In an earlier post, I looked at the first day of Gettysburg with a number of different games. I take a look now at, not one, but two of the three-day scenarios, and looking at how they take on the second day. Obviously, there is quite a departure from the historic situation. Also, perhaps inevitably, in the computer’s take the action gets started right away, at sunup, rather than taking until 4 in the afternoon. In the actual battle, the armies took most of the day to get situated but a computer game is never going to be so patient.
As I continue reading The Gettysburg Campaign: A Study in Command, the author begins his analysis of the second day with the big picture – and several of the major criticisms of leveled upon the two field commanders on that day. For Meade, an accusation was made that he was afraid to confront Lee at Gettysburg. How one of the stories tells it is that, on the night of July 1st, he met with his corps commanders and proposed withdrawing from Gettysburg and falling back on Washington. In this story, it was only the unanimous insistence of his subordinates that overrode his instincts, resulting in the victory at Gettysburg.
This scenario seems to have been dreamed up, or perhaps deliberately engineered as a political attack on Meade after he was removed from his command of the Army of the Potomac. The culprit here was Sickles, who has justifiably fallen under criticism for his actions on the morning/early afternoon of July 2nd, and his attempt to shift onto Meade the blame for his poor performance at Gettysburg. Specifically, Sickles has claimed that he moved his Corp forward (ahead of the Union line, into the “Peach Orchard”) due to a combination of a) an absence of clear orders from Meade and b) a desire to force decisive, victorious confrontation with the enemy in the face of Meade’s reluctance.
While the written record has been used to support Sickle’s version of events and helped create the narrative of Meade wanting to withdraw, the author believes the records for Meade’s preparing withdrawal are all based one of his alternate plans of forming a defensive line in Maryland. Evidence would support that he no longer considered this an option once he was engaged in Gettysburg. Furthermore, the written record shows that Meade, shortly after arrival on the field, began making plans for attacking Lee’s position. Further evidence supporting Meade’s side, the record shows that he desired to make use the 6th Corps, still en route to Gettysburg, on the attack. Waiting for the 6th to come up would allow him to attack with a full corps; one not yet depleted through engagement. The timing of the arrival meant that he was already defending against Longstreet’s attack before he could organize an attack of his own.
Longstreet, and his argument with Lee, is the other major top-level decision the author analyzes. Longstreet described in detail (after the battle) how he objected to Lee’s decision to engage on the second day, as well as objecting to the particulars. It was Longstreet’s suggestion to “move to the right,” although it is unclear whether (at the time) he intended a more sweeping flanking maneuver or whether he actually wanted the Army of Northern Virginia to disengage, move around the Union’s left flank, and attempt to draw them out of their position. The book questions whether either of these counter-proposals would have been battle-changing maneuvers, as Longstreet insisted in hindsight. In the book, ultimately, Lee’s decision is defended and it is Longstreet’s leadership that is called into question.
I left off at night on July 1st, with my union troops still in command of town of Gettysburg itself. As I stated then, my plan was to rest my armies where they were, figuring my computer opponent would do the same. My historical counterpart would never have taken that risk, but to me it seemed to be better to have fresh troops coming in than to have tired troops in place.
The result is that on sunrise on July 2nd, I’m fighting something like the historical battle of the evening of July 1st. I have weak positions north of the town, which are being overwhelmed by the Confederate numbers, but as those attacks are delayed, I’m bringing up what will ultimately result in superior numbers into a stronger defensive position south of the town.
The situation that develops as a result of all this exposes some of the weaknesses of the Civil War Generals 2 engine, problems that weren’t evident on the first day of the fight.
It is the characteristic of many single-player wargames that a computer opponent that might competently react to a player’s attacks falls short when required to go on the offensive against a player. The Battle of Gettysburg is characterized by the Confederates, despite the odds, seizing the initiative in three separate attacks spread over three days. If the battle is to play out historically (with a union player), the Confederate AI must show similar initiative. In the game, the computer does aggressively press the attack. However, the actual disposition of the armies at Gettysburg resulted, to a large extent, as the union reacted to several attempts of the South to flank and envelope their defensive position. As we can see below, the AI has concentrated its forces in my center, pushing through the town of Gettysburg. Thus my defense is more-or-less a straight line.
Concentrating all my forces in the center, in this way, exposes another feature of a brigade-level model at this scale (200 yard hexes?). Each hex is restricted to holding a single brigade (with a few exceptions, like artillery), so reserves must stacked up in hexes behind the “front line.” This creates a situation (illustrated below) where nearly every hex from the front back 4-5 or more hexes is occupied by a friendly unit; either a fresh reserve, or a depleted unit that has been pulled back to recover. In turn, that prevents much in the way of maneuvering, as all the space to move is taken up. All of this may, reasonably, simulate the command issues that would prevent rapidly shuffling armies back and forth, but it creates a much more claustrophobic feel to the battle that doesn’t really match reality, nor is it reflective of other games’ treatment of similar situations.
Noting the development of the battle moving into the early afternoon, it is a stark contrast to the real battle where nothing of note had happened yet. The fighting of July 2nd began only at 4PM and continued until darkness, thus preventing significant maneuver relative the dispositions in the morning. The game, pretty much any game, incentivizes immediate and continuous attack. At least one side has an advantage and would be wasting time/victory point opportunities is he wasn’t engaging the enemy and rolling for damage.
In reality, the morning and early afternoon wasn’t silent. There was pretty much a constant engagement of skirmishers on the Union right, but as far as the brigade-on-brigade, full scale fighting – nothing. The difference, perhaps, is in the command and control. As a player, computer or human, you simply move your units, whereas the real battle required coordination and communication up and down a chain of command. Some games in this period, The Guns of Gettysburg being one notable example, rules are put in place to put a premium on organized attacks. Such rules can be massaged to duplicate the tempo of the civil war battlefield, recreating the partial day lull that happened. A ground-up simulation may be impossible. Knowing what we know now (as any wargame commander would), Lee and Longstreet would never have delayed their attack until 4PM. So much so, that one wonders why it seemed at all feasible at the time. The heat, the miscommunications, the friction between personalities, all created a unique unfolding of events that might never be reproduced with a roll of dice or a draw of cards.
As 4PM comes, I’m actually moving my arriving 6th Corps around the Confederate flanks and launching my own attack.
For a different view, I’ve dug out an even older game on the subject, Age of Rifles, which also features a 3-day Gettysburg scenario. As I played through the first day with this version, the AI enemy very aggressively pursued me through the town of Gettysburg and, in fact, began assaulting my positions on Cemetery and Culp’s Hills. As I had mirrored my historical counterparts and began readying a fall-back position there, I easily held those defenses through nightfall. However, the computer opponent was also making a concerted effort to get around my left flank throughout the late afternoon of July 1st. As a result, I found myself having to create the full “fishhook” defensive position down Cemetery Ridge and onto the Round Tops to ward off any end runs.
The screenshot above shows that the enemy encamped within rifle range of my defensive positions. If your eyes can manage it, the minimap (orangish square in the lower right – click on the graphic to enlarge) shows the full “fish hook” beyond the borders of the tactical view. The bright white is the Confederate and the off-white is me, the Union.
The Other Sandbox
Some time ago I wrote about the trinity of games for different levels of combat. My point was that any game released dealing with the Cold War era needed to be better than the existing scenario available for the appropriate sandbox games.
Several years before releasing The Operation Art of War, (note Harpoon is nearly a decade older than the other two), Norm Kroger designed the game Age of Rifles. Released by SSI its full title was Wargame Construction Set III: Age of Rifles 1846-1905 and, as the name implies, it was supposed to be a user-extensible system for creating battles from after the time of Napoleon but before the First World War.
Age of Rifles followed two years after Wargame Construction Set II: Tanks! These two were largely unrelated, except for the publisher, to the original Wargame Construction Set released in 1986. That original package was a work based on the game designs of Roger Damon, namely Operation Whirlwind, Field of Fire and Panzer Grenadier. The concept was that, by making the statistics of units fully editable, scenarios from the dawn of civilization to present time and beyond could be created and played. The reality was mixed. The graphics were considered mediocre, even for 1986, and the engine itself was limited.
With Wargame Construction Set II, SSI and Kroger created a game more limited in scope but more successful in execution. The game brought the series (and the work of Kroger) into modern (albeit 1994 modern) graphics and user interface. The scale was that of Panzer Blitz, and thus that of the Tiller series (the first of which, Battlefield: Bulge-Ardennes game out a year later). There are those who say, even today, that it is the best treatment of armored combat at that scale. I’ll have to take a look at it in a future post.
With Wargame Construction Set III, Kroger moved into the same time frame (probably not so coincidentally) as Tiller with Battleground: Gettysburg (actually quite co-incidentally, both are 1996 releases). Like Tanks! before it and The Operational Art of War to follow, while the game released with various scenarios available, the goal was to enable the user to implement any period-appropriate (and then some) scenario desired. There are three different scales (technically 3X3 as map size and unit size can be set independently), and people continue to keep the system alive over 20 years on.
Age of Rifles was released the same year as the original Civil War Generals. Civil War Generals 2, which was released the following year, expanded the game by adding more battles to the campaign and providing an open-ended scenario editor. Particularly the latter was ripe for comparison to Age of Rifles, itself heavily focused on American Civil War battles.
Very briefly, the compare and contrast between the two is that Age of Rifles has more detail in the tactical level, but a more sparse campaign interface. Civil War Generals 2 focuses on the campaign and, for better and for worse, simplifies the tactical level.
Several of the Age of Rifles features make sense. I’ve taken the other side of the argument before, but there is a good argument that Civil War battles need a regimental level of maneuver. The more complex disposition of said regiments also seems important to this era. Do you chose to deploy for speed of movement? For offense? For defense? Do you spend turns, while making yourself vulnerable, digging in to enhance that defense? All very appropriate. But at what cost? The carpal tunnel syndrome resulting from having to drag a mouse over every regiment at Gettysburg repeatedly for almost 200 turns is a high price to pay for that additional flavor.
As this article hints at, the Age of Rifles AI seems to have the upper edge as well. The aggressiveness of the Confederates has exceeded that of their historical counterpart, and it has kept me on edge. Not to say there aren’t issues. Frequently, routed units run towards the enemy rather than away, provoking a cycle of opportunity fire and counter fire. Of course, details like this can be dismissed if, perhaps, the attrition of a broken unit winds up historically accurate in the end. I’ve also witnessed some repeated cavalry charges against dug-in infantry that ended up, predictably, in utter failure. Over all above average performance, but with a real range of results in game. For another example, on one turn the Confederate AI is aggressively trying to flank me, and on another turn I can see units camped out within the range of my rifles, deployed facing nothing in particular.
All this extra micromanagement provides more historical flavor, but it also risks creating ahistorical results. While certain aspects of command and control are modeled, one still commands every unit on every turn. It is true, for example, that units with lowered morale become more difficult to command. Further, there is a loss-of-command that comes from splitting up a brigade that I don’t fully understand and have been too lazy to research in the manual. However, because every unit remains under my direct control, I can do things like ride some cavalry into the enemy rear to spot his dispositions and then immediately began moving (within the same 20 minute turn) units from my own rear to react to that new information. There is no delay in communication and orders, and no apparently loss of control when sending a division or a corps winging through the woods in pursuit of an enemy flank.
It makes me wonder whether the computer AI benefits from some of these same cheats. If not, the answer may be simply one of discipline, holding the pace of one’s reactions to what is realistic from a battlefield communications standpoint.
The Best Defense
Returning to Day 2 of my game, I find myself suddenly adopting a 1800s mentality toward the battle. As dawn arises, the Confederates immediately resume their assault on my positions, particularly at the northern curve of the fishhook. For myself, I’ve filled out those defensive as the sun was setting the night before so, unlike Meade on July 2nd, I’m not worried about unoccupied positions on Cemetery Ridge and Little Round Top. Furthermore, as I positioned units along my line and dug them in, additional arriving brigades were held in reserve. Of note, I’ve good a good chunk of fresh units hiding behind the round tops, waiting to see where they are needed.
Similarly to Civil War Generals 2, although not quite to the same extent, a reinforced defensive position starts to run into “crowd” issues. Here, there is a hex limit of two regiments (or substitute a battery) per. This means that to reinforce a front line unit that has been weakened or, perhaps, just run out of ammo, that unit has to be pulled from the line into available space behind the line. Then a fresh unit has to be moved into the front position, it being more vulnerable to enemy opportunity fire while doing so. Both these incoming and outgoing units need the same space behind the line, which also is filled by additional reserves. Having short interior lines is great, but it makes things really crowded in the dozen or so hexes right behind those lines.
With that handicap, it seems like only a matter of time before one of the Confederate results will see some success. If they manage to drive me out of my entrenched positions, it will be difficult to set up a fallback position – there’s just not much left space to move around from the Union’s historical position*. It strikes me that the best way to relieve pressure on my defensive troops is to go on the offensive in other parts of the battlefield. Specifically, I’m going to try a double envelopment, using the reserves to hit both flanks simultaneously (although concentrating on my own left).
Fast forward a bit to see the consummation of my plans. As the hour approaches noon, the right end of the Confederate line is starting come apart as I hit it from both sides. Of course, even as I’m still getting reinforcements I can see that the Confederates are too. (Of course we knew that – much of Longstreet’s corps arrived that day). It now just seems a matter of what gives first – my boys’ forward momentum, the Confederate position on Seminary Ridge, or my wrist.
*Not really true. Meade contemplated falling back further from what he perceived as a weak position, with Confederate troops on his left flank (Big Round Top), to a more defensible position behind Rock Creek, with Wolf’s Hill anchoring his right. The map in this scenario is big enough that such a position could probably be played. However, I’d rather go for the gold in a glorious charge than slog through the mundane task of pulling back from one defensive position to another. Especially since that was what I just did on Day 1.