysburg Campaign: A Study in Command. This is a sizeable book that addresses not only the battle itself, but the bigger picture leading up to the battle.

The book starts out, while not explicitly, with a debunking of some of the foundation of the Killer Angels. With the printing of this particular version being 20 years ago, I looked at it as a sort of counter-revisionist history. At some point, however, I actually looked at the front matter and the original date of this work is 1968, meaning it preceded that novel by almost a decade. Clearly the influence of Longstreet’s writing had a hold on the popular conscientious before the Killer Angels.

In particular, Professor Edwin Coddington (the author) questions the idea that Longstreet had agreed with Lee on structuring the campaign around the defensive, and that the Battle of Gettysburg was entered into against his long-standing advice. The evidence, in Coddington’s view, suggests that this view of the Campaign was conceived by Longstreet after the campaign and perhaps after the war. Correspondence from before Gettysburg show Longstreet as an enthusiastic supporter of Lee’s plans for taking the army north. There is no evidence that such support came with conditions, such as the agreement to seek a defensive strategy once in enemy territory.

Another stark contrast between the book/film and Coddington’s research is the scene where Lee realizes that they need to turn to fight the approaching Union army. In the movie, unable to read the map, he asks “What town in this?” The answer is Gettysburg. By contrast, Lee in fact expected the battle in the North to take place near Chambersburg, York, or Gettysburg. Gettysburg may have been his preference. While moving North, he deliberately telegraphed his positions with the intent of drawing the Union army towards him. When he realized he had succeeded, he deliberately and leisurely positioned his forces around Gettysburg with the intent of fighting a battle much like the first day at Gettysburg turned out.

Furthermore, the days and weeks leading to Gettysburg were far from a peaceful march from Virginia to Pennsylvania. Various actions occurred all along the route, ranging from the cavalry battles at Brandy Station and near the passes to the overwhelming victory of Ewell and Early’s assault and capture of Winchester. The book also highlights the importance (and impotence) of Darius Couch’s Pennsylvania militia, as the defending force in Pennsylvania up until the arrival of the Army of the Potomac.

Looking at what I’ve previously thought of as simply a march north in in new light, I decided to break out my old copy of AgeOD’s American Civil War. This was a follow-on to the extremely well-regarded Birth of America, a game which covered the operational level of the French and Indian Wars and the American Revolutionary War. AACW, as it is often abbreviated, expanded on the engine and further added the management of the wartime economy making it a game of vastly greater scope and scale. Many thought this made the resulting game too complicated, losing the charm that made its predecessor so successful.


“We did not want the fight but the fight is here!” Longstreet moves his Corp into Gettysburg in the opening turn.

Disappointingly, given my goals, I found that the Gettysburg scenario starts at the end of June with both armies situated just outside of the town. For the battle itself, it seems that, while there might be a few choices for the player, the outcome is going to be largely depended on the roll of the virtual dice.

Devil’s Details and the Brigade

I also continued to work my way through the Ultimate General campaign, and have warmed to it the more I play. My original impressions were that it was a little on the simple side.  Perhaps it is better thought of as more of a complex RTS than a simplified wargame. Specifically, I did comment on the unit size; the basic maneuver unit being brigades rather than the more commonplace regiments. My other first impression was that the game was very difficult. In several plays, I lost every time.

Since that first writing, I’ve played all the way through the campaign from both sides. My result was that, as Confederates, I suffered the historical loss. As the Union, my first two days were victories (again historical), but on July 3rd I was taken by surprise by a Confederate envelopment. Pickett’s command actually hit me from the rear, and the enemy used my distraction to push me off Cemetery Ridge.

The play got me thinking about the level of abstraction, and what is appropriate. Appropriateness, of course, is relative. Does that mean ever more realism and accuracy of the simulation? Does it mean more fun to play? How important is accessibility? For the target of this particular game, the ability to pick up the mechanics quickly and to be able to complete a game within a 20-30 minutes is probably at a premium. When going in this direction, how much and what kind of abstraction will retain a good simulation of command, even if the details don’t appear to be simulated?

The original Avalon Hill Gettysburg game came out in 1958, and was one of the first of the modern wargames. To the eyes of today’s game player, it was fairly primitive and the package does not come together well, but its fair to say that all games that followed owe a debt to the design. The game was played on a square grid, over a fairly attractive map using counters representing Divisions. The rules were rewritten and the components re-released several times by Avalon Hill, culminating in much more modern-looking 1977 version using hexes and traditional counters. Still, the maneuver unit is at the Division level. Avalon Hill had a 1993 release, Roads to Gettysburg, again with primarily Division level unit markers, however this was meant to capture (as I started off with at the beginning of the article) the larger campaign.

The most modern, and probably most regarded board game treatment of this battle is The Guns of Gettysburg. It is a hex-less (and square-less), card-less, and dice-less treatment of the Gettysburg battle. It is part of a series of games taking place in this era (Napoleonic and American Civil War) that uses wooden blocks and nodal maps to represent the combat style of the time. I don’t mean to get into a discussion of the game itself except to note that, again, the maneuver unit is somewhere between the division and the brigade (multiple units per division but not necessarily one per brigade).

The point of all this is that, when it comes to boardgame representations a huge battle like Gettysburg, breaking the units down into brigades would be considered a “serious” treatment. Within that context, the Ultimate General design make not seem quite so simple. I’m further influenced by looking ahead at the next in the series, Ultimate General: Civil War. I don’t have this game; it’s a Steam Early Access game, which seems like paying for the privilege of being a beta tester. However, I can see from discussion, screenshots, and design notes some of the direction that game is headed. It also has me considering Ultimate General: Gettysburg to be more of a testing-the-waters for the fuller vision of the new game.

From what I can tell, Ultimate General: Civil War will, in fact, take the simulation down to the regiment level. The comparison I would make, rather that to the Sid Meier game, is to Sierra’s Civil War Generals 2. That game, which I still might be playing today if it worked easily on current operating systems, provided the battles of the Civil War linked together in campaigns. At the campaign level, management decisions were made about supplies, weapons, and manpower as well as high level strategic options. Those decisions then influenced the order and details of the tactical battles.  It was a very engaging game that consumed a lot of my time back when it was new. This may be the direction that this series is trying to go.

Reading Steam comments about the early access for Ultimate General: Civil War highlight some existing problems, that hopefully will be ironed out in due time. Most of the comments seem to be about the campaign system, and how well previous results are factored into the next battle. Based on my playing of Ultimate General: Gettysburg, if I had one request it would be better modelling of supply. It isn’t so obvious with infantry; when infantry units engage, they gradually become depleted through a combination of morale, physical losses, and (perhaps) ammunition. However, a battery of cannon can, if out of range of enemy units, fire on the enemy positions indefinitely.

While the standard for computer wargames tends to be at the Regiment, rather than Brigade level, I took a peek at Battleground 2: Gettysburg to remind myself of that design. Although the second of the John Tiller/Talonsoft series, it was the first of the “Age of Rifles” games. It also used the Brigade as the unit for the “stands” (it mimicked the look of the table top wargame, although played on hexes).

In the HPS iteration of the the Tiller Civil War engine, he reimplemented Gettysburg battle at a much finer level of detail as well as capping it with an operational level “decision” interface that guides the tactical level battles. While there are a dozen Civil War titles for the PC, only eleven are available on the tablet (and that includes one demo game, not related to one of the battles). Perhaps because Gettysburg was one of the first, if not the first, conversion to the new engine, it is not on the tablet.


This is most certainly not Gettysburg. It is a tutorial scenario, but one that vaguely corresponds to Buford’s July 1st Gettysburg defense.

For some further compare and contrast, I did take a look at the demo app on the tablet. For continuity sake, one of the tutorial scenarios involves a cavalry unit (under Buford) defending a ridge against attacking Confederate infantry. It certainly sounds pretty Gettysburg-like.

The Tiller offerings are thought of as at the serious end of the military simulation spectrum, which is why I think it makes an interesting compare and contrast. As I said, the unit size (regiment) matches where the Ultimate General: Civil War appears to be headed. But just look at that row of buttons across the top. Surely, with all those options and all that you do every turn, it must be a much deeper game? What “serious” options are left out of the more accessible version of the Battle?

The turn-based versus continuous time is one obvious difference. It certainly changes the way one tends manage their game. In a turn based game, a player often feels the need to consider each move carefully, whereas in a real time game you’re forced to move your focus away from some parts of the battle as you focus on others. In theory though, when a RTS is pausable, there is opportunity for equal amounts of deliberation. Further, continuous time (or at least a simultaneous execution of turns) eliminates some gaminess surrounding the order in which you make your moves. Point being, I don’t automatically consider one style or the other as superior.

Several of those interface buttons along the top have to do with facing, which is also modeled in Ultimate General, and with their cooler interface. What isn’t in the Ultimate General interface is the change from column to line, or the limbering/unlimbering of artillery. Ultimate General does this “automatically” when you make a move. For short moves, cannons are pushed forward into their new position. For long moves, cannons are hitched to their horse teams and infantry forms into a column. When playing the various tiller games, I often find that the formation is more something to be forgotten (especially in more modern settings where there is column/deployed and mounted/foot) rather than a fun addition to gameplay. The question with letting the computer do it is, is everything modeled correctly?

Similarly, the choices for targeting units in the turn based game do not necessarily add to the experience. In Ultimate General, the player can direct fire, but the default is that units choose their own targets. This actually makes more sense in terms of realism. From the battlefield commanders standpoint, units are apt to fire on the units that they feel are most threatening to them, not the most important targets from command point of view. Granted, the automatic resolution of “opportunity fire” makes the two systems, again in theory, pretty similar, the real time version (to me) feels more natural. Finally, they both have options to enter melee combat, and are thus pretty equivalent on that level.

There are other details that the Tiller version probably models better, and certainly models more explicitly. I do think that the Ultimate General system is not targeting the same level of fidelity, and so a one-to-one comparison isn’t necessarily fair. My point, however, is that something like Ultimate General could easily challenge the something like the Tiller games, if taken in that direction.