Part 1
Expanding on the work of Herman Hattaway and Archer Jones in How the North Won,1 we will evaluate the performance of four important commanders of the Union and Confederate armies. Hattaway and Jones used data compiled from Thomas Livermore’s Numbers and Losses,2 to compare relative performance of union and confederate forces. They chose 26 battles from 1861 to 1863 for which Livermore provided reliable figures and where one side had lost at least 1000 in killed and wounded. Hattaway and Jones dealt with questions of the relative efficiency of Confederate attacks and the effectiveness of General R. E. Lee's apparently aggressive strategy. The four commanders selected are U.S. Grant, W.T. Sherman, R.E. Lee and T.J. Jackson. Hattaway and Jones’ database of 26 battles has been expanded to a total of 50. This was especially important to cover the battles where either Sherman or Jackson were in command. Also a number of battles between Grant and Lee were added. Hattaway and Jones had excluded these because they had occurred in 1864 when Confederate reports were often incomplete. However, The Civil War Dictionary3 by Mark Boatner contains a number of references to Livermore's figures for engagements not included in the 26-battle database.
Hattaway and Jones developed a system based on battle statistics to evaluate Lee's performance in relation to all other confederate commanders. The system involves five measures of performance. They are: 1) Friendly force engaged as a per cent of the opposing force. 2) The per cent casualties of the friendly forces. 3) Friendly casualties as a per cent of enemy casualties. 4) Livermore's efficiency index, where the efficiency of the friendly force is expressed as the number of killed and wounded on the opposing side per thousand friendly forces engaged (also referred to as the hit/1000 index). 5) The efficiency of the opponents force (enemy hit/1000 index). The first of these five measures indicates the relative advantage of forces enjoyed by each commander. A commander who frequently has the advantage of forces available should be expected to achieve better results. The second measure is a general indicator of the commander's effectiveness in avoiding losses. A key measure of losses in relation to results is given by the third value. The two remaining figures are regarded as the best indicators of the relative combat performance of the opposing sides in any action. Comparing these measures for a number of commanders can show where their leadership proved to be superior to the average commander and that certain individuals attained a higher level of understanding in commanding military operations.
Following the lead of Hattaway and Jones, the first step in the evaluation is to compile data on the tactical performance of each of the four commanders. For the most part this was a straightforward process of tabulating figures for each engagement on numbers engaged and casualties. Some questions arose during this process. The first was the question of confederate losses at the battles of Wilderness, Spotsylvania, Cold Harbor and in the Petersburg Campaign. The Civil War Dictionary referenced Livermore's estimated number of confederate losses for the battle of the Wilderness and this was accepted. For Spotsylvania and Cold Harbor, Livermore had not compiled a complete figure. However, The Encyclopedia of Military History4 gives estimated figures for both of these actions and these are used in the expanded database. There were three major assaults during the Petersburg Campaign. Livermore gives a figure for Confederate losses only for the Battle of the Crater. The losses for Lee's forces at the First and Final assaults have been estimated from the efficiency index of the attacking Union forces at Cold Harbor and the Battle of the Crater. Since these actions involved many of the same forces and commanders and the tactical situations are similar, Union assaults against confederate entrenchments, these estimates represent the most reasonable figures available. For Grant's Assault at Vicksburg Livermore has no reliable number of Confederate losses. However, Grant in his Memoirs states "I have always regretted that the last assault at Cold Harbor was ever made. I might say the same thing of the assault of 22 May at Vicksburg."5 This indicates that Grant regarded these two engagements as nearly equal in the chance of success and the massive slaughter suffered by the Union troops. We will take that statement to mean that the relative performance of the forces were comparable as well. We've then taken the efficiency index for the Union troops at Cold Harbor and applied it to the forces at Vicksburg.
Another question arose regarding engagements commanded by Sherman and Jackson. The two basic criteria for including battles in Hattaway and Jones's 26-battle database were: 1) that Livermore has a reliable figure for numbers and losses, and 2) that one side had lost at least 1000 killed and wounded. For a number of engagements this second criteria is not met. In order to provide the most comprehensive coverage, battles were included where the basic criteria were not met. While compiling data for engagements a question arose regarding Jackson’s role in the Seven Days battle. Although playing an important part in the campaign, his forces actually participated in one significant combat action- Gaines Mill. In this action most of Jackson's command participated in the final assault that forced the Union retreat. Since the Seven Days battle is treated as a single action it was not included in data for Jackson's tactical performance. Two further questions were raised concerning Lee's record. First, Hattaway and Jones give Lee credit as the commander at South Mountain. But, since Lee was not present on the field this has been taken off his list of battles. Second, Lee is credited with a defeat at Antietam. Yet, it is clear that the Confederate forces held their position against repeated Union attacks. It is well established from federal accounts that the Corps of Hooker, Mansfield, and Sumner were much cut up and not capable of further action.6 In as much as Jackson has been credited with a tactical victory for this battle, Lee also has been given a victory for Antietam. Granted, that Lee's army retreated two days after the battle, giving the Union a claim to a Strategic Victory. For the purposes of this evaluation of tactical performance, Antietam is regarded as a tactical Confederate victory.
Here a comment should be made on comparing Jackson to the other three army commanders. First, since Jackson fought under Lee's command in seven out of ten engagements a meaningful comparison is difficult. Second, that since Jackson never had overall command of an army he should not be included with other army commanders. On the first point, it is readily accepted that since we are making an evaluation of tactical performance, and since the tactical performance of Jackson's command would be included in the overall performance of Lee's army, a valid comparison between Lee and Jackson may not be possible. Here then, the emphasis should be placed on comparing Lee and Jackson's performance to Grant and Sherman. With respect to the second point; although Jackson never commanded an army, he actually did conduct independent operations on a number of occasions. These include his operations against the Baltimore and Ohio Railroad in the winter of 1861, the Valley Campaign, and the battle of Cedar Mountain. He also carried out four semi-independent actions under Lee; the raid against Pope's supply line prior to the battle of 2nd Bull Run. The first two days fighting at that battle; the capture of Harper's Ferry in the Antietam Campaign; the flank march and attack against Hooker at the battle of Chancellorsville. Because Lee and Jackson are the only two Confederate commanders of a stature comparable to Grant and Sherman, and due to the important contribution of Jackson to the Army of Northern Virginia, the comparison is meaningful and will help cast light on tactical performance.
The data has been arranged in five general groups; attacks won, attacks lost, defensive won, defensive lost, and all battles. Values for all Union and Confederate forces from Hattaway and Jones's original work have been included for comparison, as well as Lee's performance in nine battles. Tactical Evaluation When looking at the All Battles categories, Grant had superior forces greater even than most other Union Commanders. Lee's inferiority of forces was as great as Grant's superiority. This is no doubt partially because Lee and Grant share six of thirteen battles in common. Sherman commanded forces roughly equal to his opponents. Jackson's command was inferior to the enemy he faced, but not as inferior as most other Confederates. The All Battles group also points out the ineffectiveness of Grant's attrition strategy. Grant enjoyed a 156% superiority of forces, yet his attrition strategy produced union losses 199% greater than confederate losses. This is especially brought out in statistics of Grant's attacks lost. In losing attacks his forces suffered losses at a rate of 229% of the confederate defenders while enjoying a superiority of 165%. While the argument for attrition strategy was that Grant would "bleed" Lee's army, in fact it was Grant's army that was being bled. The political reaction to Grant's losses during the summer of 1864 nearly cost Lincoln the election that November.7
Livermore's hit/1000 (efficiency) index provides a second measure of tactical performance. In the All Battles group Grant's hit/1000 value was 72 against 224 for his opponents. In contrast the hit/1000 for all union forces, according to the original Hattaway and Jones data, was 117 to 157. Grant's Record This makes Grant about 40 per cent less efficient than the average union commander. Perhaps this is due to the large number of assaults against entrenchments made by Grant. However, in the defensive won category Grant's efficiency was about half that of the typical union commander. (52/96=.54) Compared with Sherman whose hit/1000 for winning defensive was 187 vs. 73; Sherman's Record Lee 271 vs. 85; Lee's Record and Jackson 221 vs. 137. Jackson's Record Another perspective may be gained in comparing Grant's performance at Cold Harbor to Burnside at Fredericksburg. Burnside's hit/1000 was 51 to 296 for Lee; Grant scored a hit/1000 of 28 to Lee's 220. Confederate efficiency at Fredericksburg was six (5.8) times better, but at Cold Harbor it was eight(7.9) times better than union efficiency.
Battlefield success and frequency of attacks may make another measure of overall command performance. We should expect that over a long period of campaigning a commander is likely to experience a variety of situations requiring some mix of offensive and defensive tactics. Over time this mix will approach a rough balance, 50/50 mix, as a broader range of tactical circumstances is encountered. The most experienced commanders can be expected to come closest to the theoretical norm of a 50/50 mix. However, at least two modifiers to this general rule can be recognized. First; the over all mission or strategic situation may require a predominantly attack or defensive orientation that could throw the mix out of the theoretical 50/50 norm range. Therefore, the norm for the Confederates would tend more toward the defensive, for the Union more toward the attack. Second; while defensive fighting (tactical) offers the advantage of greater efficiency, offensive tactics give the advantage of initiative and an opportunity to achieve a decisive victory resulting in significant political gain. A situation requiring immediate and decisive results would naturally tend toward the attacking mode, while one where losses and force efficiency are paramount tends toward the defensive. Recognizing these modifications, the basic idea of a roughly 50/50 mix over time remains valid. Of thirteen battles Grant attacked in all but one, giving an attack frequency of 92%. For Sherman, Lee and Jackson the attack frequency is 44%, 23% and 50%, respectively. The attack effectiveness, or per cent of success when attacking, for each is 50%, 25%, 67% and 80% respectively. Overall effectiveness for all battles is 53%, 67%, 81% and 90%, respectively for Grant, Sherman, Lee and Jackson. When comparing the efficiency indices and battle effectiveness measures, both show clear superiority of Lee and Jackson over Grant and Sherman. Jackson shows a clear superiority in attack effectiveness. His ability to exploit mobility and surprise, thus creating favorable conditions before attacking, may account for this. In winning attacks Jackson achieved a hit/1000 index of 281 vs. 143 for his opponents, far better than either Grant or Sherman.
A final indicator of command performance for civil war commanders may be success in attaining tactical and strategic surprise. Among civil war battles we find 13 where one side succeeded in attaining a significant degree of surprise.* Only Lee and Jackson achieved a surprise situation more than once. Only Grant, among union commanders, was able to achieve major strategic surprise. Here we follow B. H. Liddell-Hart's distinction of strategic & tactical surprise. The first being based on a wider and deeper logistical penetration and having the enemy's Line of Communication or morale as the target. A Surprise situation is indicated more by the enemy's inability to act effectively or by faulty reactions rather than by open-mouthed amazement of the victim. Effects of strategic surprise are generally less immediate but affects a larger area and last longer (days). Tactical surprise is generally a local tactical arrangement of forces aimed at the immediate dislocation of specific enemy units, i.e.Chancellorsville. A commander's capacity and willingness to attempt a surprise operation shows both greater energy and understanding of the need to create some favorable condition for a decisive victory. The basic ingredients to attaining surprise are; security in planning and conducting the operation; some aspect of distraction or decoy; mobility; superior analysis in determining the weak or vital points of the enemy. This last contains the element of collecting vital intelligence.(*they are Kernstown, Shiloh, McDowell, Winchester, Seven Days, 2nd Bull Run, Chancellorsville, Gettysburg Campaign, Vicksburg Campaign, Chickamuga, Drewry's Bluff, 2nd Winchester, and Nashville.)
Jackson almost routinely sought and achieved surprise. He attempted and achieved some significant measure of surprise in 5 of 10 major actions in which he participated from March 1862 to May 1863. This is perhaps unique not only among Civil War commanders but also throughout modern military history. Of his seven independent or semi-independent actions, Jackson surprised and dislocated his enemy at least five times (Valley Campaign(3), 2nd Bull Run, and Chancellorsville). Only Lee among all other civil war commanders was able to achieve surprise more than once (at Seven Days, 2nd Bull Run, Chancellorsville and Gettysburg Campaign). When use of surprise is taken together with tactical performance, there can be little question that Jackson was the superior commander of the civil war. Lee also showed great ability, perhaps equal to Jackson. No other general, Union or Confederate, and including Grant and Sherman, achieved the success or displayed the mastery of Jackson and Lee.
PART 2
Thomas' Record
Longstreet's Record
NOTES
WORKS CITED