Seventeenth Century Armies - counting the troops

I'm busy with the second part of the Swedish Army post at the moment.  I had to take a few days out for recovery from having a kidney stone lasered but I'm back at the keyboard again now.  The enforced rest gave me time to think about one of the things I'm finding quite difficult.  Which is getting hard information on Swedish unit sizes, well unit sizes for any nationality really.  That left me thinking about how units were assembled for use in combat and their actual sizes.

I have already stated that the Regiment (or it's local equivalent) was an administrative body, pure and simple.  The combat formations were a different beast as they were formed from the regimental manpower pool but would have been merged or split as required to make up battlefield manoeuvre formations.  Companies or troops would have been kept together as these were bodies used to fighting as a group so I expect there to be some variation in unit sizes but not to a huge extent.  This is where the difficulties arise.  

The vast majority of the manpower figures I have seen are from pay returns or casualty returns.  These were drawn up by the administrative formation; the regiment not the combat formation so are not automatically an accurate reflection of the numbers in the combat units.  On top of which as mentioned before the practice of drawing pay for dead deserted or entirely fictional soldiers means the returns are not as accurate as we might hope.  (Yes I know you're dead Smith, now just stand there and be counted by the nice pay master there's a good lad).  I do wonder if that was worse or better in European armies which employed a large number of mercenaries as the contractor wants to increase his profit, but the engager wants to know they are getting value for money.

Either way we do know the theoretical size favoured under the various, and different, tactical doctrines from looking at the writing of the military authors of the day.  We can test that by the simple means of dividing total head counts by the number of tactical formations to see how close these come to the theoretical head counts.  And that dear reader is as good as it gets, unless you have a TARDIS about your person!