Follow by Email

Saturday, February 27, 2016

Livermore v. Dupuy: Analyzing the Relative Performance of Civil War Opponents

All students of the American Civil War should read Thomas L. Livermore's Numbers and Losses in the Civil War in America, 1861-1865 and T. N. Dupuy's A Genius for War.  Both provide comparisons and methods for analyzing the intensity of particular fights and the performance of the troops involved.

Livermore analyzes according to hits in 1,000 (casualties suffered) and hits by 1,000 (casualties inflicted).  The former figure, hits in 1,000, gives a perspective on what the troops endured, similar to what William F. Fox, in Regimental Losses in the American Civil War, 1861-1865 has in mind when he defines his fighting 300 Federal regiments on the basis of more than 130 or ten percent killed or mortally wounded and then gives charts of Confederate units that lost more than forty percent in particular fights.  The latter figure, hits by 1,000, shows how efficiently the troops performed in damaging the foe--perhaps the more important figure.  

Dupuy incorporates Livermore's analysis and takes it a couple steps further.  First, Dupuy figures out the raw performance of the troops with reference to one another.  That is, for example, the hits by 100 of one side divided by the hits by 100 of the other, and vice versa.  In a longer operation, this is broken down into the casualties inflicted per dayIn "Table 4: Combat Efficiency" below, from my Siege of Petersburg: The Battles for the Weldon Railroad, August 1864, those percentages are in parenthesis.  As you can see, General Lee's army substantially outperformed General Grant's army in August, 1864.  (Beringer, Hattaway, Jones, and Still go this far in Why The South Lost The Civil War.)  Finally, Dupuy modifies the scores for the posture of the forces involved--Attack and Hasty Defense divides by 1.2, Hasty Defense by 1.3, Prepared Defense by 1.5, Fortified Defense by 1.6.  That gives the final column in the table below, "Score Effectiveness."  

Livermore is thus a very helpful starting point, but for actions that lasted no more than a single day, such as Second Reams Station below, one ought to go further and calculate the relative performance of the troops (Why The South Lost The Civil War does this) as modified by the posture of the troops (using Dupuy's formula).  If you look at Dupuy, you will see that the intensity of American Civil War actions exceeds the limits of the examples on his charts, which come from World War I and World War II.  (See note 3, below.)  The full Dupuy analysis is the most helpful for analyzing performance in longer operations, such as the Fourth Offensive as a whole, below.  


Fourth Offensive
                                                                  Average           Percent            Inflicted
No. of        Total               Total                Casualties        Casualties        Per Day               Score
Battles       Engaged         Casualties        per day (12)      Per Day           Per 100 Men[1]  Eff.[2]
USA    3    85,877            9,922                827                   0.96                 0.4 (  28.5%)       0.33
CSA    3    56,795            4.500                375                   0.66                 1.4 (350%)          1.16

Second Reams Station
            Total                Total                Percent            Inflicted                       Score
            Engaged          Casualties        Casualties        Per 100 Men                Effectiveness[3]
USA    8,000               2,727               34.0                    9.375 (  27.5%)            6.25 (  18.4%)
CSA    8,000                  750                 9.375              34.0     (363%)             34.0   (544%) 

[1] This column corresponds to Dupuy’s “Score” and represents “the casualties per day as a percentage of the force inflicting the casualties, derived by applying the casualties of one side to the starting or overall strength of the other side.” Dupuy, A Genius for War, 328.  The figure in parenthesis represents one side’s “Score” or combat efficiency as a percentage of the “Score” of the other side.  The Unionists (at 28.5% of the Confederates) had fallen a long way from the combat efficiency displayed at Antietam (61% of the defending Rebels), while the Secessionists (at 350% of the Federals) performed better than at Gettysburg (100% of the defending Northerners).  Beringer, Hattaway, Jones, and Still, Why The South Lost The Civil War, 472.   Dupuy displays an alternate “Score” excluding prisoners.  Dupuy, A Genius for War, 330-331.  I have not shown an alternate “Score” excluding prisoners because in the campaign of 1864 a prisoner amounted to a dead man due to Grant’s refusal to exchange prisoners.
[2] “This last column, ‘Score Effectiveness,’ adjusts the Score value to reflect approximately the known operational advantage which is conferred by defensive posture. “  Ibid., 328.  Because both sides attacked and at least partially engaged also in hasty defense during Grant’s fourth offensive at Petersburg, I have applied the operational advantage of 1.2 estimated by Dupuy for such situations.  Ibid.  Looking at Dupuy’s Figure C-1/Aggregated Statistics of Fifteen World War I Battles, Grant’s fourth offensive most resembles in its results 1915’s forty-five day long Champaign II battle, where with the benefit of machine guns and modern artillery the defending Germans attained a score effectiveness of 1.06 and the attacking French had a score effectiveness of 0.27.  Ibid., 330-331.  The fifteen day Aisne II (Nivelle offensive) where the defending Germans had a score effectiveness of 1.03 and the attacking French attained a score effectiveness of 0.27, may provide a better example because of a length similar to Grant’s fourth offensive and the tendency of score effectiveness to drop as battles drag on.  Ibid., 328-331.
[3] This column adjusts score effectiveness at Second Reams Station by 1.5 for the Prepared Defense of the Federals.  Ibid., 328.  It did not amount to a Fortified Defense which would have required an adjustment by 1.6 because the Unionists failed to improve the Reams Station earthworks significantly.  Id.  Even giving the Federals every benefit of the doubt and adjusting by 1.3 for a Hasty Defense, they attained a Score Effectiveness of 7.21.  Either way, the Unionists scored surprisingly well at Second Reams Station—just not very well in relation to the Confederates.  Whether the Federals attained a score effectiveness of 6.25 or 7.21, their score effectiveness resembles that of the victorious Germans in 1914’s battle of the Masurian Lakes.  Ibid., 330-331.  Unfortunately for the Northerners at Second Reams Station, the victorious Rebels attained a far higher score effectiveness of 34.0, approaching the 38.1 of the victorious Germans in 1915’s battle of Gorlice-Tarnow.  Ibid.  Dupuy’s chart does not list a battle where the defeated attained a score effectiveness above the 3.86 of the Germans in 1914’s Marne battle.  Ibid.  The score effectiveness of the armies for the entire fourth offensive gives a better idea of their capacities compared to WWI troops than the score effectiveness for the single day’s combat at Second Reams Station.  Ibid., 329-331.