You wrote this elsewhere.Fiver said:Yes, McClellan did claim the numbers I have listed were correct Confederate numbers. You can find them in most any source except for those dedicated to proclaiming George McClellan as an unsung genius. My main source, which I have already mentioned is Edwin Fishel's The Secret War For the Union. Fishel's sources included the operational files of the Bureau of Military Intelligence, more of the BMI's reports in Hooker's papers, and 1000 pages of Pinkerton reports in the McClellan papers. Pinkerton pretty consistently overestimated Confederate numbers, while McClellan pretty consistently treated Pinkerton's numbers as underestimates.
If McClellan didn't believe his numbers then he was lying to his superiors and his lies worked against him, because those over-inflated claims of Confederate numbers led directly to the large number of troops Lincoln kept in the Washington defenses instead of sending to McClellan.
Fishel kept making the same mistake. He'd take an estimate in one strength category (aggregate present or sometimes present and absent) and compare to a lower strength category.
In fact, Pinkerton was uncannily accurate until June. For example, in March he counted rebel strength exactly by regiment:
This was accompanied by a strength estimate of 60,000 at Manassas reached by penetrating the commissary and similar estimates elsewhere averaging about 1,000 men per regiment. Obviously these are estimates of effective strength, or PFD, and at best are estimates of aggregate present. The int was dated 25th December 1861 (the agent used Christmas to get details out of a Commissary clerk), and the return for that period shows it it was accurate.
Johnston in his returns was lowballing his strength (as Newton points out) and in February-March huge numbers had been furloughed causing a temporary drop in strength of 20,000 present.
Pinkerton had accurately reported exactly the number of regiments present, and had also reported numbers from Commissary rolls, which were probably grand aggregates, and McClellan could certainly read that.
On the Peninsula McClellan three times gave estimate of enemy strength at Yorktown, and he was pretty much right:
Pinkerton gives a fairly accurate report in early May (see Newton link above). On 26th June the general estimates of Pinkerton are:
150,000 at Richmond
30,000 under Jackson
possibly 20,000 under Beauregard transferred east.
Now, here Pinkerton is counting high by 36 regiments at Richmond. In the strength category he was counting he should have said about 120,000 in Richmond, but the pertinent fact is that Lee really did outnumber McClellan. To be more precise his force at Richmond was of equal strength to McClellan's force, and Jackson was the edge.
The exaggerated reports from the Valley come from Banks, who was estimating Jackson at 30-40,000. The wrong intelligence that Beauregard was Richmond with his army originated from Washington and McClellan started openly he didn't believe. Washington had completely lost track of Beauregard when they moved from Corinth to Chattanooga and read in the rebel papers that Beauregard had been summoned to Richmond, which he was but to explain himself to Jeff Davis, not with his army.
At Antietam the question is simply what are the "46 regiments not counted"? These are a mix of miscounts and GW Smith's division of 4 brigades left at Richmond. Aside from this line the estimate is uncanny.
McClellan in his communications with Halleck etc. said 100,000 and 120,000. Here McClellan is trying to convince Halleck that there is not a whole other rebel army on the Rappahanock. The latter is correct for McClellan's interconversion between PFD and Present, and thus is the same figure viewed in two different columns.
There is good reason to believe McClellan doubted the "46 regiments". However, what we're seeing is what was sent for external consumption.
Fishel can pour as much scorn as he likes, but Pinkerton was much more accurate than credited with. Sadly by accepting fundamental tenants of the Lost Cause he misjudged the estimates badly.