The Peninsula Pinkerton and Mac

[MikeyB]

For those who have studied McClellan in detail, do you think Mac really believed the Pinkerton reports? Or was he just using the inflated estimates as excuse/cover to justify his actions?

If he was a true believer, then his cries for reinforcement and warnings how if the army is destroyed he will die with it but it won't be his fault are a little more sympathetic.

Conversely, love him or hate him, he was an intelligent man. And he clearly understood logistics. How did it not dawn on him that the estimates just didn't make sense?

 

[67th Tigers]

Pinkerton's reports were just one input into the intelligence cell. They were generally more accurate than is commonly believed. For example, at Yorktown, the numbers McClellan gave tracked the true numbers very closely.

By the time of the Seven Days, Confederate counter-intelligence had eliminated many of Pinkerton's best agents. A lot was placed on John C. Babcock's estimated order of battle, which started to accumulate errors. By the Seven Days, Babcock was estimating an additional 36 infantry regiments over those actually present. If you remove these and run the numbers, they're actually quite reasonable and ca. 125,000 rebels present (close to the true figure).

It is important to note that, by this time, Lee really did outnumber McClellan. It was just by a smaller margin than believed.

 

[Saphroneth]

If one tracks actual Confederate companies present in the Seven Days (companies being 1/10 of a regiment), the total that this produces is 204.7 regiments of all arms in the field army, plus 11.1 in the Richmond militia and garrison (215.8 in total). Another 26.1 arrived in the next month, demonstrating that most of the "Babcock extra" was plausible even if it hadn't actually arrived yet and so it couldn't be thrown out as impossible.


McClellan at this time had about 170 regiments of all arms in his field army, disposing of 117,226 Aggregate Present or 105,825 Present For Duty. Assuming roughly equivalent average company sizes, this comes to:

148,900 Aggregate Present with correct intel (of which the garrison force is about 8,000)
134,400 PFD with correct intel (of which the garrison force is about 7,000)

166,200 AP with Babcock ORBAT
150,000 PFD with Babcock ORBAT

Basically McClellan is outnumbered 5:4 in regiments and companies, instead of 3:2. It still means he's outnumbered.



Other ways of getting at Confederate strength come out much the same way. e.g.


Lee was complaining that he didn't have enough officers for his regiments after the Seven Days, at a time when he had already sent Jackson (but not AP Hill) off back to the north.
On July 20, Lee had 4,333 officers in his army (not counting Jackson) which was up 12 regiments on the Seven Days.


As it happens, McClellan's report for the 10th of July (which includes Fort Monroe and is up nine regiments on Seven Days strength) has a headline number of 4,327 officers PFD. Looking at all McClellan's infantry units gives a range of 19-25.3 enlisted men per officer, with the average being 23 (so officer number * 24 is a rough estimate of PFD strength for an infantry formation).

The same figure for the whole of McClellan's force, sans Fort Monroe, is about 22.3 (so officer number * 23.3 for whole-force PFD) and including Fort Monroe makes it 22 men per officer (so officer number * 23).

Based on this estimate our first-order calculation for Lee's strength on July 20 would be on the order of 99,700-104,000 PFD.

Doing the same calculation with Lee's infantry units using the July 20 strengths means he comes out as about 13.5 to 18.5 officers per man (averaging 15.2), which is far lower (and definitely supports the idea he's not using the same measure of PFD).
Since the AotP corps in March pre-Peninsular campaign were in the range 20.5-23 officers per man (average 22.3) it looks like the reason for the discrepancy is not that McClellan had ended up with a disproportionate number of non-PFD officers; it must be the case that Lee's officer count is proportionate (or low), rather than high.

This means we can estimate Seven Days strength as:
Same number of men per officer as McClellan, for Lee's July 20 strength
+ Jackson and Ewell from Schulte
+ 7 days casualties
- reinforcement regiments (which had arrived by July 20 but which were not at the Seven Days)
- recovered Confederate casualties

4333 officers at Richmond July 20, x 23 by comparison with McClellan = 99,700 PFD
Jackson and Ewell August 1 (7500 + 5700) / 0.8= 16,500 PFD (this is from the Schulte ORBAT and assumes no cavalry went north with Jackson)
Seven Days casualties 20,000 (assume PFD)
12 reinforcement regiments (9152 * 12/13.4) = 8,200 PFD <- this is the reinforcements Lee got post-Seven-Days and by July 20, calculated by comparison with the 13.4-regiment Pennsylvania Reserve division on June 20

Estimate of recovered casualties based on Union...
5th Corps on June 20 exc/ of attached cavalry is (3+493+10940+268+5477+85+2023) (PFD main 5th Corps) + 382+9132 (McCall) for 28803 total
Suffers 56+939 killed and 101 + 2700 captured (3796 unrecoverable casualties) n.b. cav casualties ca. 40 overall, do not affect conclusion
Suffers 194+3611 wounded (3805 recoverable casualties)
Strength July 10 is 20203 + 817 for 21020 total
If no wounded recovered, would be 21202

So recovered casualties 10 days after end of fighting are negligible and/or compensated for by sickness
Will assume 2,000 Confederate casualties recovered by July 20, to argue a fortiori
Thus:
99700 + 16500 + 20000 - 8200 - 2000 =
126,000 PFD

 

[OpnCoronet]

If any of the calculations of Mac's apologists on this thread, are anywhere near correct, it indicates to me that McClellan's entire Peninsula Campaign was misbegotten, from the very first.

 

[Saphroneth]

If any of the calculations of Mac's apologists on this thread, are anywhere near correct, it indicates to me that McClellan's entire Peninsula Campaign was misbegotten, from the very first.

I don't think it does.

The original plan was Urbana, not Peninuslar. But when the Peninsular plan was adopted (against McClellan's preferences - he rated Urbana first, Peninsular second, and overland third) it was with the understanding that McClellan's offensive force would consist of four full corps, plus the forces at Fort Monroe, and that there would be naval support to overcome Fort Monroe in short order by bombardment.

Naval support was not forthcoming in that manner, contributing to the delay at Fort Monroe. There was terrible weather which hampered operations and added delay (giving the Confederates time to concentrate forces at Richmond). McClellan was stripped of four divisions and only ever got two of them back. And, despite those factors, the Peninsular Campaign came very close to working - it had artillery moving up onto the heights over Richmond when the Seven Days Battles pushed McClellan away from Richmond, in the largest attack ever launched by the Confederates.

And even after that, the cancellation of the Peninsular Campaign was based on the idea that it was better to combine with Pope's force than to wait the extra few months it would take to be able to send 50,000 reinforcements to McClellan and operate against Richmond from near Petersburg.
(This idea, it scarce needs to be said, was totally incorrect.)



The Peninsular Campaign nearly worked, and any alternative fought under the same parsimonious conditions against the same enemy strength would have had similar problems - as well as a few the Peninsular Campaign avoided, such as how to supply the army close to Richmond or how to get over the Rappahannock and into good supply.

As it stands, the Peninsular Campaign merely inflicted more casualties on the Confederates than were suffered by the Union and opened up the York and James rivers to Confederate supply. This is a good outcome and it's hard to name another campaign against Lee which achieved more under similarly trying conditions; for one thing, it'd be impossible to achieve an Overland level of superior numbers unless you gave McClellan everything he wanted for his original strategy memo...

 

[Belfoured]

I don't think it does.

The original plan was Urbana, not Peninuslar. But when the Peninsular plan was adopted (against McClellan's preferences - he rated Urbana first, Peninsular second, and overland third) it was with the understanding that McClellan's offensive force would consist of four full corps, plus the forces at Fort Monroe, and that there would be naval support to overcome Fort Monroe in short order by bombardment.

Naval support was not forthcoming in that manner, contributing to the delay at Fort Monroe. There was terrible weather which hampered operations and added delay (giving the Confederates time to concentrate forces at Richmond). McClellan was stripped of four divisions and only ever got two of them back. And, despite those factors, the Peninsular Campaign came very close to working - it had artillery moving up onto the heights over Richmond when the Seven Days Battles pushed McClellan away from Richmond, in the largest attack ever launched by the Confederates.

And even after that, the cancellation of the Peninsular Campaign was based on the idea that it was better to combine with Pope's force than to wait the extra few months it would take to be able to send 50,000 reinforcements to McClellan and operate against Richmond from near Petersburg.
(This idea, it scarce needs to be said, was totally incorrect.)



The Peninsular Campaign nearly worked, and any alternative fought under the same parsimonious conditions against the same enemy strength would have had similar problems - as well as a few the Peninsular Campaign avoided, such as how to supply the army close to Richmond or how to get over the Rappahannock and into good supply.

As it stands, the Peninsular Campaign merely inflicted more casualties on the Confederates than were suffered by the Union and opened up the York and James rivers to Confederate supply. This is a good outcome and it's hard to name another campaign against Lee which achieved more under similarly trying conditions; for one thing, it'd be impossible to achieve an Overland level of superior numbers unless you gave McClellan everything he wanted for his original strategy memo...

"As it stands, the Peninsular Campaign merely inflicted more casualties on the Confederates than were suffered by the Union ..."

I assume your calculations include Leon Tenney's substantial re-figuring of losses during the Seven Days in his 2012 book, contrary to the long-accepted "conventional wisdom"?

 

[Saphroneth]

"As it stands, the Peninsular Campaign merely inflicted more casualties on the Confederates than were suffered by the Union ..."

I assume your calculations include Leon Tenney's substantial re-figuring of losses during the Seven Days in his 2012 book, contrary to the long-accepted "conventional wisdom"?

As it happens, no, they're based on the Livermore numbers (which give 23,119 Union to 29,298 Confederate by my understanding) and by the way that the individual major battles are:

Williamsburg 600 more Union net (NPS Union, Sears Confederate)
Seven Pines 700 more Confederate net (Sears both)
Hanover Court House sees more CS captured than Union total casualties by several hundred
And the Seven Days give greater Confederate casualties than Union ones (with a significantly higher Confederate casualty count at Beaver Dam Creek, Gaines Mill, Goldings Farm and Malvern Hill which Glendale and Savage's Station do not redress).

Is there anything specific in Tenney 2012 you're thinking of?

 

[Belfoured]

As it happens, no, they're based on the Livermore numbers (which give 23,119 Union to 29,298 Confederate by my understanding) and by the way that the individual major battles are:

Williamsburg 600 more Union net (NPS Union, Sears Confederate)
Seven Pines 700 more Confederate net (Sears both)
Hanover Court House sees more CS captured than Union total casualties by several hundred
And the Seven Days give greater Confederate casualties than Union ones (with a significantly higher Confederate casualty count at Beaver Dam Creek, Gaines Mill, Goldings Farm and Malvern Hill which Glendale and Savage's Station do not redress).

Is there anything specific in Tenney 2012 you're thinking of?

Yes there is. Tenney has looked into this (as I'm sure you know, his master's thesis at GMU has been cited by Harsh and others). In his 2012 self-published book (which is an expansion of his thesis), he ended up recalculating the Seven Days and virtually eliminated the c.4K difference that favored McClellan (especially significant because throughout Lee was the attacker).

 

[Saphroneth]

I should note that the Gaines Mill casualties on Wikipedia are a bit odd, because the info box says 7,337 casualties for the Union but the references linked all say 6,837 (Burton) or 6,800 (Salmon). Similarly, the info box says 7,993 casualties for the Confederates but the references linked say 8,700 Confederate.
So someone's just added exactly 500 in round numbers to Union Gaines Mill casualties.

The Union casualty return in the ORs for the Seven Days as a whole says 15,849 Union broken down by individual regiment, but the Wikipedia infobox says 17,849 without citation.

Similarly someone seems to have trimmed a round 1,000 from the Confederate total in Sears to put in the infobox.

 

[Saphroneth]

Yes there is. Tenney has looked into this (as I'm sure you know, his master's thesis at GMU has been cited by Harsh and others). In his 2012 self-published book (which is an expansion of his thesis), he ended up recalculating the Seven Days and virtually eliminated the c.4K difference that favored McClellan (especially significant because throughout Lee was the attacker).

That's interesting. Does he go into the specifics about which of the Union casualty returns in the ORs (by regiment) are incorrect? Or is it that the Confederates suffered fewer casualties than generally believed?

 

[Dead Parrott]

If any of the calculations of Mac's apologists on this thread, are anywhere near correct, it indicates to me that McClellan's entire Peninsula Campaign was misbegotten, from the very first.

I've seen this type of data abuse too many times already.

Not worth the repeated electrons. On to other threads.

 

[Saphroneth]

I've seen this type of data abuse too many times already.

It seems a bit poor form to call it "data abuse" when the first question we should ask about an erroneous estimate is "why was it an error".

People don't just make up numbers out of thin air. They have methodology. And if we apply their methodology with the actual Confederate regimental positions (which we know, unlike them, because we are viewing it as history) then we can get what their numbers should have been if their methodology was applied correctly.

The number of regiments and companies at the Seven Days (~215 versus ~170) is either accurate, or inaccurate. If it is inaccurate that is because the 215 number includes units which were not actually there, and we could correct that.

It is either the case that Union and Confederate regiments were of similar sizes, or it is not. If it is not, there should be a reason why and evidence that this is the case.


This is understanding history. You're quite at liberty to prefer not to do the analysis, but it's hardly worthy to criticize those who do go into that detail.

 

[Dead Parrott]

This is understanding history. You're quite at liberty to prefer not to do the analysis, but it's hardly worthy to criticize those who do go into that detail.

And Mac's apologists are quite at liberty to pretend their 'analysis' isn't number shuffling to blame anyone other than Mac.

There are threads and threads and threads of it throughout this (actually very good) ACW site.

The analysis, criticisms and refutations have already all been done … countless times.

But this is a fan discussion site, not a courtroom with rules, so statements /misstatements and application/misapplication of data can (and apparently will) continue to be repeated.

Enjoy.

 

[67th Tigers]

Yes there is. Tenney has looked into this (as I'm sure you know, his master's thesis at GMU has been cited by Harsh and others). In his 2012 self-published book (which is an expansion of his thesis), he ended up recalculating the Seven Days and virtually eliminated the c.4K difference that favored McClellan (especially significant because throughout Lee was the attacker).

Apparently double counts. We know that many of the wounded from 25th-27th June were captured when Sumner abandoned the hospitals, but they should only be counted once.

 

[Saphroneth]

And Mac's apologists are quite at liberty to pretend their 'analysis' isn't number shuffling to blame anyone other than Mac.

Quite frankly, if a historical analysis of actual regimental positions indicates that there were more regiments at Richmond (and thus more strength) than the common picture that is presented, then that is actually quite a good defence of McClellan, in so far as he is criticized for believing there to be stronger Confederate forces than were actually the case. If someone is blamed for overestimating by 100%, and it turns out that actual regimental data suggests that someone else was overestimating by 30%, that makes the original criticism either false or at the very least questionable.

 

[Saphroneth]

For what it's worth, I just checked the Confederate casualty reports on a brigade by brigade basis, for the Seven Days (the one which is mostly broken out into individual regimental casualties), on 973-984 of the relevant ORs, and the results are that

Oak Grove​	Beaver Dam Creek/ Mechanicsville​	Garnetts and Goldings​	Gaines Mill​	Savage Station/ Peach Orchard​	White Oak Swamp​	Glendale​	Malvern​	Total where not given by battle (i.e. Seven Days casualties consolidated)​
441​	812​	426​	4126​	391​	13​	1561​	4214​	7991​

Grand total 19975.

Given that the Union reports (which do have a bottom line) give their own grand total as 15,849, then either this indicates over 4,000 more Confederate casualties than Union or there are significant discrepancies in these reports - which should be possible to identify.


For the Antietam numbers, for example, we had the actual counts of prisoners, actual counts of burials, and a reason why those numbers might be different (i.e. severe disruption to the Confederate army) along with significant units with no reported casualties and thus a place where the numbers could be in the data.

I'd be interested to hear how the same could be true here.

 

[Belfoured]

That's interesting. Does he go into the specifics about which of the Union casualty returns in the ORs (by regiment) are incorrect? Or is it that the Confederates suffered fewer casualties than generally believed?

No - he gets McClellan to c. 20,000 because he makes the case that the Federals were undercounting during that week while they were in the process of retreating. IIRC, much of it is understated MIA. The book is at home so I'll have to check before laying out expressly how he got there.

 

[67th Tigers]

No - he gets McClellan to c. 20,000 because he makes the case that the Federals were undercounting during that week while they were in the process of retreating. IIRC, much of it is understated MIA. The book is at home so I'll have to check before laying out expressly how he got there.

The OR contains two sets of casualty figures, both being 15,xxx.

McClellan's report of August '63 gives 15,249 (1,582 KIA, 7,709 WIA and 5,928 MIA/Captured) suffered 26th June - 1st July (i.e. excluding casualties on the 25th June and 2nd July)

The compiled returns give 15,849 (1,734 KIA, 8,062 WIA and 6,041 WIA/Captured) suffered 25th June - 2nd July.

If memory serves, Tenney took these figures, and added those captured when Sumner abandoned the hospital. However, those men were already counted. On 3rd July, the rebel medical director estimated they had 4,900 captured Federal wounded, including 3,000 taken at Savage's. They were being shipped to Richmond so fast that the next day, half of those estimated to be at Gaine's and Savage's had gone to Richmond.

Those captured at Savage etc. were overwhelmingly injured by gunshot, and are certainly already counted.

On 16th July, Winder reported 7,847 prisoners in Richmond and environs. These include the Seven Days captures, but also some of the captures from Williamsburg, Seven Pines, Jackson's early valley operations etc. who had not yet been exchanged. This doesn't jibe with the idea of 5,000 unreported extra PW's. I generally find the numbers reported by the Federals to be consistent with the number of prisoners the rebels took, and inclusive of the captured hospital etc.

Tenney's research was into the rebel records, and I hope he has some good evidence for his claim of increased Federal casualties, which he has added post-facto, and isn't in his original MA thesis. In all probability, it's a double count.

 

[Saphroneth]

McClellan's report of August '63 gives 15,249 (1,582 KIA, 7,709 WIA and 5,928 MIA/Captured) suffered 26th June - 1st July (i.e. excluding casualties on the 25th June and 2nd July)

The compiled returns give 15,849 (1,734 KIA, 8,062 WIA and 6,041 WIA/Captured) suffered 25th June - 2nd July.

And, indeed, the Oak Grove casualties are on the order of 600 (626).

No - he gets McClellan to c. 20,000 because he makes the case that the Federals were undercounting during that week while they were in the process of retreating. IIRC, much of it is understated MIA. The book is at home so I'll have to check before laying out expressly how he got there.

This is basically the same argument as the one that was used to argue that Confederate casualties at Antietam were an undercount (that is, that the number of casualties suffered was undercounted because of disrupted reporting) but in the case of the Maryland Campaign casualties we have very solid evidence that under reporting took place:


- We have many regiments which gave no loss report for Antietam/Maryland, even in cases where the regimental colonel was killed, so we know for a fact that there were units which were under reporting casualties.
- We have Federal counts of prisoners of war (both captured wounded and captured unwounded) which are greater in total than the number of MIA in the Confederate reports (by more than 3,000), even though there had been a clean-out of POWs in a big exchange shortly before the Maryland Campaign.
- Most tellingly, we have Confederate graves at Antietam itself which are more numerous than the KIA reports for the Confederates, by about 800 on the field itself irrespective of anyone who died of wounds.


This makes a strong case for there being under-reporting of casualties - there's over 5,000 prisoners and 2,468 corpses as against the report total of 2,292 and 1,674 respectively, and we can see that at least 10% of Confederate regimental organizations filed no return of casualties - even in otherwise hard-hit units.


In order to establish under-reporting of MIA, we would need to confirm the scale of exchanges which took place before the Seven Days (to get at the extent to which the prisoner number had been "reset") or otherwise have a before-after comparison of how many prisoners were recorded.

 

[67th Tigers]

If I take the 20th June return and compare to the 10th July return, PFD has decreased by 15,791. That's fairly consistent.

An argument may be made that 4-7 July, eight regiments have joined. Inspection of the grand aggregate shows these two brigades are added to Casey's/Peck's division (Ferry's brigade) and Baldy Smith's division (Kimball's brigade). Those two divisions show 3,176 more PFD than would be expected by taking the difference between the 20th June state minus casualties, and the 10th July state.

To make the numbers consistent, without increasing the casualty count, about 3,000 men would have to have gone sick in the 8 or so days between 2nd and 10th July. This seems perfectly reasonable, especially with Keyes arguing that there was a lot of sickness at Harrisons on the same day (which he overstated), and McClellan issuing a general order the next day to prevent soldiers leaving without proper medical certificates.

I see no evidence of underreported casualties here.