I support the numbers I have found, backed by historians... the windmills are calling you...
They are not backed by anyone other than Bonekemper, and he uses his own numbers in preference to historians, but you have not read his book so you dont know what he passes of data.
As to Bonkempers cherry picked statistics. His argument and what he wants readers to believe is that the % of an armies losses is the prime indicator of a Generals abilitys. It counts only what is lost, not the cost in inflicting it, which is stand practice in measuring effeiecency. Cherry picking is to not count the total number of men under Frants command, that he can then use, but use some battles and others as campaigns, it follows that this method isa dilbberate math matical tactic to distort the data and supporta existing bias, because the outcome is comparerd only to other general single battles. Example in 64 he lumps many battles in as campaigns, so undercounts Grants number avaiable to use in many battles, resulting in grant haveing in Bonkempers book 1000s of 000s less troops to use than he did in reality and then compares that to others who dont get there data treated the same way. This method, of counting the totals present at each battle/campaign and adding them up, means instead of the PFD and replacemenmt strength Grant had in fact, and was the number he could use, is replaced by adding the strength at each battle up, this inflates the strength by massive orders of magnitude, and then compares actual calaties to this inflated numbers to geta lower end number, which is the whole reason he chose to do it that way, to confirm his pre existing bias by distorting the dat, Grant did not have and use nearlya million men in the East using his method.
Exasmple
CS achieves all its goals South mountain 1862, who however perfoms better?. Bonkemper would have you believe that the totals of CS losses are higher than US totals therfore the US was doing better.
Otoh is you look at what it cost to inflict as % of your own strength on the enemy the % it lost you get a totaly different perspective on effieceny and who was better.
US 28000 2,325 losses +10%-8%
CS 18000 2685 losses +13%-15%
Antietam
US 67,000 12,410 losses +15%-19%
CS 40000 10,316 losses +31%-26%
So for lee the campaign cost 13287. Inflicting 27371.
For Mc it cost 12641. Inflicting 13001.
Or Mc lost what it cost him to inflict, while Lee was inflicting 2 for every 1 he lost in doing so. Mac did not have 67000+28000=95000 as he had 87,164 PFD to fight both battles with, while lee did not have 58,000 but had 75,305 PFD both using the last PFD numbers in sept for each side.
Another example of Bonkemper wanting a reader to believe lee was worse than hooker, at Chancelrsville lee lost 21% to hookers 13%, so Hookers the better General according to Bonkemper as hooker army lost the lower number.
This goes into overdrive in 64 in the east as Bonkemper omits double counting grants numerical advatage.
Statistical issues by Bonkekemper
He has Grants losses at Belomont at 501, Grants casulaty return has 607, and for a skirmish at 2 days before as 85, total reported 692, an undercount of 191 by Bonekemper. Not a good start.
He continues, for Fort Donelson 2832, which is correct, ignores Fort Henry with its 47 ignores new Madrid with its 51 ( Grant was in overall command and Pope just follwed his order) a 98 undercount.
Shiloh he has 13047 while grants report has 13844, an undercount of 797.
Iuka he has 790
He misses out first Corinth with its 1000 casualties.
2nd Corinth/ Hatchie River he has 3,120 no isues.
He misses out Chicksaw bayou 1176 under count
he misses out Dover 110 undercount
he misses ouit Grand Gulf 80 undercount
he misses out port Gibson 861 undercount
he misses out Raymod 446 undercount
he misses out Jackson 286 under count
he misses out Champion hill 2447 undercount
he mises out Big Black river 276 undercount
Vicksburg Campaign he has a total 9,362 which is a an over count of 2000.
Chattanooga 5,814 no issues.
Return to Jackson he has 1122 no issues.
He goes East
Wilderness 17,666 18000
misses out Walthall junction 275 undercount
Misses out Swift creek and Fort clifton495 undercount
misses out chester station 285 undercount
misses out Cloyds Mouintain 688
misses out Cove mountain 150
Misses out yellow tavern 400
Spotsylvania Court House 18,399 no issues
misses out Ware bottom/hawlets farm 750
North Anna River, etc. 3,986 no issues
misses out Wilson warf 26
misses out haws shop 344
misses out tototemploy and Bethseda 821
misses out Matadequin creek 90
Cold Harbor 12,737no issues
misses out piedmont 875
Petersburg Assaults 11,386 no issues
misses out Trevalian station 1007
misses out Jeruslam plank road 2962
misses out sammaria bridge 350
Misses ouit reams station 375
misses out Petersburg II assault 10600
Misses out petersburg 60
mises out rutherford farm 242
Petersburg Siege/Campaign 42,000 no issues
Appomattox Campaig 10,780 no issues.
A combined undercount of 27000 for Grants losses, which is a 18% undercounts of grants casualties, if you add in the strength on hand for the engagements not used grants total use of US troops rises to over a million, resulting it a c50% undercount of grants manpower, so his entire diatribe is baseless math as he cherry picked the data to get the result he wanted, as he was not interested in understanding history, but promoting his bias.