"The Rise of Rail-Power in War and Conquest, 1833-1914" - re AACW

[Taillebois]

I just came across and have started reading the above book by the unfortunately named Edwin A Pratt.

It has quite a lot on the Civil War being the first major trial of railways in war. The book is free on Project Gutenberg so you can get it for almost any type of electronic reader.

I posted here just because traffic seems a bit slow so feel free to move it.

Long term players may well know of the book anyway.

[khbynum]

Interesting, thank you, Sir. I'm a long-time amateur student of the war but had never seen this book. I also had to go on-line to find out why Pratt is an unfortunate name. It's not a part of American slang.

[John Schilling]

I hadn't seen this one before, and it is an excellent source. Thank you for bringing it up.

Chapter III is of particular relevance to AACW, given the persistent controversy over various raiding tactics. In AACW, it pretty much always takes a full month to repair a damaged railroad - two weeks to deploy a suitable "repair crew", and two more weeks to get the job done. In fact, as Pratt points out, railroad repair was only slightly slower than railroad destruction, with damaged tracks and even burnt bridges being put back in service in no more than a few days. This discrepancy makes raiding a much more damaging tactic in the game than in reality.

How might this be changed for AACW2?

I can see two possibilities. First, we could keep railroad destruction as it is, but make repair a fairly rapid and automatic proposition. Tracks are marked as damaged rather than destroyed, and presuming sufficient War Supply is available will be automatically repaired in one turn. And for the duration of that turn, the effect of the damage is limited to perhaps a 2-day delay in rail movement along the track, plus perhaps a prohibition on that rail segment being used during the first "push" of the supply segment. If there is enough rail capacity to support two or three supply pushes in a turn, assume that this includes not just locomotives and rolling stock but also repair crews and stockpiled supplies as described by Pratt; the quartermasters will have things running smoothly by the end of the fortnight.

Second, we could assume that such minor interruptions are not even worth bothering about in a game of this scale, but that we do need to concern ourselves with the possibility of wholesale destruction of railroads beyond repair. Which did on occasion occur, but not just because a regiment of cavalry decided to pull up some track. In this case, we keep the railroad-repair process as is, but make railroad destruction a much more difficult process. Much harder than depot-destruction, and probably about as hard as railroad-repair. A brigade wholly devoted to the task for a full two-week turn would be about right. And a cavalry regiment that tries such a thing, will still be at it when enemy reinforcements arrive in strength.

Of the two, my preference would be for the second. It would in principle be possible to implement both in parallel, tracking railroads "damaged" and "destroyed" separately, but that would be complicated and confusing.

Other thoughts?

[caranorn]

Concerning your first possibility you did not consider that the damaging unit can only damage rail in one region under current rules (and never 100% certainty iirc). So in a 15 day period a % to damage one section of rail. In reality, a raiding unit would tear up small sections of track, pull down water towers, burn coal/wood stocks, break up bridges all along it's movement. So either we'd have to make damage railroad an action that will be executed in every enemy RR region crossed by the unit or (would actually require a separate order, damage_friendly_rail) the friendly RR region it starts it's movement in. In that case automatic repair of friendly RR (obviously not enemy) might be relatively realistic. Alternatively keep the current system damaging first region's rail regardless of control and then have slower repair (why not auto in friendly, just not during the turn)...

Essentially AACW-I has slow damage and slow repair. Fast damage and fast repair would be more realistic, but slow damage and fast repair would be unbalanced.

[John Schilling]

Concerning your first possibility you did not consider that the damaging unit can only damage rail in one region under current rules (and never 100% certainty iirc). So in a 15 day period a % to damage one section of rail. In reality, a raiding unit would tear up small sections of track, pull down water towers, burn coal/wood stocks, break up bridges all along it's movement. So either we'd have to make damage railroad an action that will be executed in every enemy RR region crossed by the unit or (would actually require a separate order, damage_friendly_rail) the friendly RR region it starts it's movement in. In that case automatic repair of friendly RR (obviously not enemy) might be relatively realistic. Alternatively keep the current system damaging first region's rail regardless of control and then have slower repair (why not auto in friendly, just not during the turn)...

Essentially AACW-I has slow damage and slow repair. Fast damage and fast repair would be more realistic, but slow damage and fast repair would be unbalanced.

How do you get "slow damage" in AACW-I? Unless I am missing something, railroad destruction under the current rules is literally instantaneous. Give the order, even to just a single regiment, and an entire region's rail network is devastated at the beginning of that regiment's first day of movement. The unit's movement is not slowed in the least, except insofar as the granularity of the game forces it to schedule its passage through the target region to correspond with a turn break. There is no possibility that an enemy reaction force can catch them in the act and defeat them in battle, or even force them to leave the railroad unmolested.

Your point that the current game mechanism prevents the same regiment from destroying railroads in multiple regions is true, and somewhat interesting, but I don't think it really matters. It is not practical for even a corps to wholly destroy the rail network of an entire AACW region in less than a fortnight, so I think it is reasonable to abstract whatever damage a force might do in a single turn, as having been concentrated in whatever region the force began its turn. At most, this would be a small penalty to overall realism, for the sake of ease of play.

What we have now is a much greater offense against realism, and I don't see any compensating enhancement in playability.

If railroad destruction is to be instantaneous, particularly when implemented by regiment- or brigade-sized forces, it must not be severe. Repair should be handled automatically, just as supply movement is, and the effect should not be more than a few day's delay. This would more realistically depict the effects of raiding-style railroad destruction, and it would make for easier, not harder, gameplay.

If railroad destruction is intended to reflect scorched-earth style wholesale destruction of infrastructure, such that a month or more will be required to restore service, then it must not be instantaneous - and it must not be available at all to the smallest forces, such as lone cavalry regiments. And we've already got two perfectly good mechanisms for that; depot destruction, with its five-day waiting period, and the railroad repair mechanism itself, which could easily be reversed to support lengthy railroad-destruction operations.

Right now, we have instantaneous destruction and slow repair. Of all the possible combinations, that one I think is the worst, and reading Pratt only reinforces my belief.

[Ace]

Your suggestions are good. For the sake of simplicity, I would not differentiate road damage and road destruction. But the auto repair feature would be nice to reduce the micromanagement. I would implement few days delay for the destruction, and I would keep one turn delay, not few days delay, for the repair. You are suggesting raiding had no influence over Union supply system in the war. In the West, it had all the influence, and by keeping a turn delay for the repair it would be simulated.

[orca]

How do you get "slow damage" in AACW-I? Unless I am missing something, railroad destruction under the current rules is literally instantaneous. Give the order, even to just a single regiment, and an entire region's rail network is devastated at the beginning of that regiment's first day of movement.

That's not how it works. You give the order, it takes some number of days, and you have a success rate equal to the power levelof the stack assigned to destroy the rail.

I never bother to destroy rail unless I have 4 regiments of cavalry and so nearly 100% chance of sucess. For a single regiment you're just too unlikley to suceed and tying the regiment in place isn't a worthwhile risk IMHO.

So slow destruction, slow repair.

[Taillebois]

Chapter 18 is for those who find WW1 (Schlieffen Plan) and WW2 (1940 and 1944) Ardennes interesting.

"In 1908 the Prussian Government suddenly decided to double the line ... to St Vith, notwithstanding that there was no apparent justification for such a procedure."

"..the Germans succeeded in persuading the Belgian Government, not only to agree to the Weismas-Malmedy branch being continued to Stavelot, but themselves built the greater part of this connecting link"

"In helping provide this connection, Belgium, as subsequent events were to show, was in the position akin to that of a man forced to dig the grave in which he is to be buried after being shot"

[John Schilling]

That's not how it works. You give the order, it takes some number of days, and you have a success rate equal to the power level of the stack assigned to destroy the rail.

I never bother to destroy rail unless I have 4 regiments of cavalry and so nearly 100% chance of sucess. For a single regiment you're just too unlikley to suceed and tying the regiment in place isn't a worthwhile risk IMHO.

So slow destruction, slow repair.

This does not appear to be the case. I had forgotten that the latest revisions include a probability of failure for the smallest forces attempting to destroy railroads, and that is at least a step in the right direction. But where destruction does occur, I have never seen any delay, so as for "takes some number of days" and "tying the regiment in place", no.

I have just finished verifying this with one of the tutorials, using I believe the 1.17a revision. Ran several variations, ordering forces of various sizes to destroy railroads and then conduct a full turn's movement. Results:

A. There is no effect on movement plotting. Order fifteen days' projected movement, select the "Destroy RR" option, and the projected movement remains unchanged.

B. As noted, railroad destruction is not automatic for forces smaller than a 100-point brigade. The success rate is approximately equal to the power level of the stack, as a percentage, whether the force is moving or not.

C. Railroad destruction, when it occurs, appears to be instantaneous. The destroyed-RR symbol appears on the map at the very start of movement, or not at all.

D. The moving force, whether it succeeded in destroying the railroad or not, completed its full fifteen days of plotted movement.

E. In one test case, another friendly unit ordered to move through a region with plotted RR destruction on day 2, was stopped in its tracks - or rather lack thereof.

F. When a force of cavalry in clear terrain was ordered to destroy RR and move out, entering the next region on day 2, RR destruction occurred (or not, if the force was small and unlucky) on day one, and the cavalry entered the next region on schedule on day 2.


If the intent was to impose a delay on railroad destruction in one of the latest revisions, then this was well-meant but does not appear to have been successfully implemented. And if the intent is that brigade-sized forces should be able to destroy railroads rather than simply disrupt them for a few days, then to match the historic results reported by Pratt and others the appropriate period for destruction should be a full turn or thereabouts. Which would seem easy enough to implement - simply have the "Destroy RR" button deselect any plotted movement, and vice versa.

What we have now does in fact seem to be instantaneous destruction and slow repair. Which ahistorically exaggerates the effect of raiding forces, something I do not think this game needs.

[Ace]

E. In one test case, another friendly unit ordered to move through a region with plotted RR destruction on day 2, was stopped in its tracks - or rather lack thereof.

That is not true. Any damage to the rail affects only movement in the next turn. You can freely travel over damaged rails if they are damaged during a turn. That is an abstraction I can live with.

[UnionBlue]

I have not read the book but I figured this is a good spot to post my beef with the AACW rail system. What I am about to say may sound nit picky but for the sake of accuracy and argument I feel it is constructive.

First off, the game does not take into account for the different rail gauges present in the country. Especially in the south (and to a lesser extent the north) the tracks were different widths according (more or less) to each state. This meant that when a train switched lines it would have to stop and all its cargo would have to switch trains. This could take all day depending on how much and what type of cargo it was carrying.

Second The only way you lose rail capacity is by using it and by accidents. In reality trains would often break down and towards the end of the war the south had trouble finding engineers and the parts necessary to build them (trains). In short to be more accurate the game should have added more triggers to lose rail capacity.

In all the game could have made the transport system more accurate.

[khbynum]

There is already attrition of rail transport built into the game and for me that abstraction is quite enough. What I would like to see is a chance to build new rail links, perhaps coupled with an option to NOT build ironclads. In fact the CSA tore up some seldom-used railroads to provide armor for ironclads. Perhaps allow the CSA player to forgo some or all ironclads in exchange for lowered attrition and/or the ability to build new railroads, like one between Montgomery and Selma.

Edit: On second thought, give me the option not to build ironclads at all. I think CSA resources in materials and artillery could have been much better spent. The game forces you to build ironclads, when in fact it was a deliberate and controversial path they chose to take. Ah, but that's a whole 'nother topic.