Gettysburg - cavalry speed & fatigue

After Saturday’s scenario, a group of us were chatting on T/S about the treatment of cavalry in the stock game. Jeff had played the splendidly-named John Sappington Marmaduke, in command of Price’s Confederate cavalry, and been rather frustrated at his inability to both move and fight. When producing the scenario, I had increased the stock ‘callisthenics’ ratings for the Rebel cavalry up to the maximum, in an attempt to reduce fatigue and allow them to function properly as mounted infantry. In the event, this did not prove to be sufficient. As Jeff noted, if they move at the walk, they are little (if any) faster than infantry, and if they move at double-time, then they still get fatigued very quickly.

Kevin was hopeful that we can mod the necessary changes, so here are my initial thoughts on where we might go. I think there are 2 issues here: speed & fatigue:

Speed

The 19th C kriegsspiel manuals suggest that the game is correct to make cavalry at the walk no faster than infantry. These indicate march rates of 100 paces per minute for both infantry and cavalry (of all types). But that doesn’t really take us very far, as cavalry only moved at the walk for long periods during extended marches. For shorter marches, it often moved at the ‘trot & walk’, which meant alternating the two modes. The average speed for this was 200 paces per minute for light cavalry. The faster gaits of canter & gallop were mainly used in the charge. At the gallop the final speed was 900 paces per minute for light cavalry. [At these faster gaits heavy cavalry moved a bit slower, but that’s not really an issue for our ACW games.]

How could we reflect this in the game? My suggestion would be to treat the ‘double-time’ button as equating to the average speed for ‘trot & walk’ – ie 2 x walking speed. Rather than enforce the need for alternating gait on the players, – probably not practical with many units to manage – it’s perhaps better to just use the average speed. It may even be that this is the default speed. The key point here is to reduce fatigue (see below), which would restore the cavalry’s mobility and allow more tactical choice. I think this would make for an even more interesting game.

To reflect the speed of a charge, the ‘charge’ button should be modded to produce a higher speed of say 4 x walking speed.


Fatigue

Whilst ‘trot & walk’ was considerably faster than the walk, it was felt to not unduly tire the mounts – indeed in same ways alternating the gait was thought to be beneficial. This implies that it should come with a higher (but not greatly so) fatigue penalty than moving at the walk – ie a much lower fatigue penalty than it currently suffers

Charging on the other hand was much more taxing, and one full-blooded charge more often than not left the horses winded. That implies increasing the fatigue penalty substantially. However there is a problem with that. To take a battery requires a succession of 4 (or 6) successive charges, as a regiment ceases charging after each gun capture. That may require a fudge as regards fatigue.


Horse & rider

There is third issue, which is probably beyond our modding skills, and may need to await a new game engine. That is the separation of horse & rider. As an extreme example of what I mean, here is an incident from the Battle of Brice’s Cross Roads, fought in 1864 by Forrest’s Confederate cavalry against a superior Union force of infantry and cavalry. As per usual, Forrest was on the attack throughout, even though initially he had only a small force with him. He had called up all his nearby troops. One brigade was 14 miles away and (supposedly) rode at the gallop to join him. We may doubt that any horse could gallop for 14 miles, but that isn’t really the point of the story. What happened next was that, on arrival, the men immediately dismounted and charged on foot. Their horses were no doubt blown, but they were not. At the moment there is no way the game can reflect this.

Thoughts anyone?

Martin (J)

Fatigue:
The file to examine is statetables.csv. It contains the fatigue modifications for each type of unit. Except for charging, the cavalry should be a bit more robust than the infantry. From my limited experience with cavalry, I don't believe this is the case. Fatigue is suppose to go from 0-1000, with 1000 being fully rested for all units. My guess is that it is not 1000 for cavalry, but something less.

The first thing to do is to find out how much less this value is. That can determined by running infantry and cavalry across the same 'obstacle course till they become exhausted. I'll see if I can do this in the next day or two. Once that number is determined, we can begin the debate as to how robust cavalry should be compared to the infantry.

Speed:
The speed of the unit types is contained on unitglobal.casv. The run/charge speed, (double time), of the cavalry is 15. For the infantry it is 8. That's not unreasonable and it may be that if we fix the fatigue issue and allow the cavalry to run farther without becoming exhausted, the speed may not be an issue.

I'd recommend working on the fatigue issue first and leave the speed question for later.

That sounds a sensible way forward Kevin, and many thanks for your offer to test. I will be very interested in what you discover.

Also, if changing the speed requires modifying the unitglobal.csv file, that's another argument for leaving it alone, as I believe that file is currently barred to MP modding.

I have been discussing Davinci's uniform mod with him recently. Within that he has also adjusted speed settings to make units march more slowly. More historical, but I'm not sure it makes for as good a game. He is saying he did it in the drills.csv file (Column K).

Martin (J)

As is often the case, it always pays dividends to make a quantitative measurement. Perception in this case is flat wrong. I created a regiment of infantry and type 1 cavalry with the exact same characteristics. I then varied the speed and fatigue values of the cavalry. I then ran the two units, one behind the other until they were completely exhausted, as shown by the fatigue bar.

Same running speed, same fatigue value:
The cavalry fatigues 20 seconds sooner than infantry. They keep nearly the same fatigue levels until they reach the exhausted state. At that point the cavalry fatigues faster. This makes no sense but it is repeatable.

Cavalry speed X 2 same fatigue values as infantry:
Cavalry reaches full fatigue ~1.5 minutes sooner than infantry. It is about 88% of the time needed for the infantry to fatigue. This is probably due to the fact that the cavalry travels much farther. It crosses more fatiguing terrain than the infantry, so it also gains fatigue faster.

Cavalry speed same, fatigue 90% of infantry, (stock value).
Infantry reaches full fatigue ~25 sec sooner than the cavalry. this is ~96% of the value of the cavalry time needed to fatigue. It should be 90%. This confirms that the cavalry rate of fatigue increases once it reaches the exhausted state.

Conclusions:
1. Stock type one cavalry are a bit more robust running than infantry and the same while walking.
2. Cavalrymen are whiners.

So now all that is left is to decide how reasonable #1 is. I believe conclusion 2 is a canonical law and therefore unchangeable.

I have, in my recent battles with cavalry (read that either way!), not used double time at all, IIRC. I walked them everywhere and they still got very tired very quickly. They were also very very skittish, retreating with almost no provocation.

From your figures, Kevin, cavalry double time is twice as fast as infantry double time, which sounds fine. In both cases this would have to be used very sparingly.

I'd like cavalry's default speed to be twice as fast as infantry's, as per Martin's KS sources. Do you know how these two speeds compare (in-game), and how fatigue compares?

Sorry if this doesn't make sense. Pre-Christmas tiredness is numbing my brain, which may also be dulled a bit by my recent secondment to the cavalry ...

Jeff

If I had played Cavalry in the last Scenario I'd have preferred taking them on a scouting mission North. Double-timing should never be out of the question in my book. Better to be present with a tired regiment than being not present with a fresh one.

Here are the stock fatigue values used in the game.

Unit	Walk	Run	Charge	Melee
Infantry	-2	-10	-12	-15
Cavalry with sabers (type1)	-2	-9	-11	-45
Cavalry w/o sabers (type2)	-5	-12	-15	-20

Type 1 cavalry is just a bit more robust than infantry when running or charging. Type 2 is very much at a disadvantage to infantry when walking and only a little poorer when running. These numbers are just factors that are added to others to determine the final fatigue loss. But those other factors are the same for all types of units.

With regards to speed, both infantry and cavalry walk at 4mph. Cavalry run at 15mph and infantry at 8mph.

What difference do the different formations make to speed and fatigue?

We will need to make couriers faster than any other unit. In fact we need to do that now.

I remember Charles Grant's research for his original 1960s rules "The Wargame" involved quite a lot of investigation about cavalry speeds. While I'm happy to have SoW cavalry move faster on the battlefield than infantry I think 2x is too quick and 1.5x is about right. Basing move distances on paces per minute and other parade ground data isn't a good idea as the ground of a battlefield was always worse than any parade ground and units would need to halt at intervals to dress ranks.

Grant also observed that mounted charges were nothing at all like we are accustomed to seeing on our TV screens and in fact were quite slow and controlled affairs; this latter reason being that if everyone went at it hell for leather and screaming and waving their swords, the officers at once lost control of their units. The charge would begin at the walk and only when within some 200 yards of the enemy would the trot and then the canter be sounded. The gallop was reserved for the final 50 yards or so.

In wargame terms this means that at the moment of impact mount and rider would be moving pretty fast but overall, the speed of the charge on average from start to finish was not that much faster than a normal pace of walk/trot or walk/canter the unit used to move about the battlefield.

Based on this I think we still only need two speeds, the basic manouvering speed which would be 1.5x infantry speed and would equate to walk/trot; and secondly a charge or run speed that would be somewhat faster than an infantry run, say 2x (about the speed we have now). This faster pace should tire cavalry more than it tires infantry.

I see from Kevin's table that cavalry suffers a massive fatigue penalty from melee, which should probably remain as we want to discourage actual melee and the very short instancess of "melee" when caavlry capture guns hardly counts and sso won't affect them much.

Separating mount from rider is I fear, impossible in SoW. We'll have to prompt Norb to include this in SoW v2.

Mr. Digby wrote:What difference do the different formations make to speed and fatigue?

We will need to make couriers faster than any other unit. In fact we need to do that now.

Formation speeds are set in drills.csv The addition/subtraction to the base speed is the same whether the unit is infantry or cavalry. In other words a line of inf. or cav. will walk at exactly the same speed. Formations do not affect fatigue other than via the time it takes to cross bad terrain.

Couriers are already the fastest unit in the game. They have a speed of 24mph. It doesn't matter how fast you make them, they will never deliver their message to a running commander or cav. regiment while chasing them from behind. Xeno's Paradox in action.

Surely if the couriers are made faster still, they will catch any slower unit?

Mr. Digby wrote:Surely if the couriers are made faster still, they will catch any slower unit?

Nope, never. They catch up, stop and try to deliver their message. In the time it takes to hand over the message, the commander has moved on just enough to put him out of reach. The speed of the courier doesn't matter. The time it takes to hand over the message does. Is tomorrow a holiday?

Yes, as are the next 2 weeks

Mr. Digby wrote:Yes, as are the next 2 weeks

Must be nice! Is it "Queen Buys Another Ugly Hat Festival" or "Prince Phillip Ain't Dead Yet Month"?

Niether, just Christmas Hols.

I had a sleep on your answer about couriers pursuing units to deliver messages and realised your answer cannot be right, because if it was then any moving unit could not be contacted by a courier. Couriers happily contact walking units, it is only running mounted units they can't catch. They catch running infantry after a short time.

I had this problem in the Price Goes Home battle when I accidentally ordered a battery commander forwards and sent him off at the run. I sent a courier countermanding that order to change it to a walk and as you'd expect the courier followed the artillery officer for a while across the map. However when the officer reached a creek bed, he slowed down and the messenger delivered his orders.

So it must be a function of relative speed and therefore increasing the courier speed surely must fix the problem. If couriers currently move at 24 mph, lets test by making it 26, then 28 then 30 until we get to a speed that will catch running units.

Also I edited my main post above as you were answering, Kevin, so my thoughts on cavalry speeds and fatigue might need addressing. I'm wary of making cavalry into superfast supermen.

I do agree we need to lower the rate at which they become fatigued, though I think this may be a function of them being ordered around at the run because their walk pace is too slow. If we make the basic walk speed faster there will be less need for players to order them around at the run and so less fatigue generally. I'd vote that we try that first and see if it helps, perhaps in time for a Price Goes Home refight.

I'm of the view that cavalry are useful in scenarios like Martin designed where they perform a true function of finding the enemy. I think we need to make sure we don't make them too effective at actually fighting him. Their weakness now in that respect is about right.

I had a sleep on your answer about couriers pursuing units to deliver messages and realised your answer cannot be right, because if it was then any moving unit could not be contacted by a courier. Couriers happily contact walking units, it is only running mounted units they can't catch. They catch running infantry after a short time.

I had this problem in the Price Goes Home battle when I accidentally ordered a battery commander forwards and sent him off at the run. I sent a courier countermanding that order to change it to a walk and as you'd expect the courier followed the artillery officer for a while across the map. However when the officer reached a creek bed, he slowed down and the messenger delivered his orders.

Your observations are correct. It is the speed of the recipient that is important. If you watch the courier closely you'll see that he catches up, stops, the horse begins to shimmy. The horse dancing is the sign that the courier is trying to deliver his message. Running infantry are still close enough to the courier so that after the dance they still get the message. Running commanders are moving too fast and are out of range when the dance finishes.

I have already tried increasing the courier speed to 100mph. It doesn't help. You can try it yourself. Just download the SDK and edit unitglobal.csv About half way down is the courier entry. Change the running speed from 24 to what ever number you like. Save it and put the file in the \Logistics folder of Courier&Maps. At 100mph the couriers are fun to watch. Very zippy.

Mr. Digby wrote:I remember Charles Grant's research for his original 1960s rules "The Wargame" involved quite a lot of investigation about cavalry speeds. While I'm happy to have SoW cavalry move faster on the battlefield than infantry I think 2x is too quick and 1.5x is about right. Basing move distances on paces per minute and other parade ground data isn't a good idea as the ground of a battlefield was always worse than any parade ground and units would need to halt at intervals to dress ranks.

Let me ride to the defence of kriegsspiel here

I don’t think it’s right to describe the material as parade ground data. The game was designed to reflect battlefield conditions, and used by most European armies for many decades to train officers. Had movement rates been found inappropriate for that purpose, the game would have been modified, as it was later in other areas (eg combat, with the development of new weaponry).

Your comment about battlefield conditions is spot on though, and the kriegsspiel manuals do go into some detail on the adjustments which need to be made for slopes, woods etc etc. And I agree re the need to stop and adjust ranks. But these factors applied to all march speeds, and would have reduced the speed of walking too. What is important is the ratio between them I think. Have looked again at the adjustments, and in broad terms they don’t much affect the ratios between walk, trot & gallop.

I’m not seeking to denigrate Charles Grant's research, but would make two points:
(a) the original Prussian ‘Reisswitz’ rules were only unearthed from the archives and published in English in the mid 1980s, so were probably not available to him
(b) he was not a serving officer in the horse & musket period, as those who authored kriegsspiel in 1824 (and the later updates) were. I think it’s reasonable to assume they would have had a good grasp of military practicalities. The initial version was personally endorsed by the chief of the Prussian General Staff (von Muffling), who was so impressed with it that he ordered copies disseminated to every regiment in the army.

Mr. Digby wrote:Grant also observed that mounted charges were nothing at all like we are accustomed to seeing on our TV screens and in fact were quite slow and controlled affairs; this latter reason being that if everyone went at it hell for leather and screaming and waving their swords, the officers at once lost control of their units. The charge would begin at the walk and only when within some 200 yards of the enemy would the trot and then the canter be sounded. The gallop was reserved for the final 50 yards or so.

In wargame terms this means that at the moment of impact mount and rider would be moving pretty fast but overall, the speed of the charge on average from start to finish was not that much faster than a normal pace of walk/trot or walk/canter the unit used to move about the battlefield.

All fair points. I tried to account for this with my 4 x walk suggestion, but did not go far enough, particularly as I misstated the gallop speed. It should actually be 450 paces per minute, not 900!

I too have the perception that they only moved up to canter and then gallop in the final phase, but was always under the impression that they moved to a trot asap before that. Even a trot was 300 paces – ie 3 x walking speed. There are a number of reasons why they might have wanted to move faster, earlier in the charge. Any charge made under fire would suffer fewer casualties & disruption if the ground was covered quickly. And of course if you were charging infantry, it was imperative to close quickly before they formed square (Napoleonic) or redeployed to defensive terrain (later). I must confess that I lack supporting facts on this one, and no doubt the circumstances of each charge were different anyway. Do you have any evidence for the 200 yard figure?

Mr. Digby wrote:Based on this I think we still only need two speeds, the basic manouvering speed which would be 1.5x infantry speed and would equate to walk/trot; and secondly a charge or run speed that would be somewhat faster than an infantry run, say 2x (about the speed we have now). This faster pace should tire cavalry more than it tires infantry.

There is an attraction in leaving the game at 2 speeds, rather than moving to 3, if we can make it work. Personally I'm not sure of that yet.

Mr. Digby wrote:I see from Kevin's table that cavalry suffers a massive fatigue penalty from melee, which should probably remain as we want to discourage actual melee and the very short instancess of "melee" when caavlry capture guns hardly counts and sso won't affect them much.

Yes that’s my feeling too.

Mr. Digby wrote:Separating mount from rider is I fear, impossible in SoW. We'll have to prompt Norb to include this in SoW v2.

Sadly yes.

Martin (J)

Mr. Digby wrote:I do agree we need to lower the rate at which they become fatigued, though I think this may be a function of them being ordered around at the run because their walk pace is too slow. If we make the basic walk speed faster there will be less need for players to order them around at the run and so less fatigue generally. I'd vote that we try that first and see if it helps, perhaps in time for a Price Goes Home refight.

I thought about suggesting this idea myself, but can see two problems with effectively dispensing with cavalry ‘walk’ speed:

1. doing so is unhistorical, because 'walk' was frequently used. Although ‘trot & walk’ gave a good bang for the buck, it was still more tiring than walking. So you didn’t use it when you didn’t need to.

2. For mixed infantry/cavalry units (brigades, divisions or corps), you cannot road march properly if the different components have different base speeds. Any attempt to do so will end up in a jumbled mess. I suspect in game terms that would not make a difference, because I don’t *think* the game penalises units occupying the same stretch of road as it should. But again it would be unhistorical, and also not visually appealing.

Mr. Digby wrote:I'm of the view that cavalry are useful in scenarios like Martin designed where they perform a true function of finding the enemy. I think we need to make sure we don't make them too effective at actually fighting him. Their weakness now in that respect is about right.

Yes I agree on both counts.

Martin (J)

I should point out that we have the ability to adjust cavalry fatigue moving over non-open/roads terrain. For instance, it would be impossible for formed mounted cavalry to move through heavy woods for instance. We can give mounted troops an additional penalty for such behavior while not penalizing dismounted troops. Note, this penalty would apply to all terrain that has a fatigue value greater than 0.

A bit of (hopefully) relevant horse trivia pulled from local horse experts where we board our mare, ("Lonesome Lola"). ;-)

-A horse's normal walk is only slightly faster than a human's: about 4 mph vs 3 or thereabouts (interestingly, Hardee's School of the Soldier used prominently in the Civil War called for infantry to be trained to achieve, at long marches on the double-quick with knapsacks and weapons, 5 miles in 60 minutes -- Those guys were tougher than most of us today )
-Most horses trot at around 8 mph; canter/lope ~12-14 mph
-A rider on a walking horse tires about as quickly as a person walking;

Although it seems counter-intuitive, as a kid growing up in the rural West, I clearly remember talking with an old cavalry veteran who insisted that in the army's experience, a horse tires more quickly walking than trotting. While he may have been exaggerating a bit, his point that a horse can go a very long time at a slightly faster pace is valid, I think.

Naturally, there are lots of other variables involved in coming up with an historic rate for cavalry travel, e.g., the load, available feed, terrain.

-Ron

MajorByrd wrote:If I had played Cavalry in the last Scenario I'd have preferred taking them on a scouting mission North.

At the risk of ignoring cavalry precedent, I was following my orders

MajorByrd wrote:Better to be present with a tired regiment than being not present with a fresh one.

At the co-op game last week, my regiments of cavalry were, to all intents and purposes, useless. With hindsight, I would probably have been more use to my commanding officer as a lone scout riding around the battlefield and sending couriers back with info.

I think we use cavalry for two quite different roles. Reconnaissance can actually be performed by a lone commander, but it looks and feels better if the commander is accompanied by cavalry. Once in battle, on the other hand, you want your men to be able to contribute to the fight. This may be a bit ahistorical but it means you can take part, as a cavalry commander, in a complete game, not just the recon stage. I spent almost the entire battle last week on Prairie Grove nursing and resting my regiments.

So, I don't want cavalry to become über cavalry but I would like to be able to command cavalry and feel that they can move fast around the battlefield and take part in the fighting. The alternative is to accept that cavalry, if at all tired, are useless for fighting and stay in a scouting role for the duration of the game. For that, though, the cavalry are essentially decorative and it would be best to just have a small force, or no troops at all, just your aides.

I would like to see what happens if we make cavalry's walk double the speed of infantry's and remove the correspondong fatigue penalty relative to infantry.

I think this is still historical as our games represent actual battle and the immediate preliminaries. Let the cavalry walk on the campaign map and walk/trot on the battlefield!

Jeff

I would like to see what happens if we make cavalry's walk double the speed of infantry's and remove the correspondong fatigue penalty relative to infantry.

Are you referring to the penalty that the type 2(T2) cavalry has? I agree, I cannot think of a reason that T2 should have more than double the walking fatigue penalty as T1 or infantry. Maybe it was suppose to represent the mounting/dismounting that they do, but that doesn't seem reasonable.

Uncle Billy wrote:

... remove the correspondong fatigue penalty relative to infantry.

Are you referring to the penalty that the type 2(T2) cavalry has?

I was, though I may not have remembered that it was only T2 when I wrote it

Uncle Billy wrote:Are you referring to the penalty that the type 2(T2) cavalry has? I agree, I cannot think of a reason that T2 should have more than double the walking fatigue penalty as T1 or infantry. Maybe it was suppose to represent the mounting/dismounting that they do, but that doesn't seem reasonable.

I don't understand that either. It wasn't as if there were two arms of the cavalry service with mounts sourced from different suppliers. Most cavalry units on both sides acquired horses where they could. This was so unsatisfactory that throughout the war a sizeable proportion of all large commands (divisions & corps) were usually dismounted.

Martin (J)

WJPalmer wrote:Although it seems counter-intuitive, as a kid growing up in the rural West, I clearly remember talking with an old cavalry veteran who insisted that in the army's experience, a horse tires more quickly walking than trotting. While he may have been exaggerating a bit, his point that a horse can go a very long time at a slightly faster pace is valid, I think.

-Ron

That's interesting. Thanks.

I suppose one advantage of the walk is that the horse can get a break from carrying the rider. The rider might not see it that way of course, and cavalry boots might not be ideal for marching........

Martin (J)

Uncle Billy wrote:I should point out that we have the ability to adjust cavalry fatigue moving over non-open/roads terrain. For instance, it would be impossible for formed mounted cavalry to move through heavy woods for instance. We can give mounted troops an additional penalty for such behavior while not penalizing dismounted troops. Note, this penalty would apply to all terrain that has a fatigue value greater than 0.

That's good news. Just to be clear though, are you saying that we can give only one terrain penalty, which has to cover everything from woods, slopes to streams etc? Or have I misunderstood?

Martin (J)