just for flavor, what was the size of forces at manzikert? 100,000 byzantines vs. 40,000 turks IIRC why didnt they die on the way?
-
We have updated our Community Code of Conduct. Please read through the new rules for the forum that are an integral part of Paradox Interactive’s User Agreement.
Akka le Vil said:
Sure they do look what happened to Nappies great army. In fact attrition, which includes desertions, were many times lots higher, The vast majority of soldiers have always died from sickness, starvations or deserted. Few died in battles before the raiways changed the consept of warfare logistic. Before the 18th century noone ever tried to support their armies in hostile territory. Armies were feed by plunder and tributs. Which is why you over and over again can see why small commander makes seperate peaces with other small commanders in order to get their supplies. If you study what happend to the forces of Gallas during the thirty years war (a general that regulary tried to enter pilliged land) you will see what I mean.
Finally during the middle age noone ever tried to march a army of teens of thousends. Not even armies of thousands throug roadless territory. Please do remember that for every fighting soldier two other followed the army. Wifes, children, whores, bums etc. So why should you be able to do it in the game. Its not funny, not realistic, not even nessesary. Move your troops in many small stacks instead. And even thougt its true noone could coordinate several forces back then they still moved seperatly, just very uncoordinated.
Edit. An armies average marching distance remained unchanged from the romans up to our times. Is 4.5 kilometers a day. Or so I was taught at the swedish Officers school of Karlsberg.
Surely that's 4.5 km/hr, not per day. The crusaders were making 8 km/day as the crow flies through rough country in awful conditions.
Not to be vindicative, but you obviously didn't bother neither to read the previous posts - which already include, often several times, the counter of most, if not all, your points - neither really thought about what you said.BueDigre said:
I don't feel like repeating myself uselessly. Just bother to actually read the thread before answering it, which would spare me from copy-pasting a millionth time the same thing -_-
Akka, I have to agree with Jinnai on this one. Whilst it would have been technically possible to arrange to supply an army in hostile territory, it didn't happen all that often, and wasn't always all that effective. Plus, warfare wasn't approached in that kind of manner for a lot of the time period. On top of the sheer cost of warfare alone, paying to arrange for regular shipments or trains of food and other supplies from provinces depleted of their normal manpower (as they're now soldiers) would be slow, cumbersome and costly.
The alternative? Have the army live cheaply off the land, pillaging as it goes. No extra cost and who cares what happens to the peasants of the enemy, anyway...that was pretty much the line of thinking throughout the era.
If there were horrific effects to plunging into debt akin to the bankruptcy in EU2, then I might agree with your suggestion to base more of the factors on money, but as it stands the worst that can happen to me for being in debt is that a few buildings might be lost until I can convince the Estates General to bail me out (and it has to be said, its ahistorically easy to convince them...). So what if the war costs 20%-30% more? If that guarantees my victory, it'll easily pay for itself in the future...
My thoughts on using money to soothe attrition were just that - to allow you to support a couple of extra thousand men in one place before attrition kicked in because you were doing your level best to keep them supplied, but I think it would probably cause more issues than it would solve.
In terms of the attrition curve, bell-shaped is realistic - or, more accurately, steeply curved. If the province can support 4,000 men through either forage or natural supply (which is really what the supportable troop limit is about), then having an extra 1,000 men means that you're 20% short on food. If it was only a hundred men over, people could probably tighten their belts and attrition would scarcely show, but 1,000 would cause health deterioration and death. 2,000 men over would be far worse.
These effects wouldn't be linear in nature. Once you truly reach your maximum limit of supply then every extra person is condemned to death. Plus, the closer you get to that point, and especially beyond it, the effects would become disproportionately worse. The odd bit of starvation here and there might be forgiven, but you can't expect 5,000 people to casually accept their impending doom. And no-one knows which 5,000 its going to be...there would be fights over food, mutinies, mass desertions of mecenaries and peasants alike, rife conditions for disease...in short, its not a linear relationship, and the curve would be likely be steep as hell. The issue is that its almost certainly kicking in at too low a value, and that it can spiral upwards out of control.
On the side issue of leaving men behind in castles, the entire beauty of a castle was that it could be held and defended with very few men at all, yet tie up a far larger force for quite some length of time. Its unlikely that you would leave thousands of men behind unless you were leaving a garrison to prevent against active rebellion or the threat of re-invasion. In these times of swaying loyalties and pre-nationalism, most peasants weren't over-concerned with who ruled them. At least, not at the moment that they'd been conquered. There would be more pressing worries, like not starving. That action is better represented by leaving a regiment behind, rather than deducting troops from the total. At present, that isn't necessary...but who knows what CK2 will bring.
Regarding your question, Arasul, I'm not sure if martial skill is dependent on attrition. That kind of thinking flows through the other Paradox games, but I don't recall ever seeing a good marshal's army suffering less attrition (at least, the tooltip has never said so), so I expect the answer is no.
Finally, just to remind, the rate cap idea was Byakhiam's (unless it was suggested elsewhere). I just did a bit of work with some figures and think its a good idea.
The alternative? Have the army live cheaply off the land, pillaging as it goes. No extra cost and who cares what happens to the peasants of the enemy, anyway...that was pretty much the line of thinking throughout the era.
If there were horrific effects to plunging into debt akin to the bankruptcy in EU2, then I might agree with your suggestion to base more of the factors on money, but as it stands the worst that can happen to me for being in debt is that a few buildings might be lost until I can convince the Estates General to bail me out (and it has to be said, its ahistorically easy to convince them...). So what if the war costs 20%-30% more? If that guarantees my victory, it'll easily pay for itself in the future...
My thoughts on using money to soothe attrition were just that - to allow you to support a couple of extra thousand men in one place before attrition kicked in because you were doing your level best to keep them supplied, but I think it would probably cause more issues than it would solve.
In terms of the attrition curve, bell-shaped is realistic - or, more accurately, steeply curved. If the province can support 4,000 men through either forage or natural supply (which is really what the supportable troop limit is about), then having an extra 1,000 men means that you're 20% short on food. If it was only a hundred men over, people could probably tighten their belts and attrition would scarcely show, but 1,000 would cause health deterioration and death. 2,000 men over would be far worse.
These effects wouldn't be linear in nature. Once you truly reach your maximum limit of supply then every extra person is condemned to death. Plus, the closer you get to that point, and especially beyond it, the effects would become disproportionately worse. The odd bit of starvation here and there might be forgiven, but you can't expect 5,000 people to casually accept their impending doom. And no-one knows which 5,000 its going to be...there would be fights over food, mutinies, mass desertions of mecenaries and peasants alike, rife conditions for disease...in short, its not a linear relationship, and the curve would be likely be steep as hell. The issue is that its almost certainly kicking in at too low a value, and that it can spiral upwards out of control.
On the side issue of leaving men behind in castles, the entire beauty of a castle was that it could be held and defended with very few men at all, yet tie up a far larger force for quite some length of time. Its unlikely that you would leave thousands of men behind unless you were leaving a garrison to prevent against active rebellion or the threat of re-invasion. In these times of swaying loyalties and pre-nationalism, most peasants weren't over-concerned with who ruled them. At least, not at the moment that they'd been conquered. There would be more pressing worries, like not starving. That action is better represented by leaving a regiment behind, rather than deducting troops from the total. At present, that isn't necessary...but who knows what CK2 will bring.
Regarding your question, Arasul, I'm not sure if martial skill is dependent on attrition. That kind of thinking flows through the other Paradox games, but I don't recall ever seeing a good marshal's army suffering less attrition (at least, the tooltip has never said so), so I expect the answer is no.
Finally, just to remind, the rate cap idea was Byakhiam's (unless it was suggested elsewhere). I just did a bit of work with some figures and think its a good idea.
Exactly. And the amount of money you are talking about to get around the lack of infrastucture, corruption (yea...hmm how easily we forget that little thing), degradation and loss of of supplies in route (from normal use, wear and tear and accidents), weather problems, compendating for illnesses, making sure the troops are happy, and then after all that, making sure that train knows where the hell you're at and your talking bankruptsy not for just a few decades but a few centuries. Even the estates general wouldn't be able to help you out.Woz Early said:
And then of course, nothing is guaranteed, especially a supply line.
At or below the support limit is fine, after all it doesn't have a support limit displayed for nothing. Above that and your straining your resources. You might be able to feed a couple hundred more, for one month or so, but after/more than that things are going to get ugly. The longer and higher the amount the uglier it will get. And this goes for everyone. And that support limit isn't just for your army, its for everyone.Woz Early said:In terms of the attrition curve, bell-shaped is realistic - or, more accurately, steeply curved. If the province can support 4,000 men through either forage or natural supply (which is really what the supportable troop limit is about), then having an extra 1,000 men means that you're 20% short on food. If it was only a hundred men over, people could probably tighten their belts and attrition would scarcely show, but 1,000 would cause health deterioration and death. 2,000 men over would be far worse.
These effects wouldn't be linear in nature. Once you truly reach your maximum limit of supply then every extra person is condemned to death. Plus, the closer you get to that point, and especially beyond it, the effects would become disproportionately worse. The odd bit of starvation here and there might be forgiven, but you can't expect 5,000 people to casually accept their impending doom. And no-one knows which 5,000 its going to be...there would be fights over food, mutinies, mass desertions of mecenaries and peasants alike, rife conditions for disease...in short, its not a linear relationship, and the curve would be likely be steep as hell. The issue is that its almost certainly kicking in at too low a value, and that it can spiral upwards out of control.
On the side issue of leaving men behind in castles, the entire beauty of a castle was that it could be held and defended with very few men at all, yet tie up a far larger force for quite some length of time. Its unlikely that you would leave thousands of men behind unless you were leaving a garrison to prevent against active rebellion or the threat of re-invasion. In these times of swaying loyalties and pre-nationalism, most peasants weren't over-concerned with who ruled them. At least, not at the moment that they'd been conquered. There would be more pressing worries, like not starving. That action is better represented by leaving a regiment behind, rather than deducting troops from the total. At present, that isn't necessary...but who knows what CK2 will bring.
That said a cap of 99% should be placed. No matter how bad things get a few people will go to any lengths to ensure they survive.
A bell curved graph would certainly stop the use of massive stacks of armies which is also unhsitoric as long as the attrition doesn't start to kick in bad until you actually cross that support limit threshold.
Woz Early said:
You forget that the army was only a miniscule percentage of all inhabitants in a province. Increasing over what you call the "natural" supply limit would result in more deaths among peasants than it did among soldiers, as a hungry army would take what they needed, including seed stock or livestock and so on. This would wreak havoc with discipline and fighting effectiveness, but not result in automatic death for the "excess." (Automatic creation of highway robber bands and "looting" of even friendly provinces by a large army would not be uncalled for.)
Well, I'll separate this post in three parts
1) The sake of the argument :
Indeed, going bankrupt should have a serious effect. Hopefully, it's something that is very easy to alter, even for an amateurish modder like me : just put an event that fires often once you're in at 0 gold, and which randomly kills a regiment.
Totally wrong.
First, you have to remember that hunger isn't some hard-defined treshold. You can eat a little less food and not be worse. You can drastically reduce your food consumption and be quite worse and much weaker, but still not starving nor dead. As such, saying that only 10 % more people would suddendly make the whole army enter into hunger-madness is absurd : they would simply, as you said, tighten their belt. Take a morale loss. Perhaps a few (not 15 % as in the game) would desert, but that's all.
Second, as Jaok pointed, the population of a province is still quite large compared to anything but a huge army, and the province is able to feed them for the WHOLE YEAR, not just a few monthes. So there is more food, even in semi-desert province, than you may think.
Third, there is supplies arriving. They may be largely not sufficient, but they are still here, and compensate partially for the lack of food.
Fourth, if your army starts to be in really dangerous conditions regarding supplies, then people will die and desert, but they won't do it proportionnally higher if they are more numerous. Again, as I said it previously, you seem to forget that a 15 % of losses ALREADY kills more people in a big army than in a small. Making the percentage skyrocket is useless, considering how fast an army diminish in size with such an attrition. And additionnally, the very point of this thread is precisely that attrition gets enormous too fast. It is the main point to correct (a cap is good, but as again I said previously, if the curve is not flattened, then a cap simply means we'll always be at the maximum attrition, until the army has been reduced in to the little size that suffer no attrition, which will be VERY fast).
Anyway, I consider that "castles" don't simply represent a single castle, but the whole network of fortifications, including cities (it seems to me a bit absurd to think that "huge castle" in Byzantion refers to some lone castle somewhere, and not Constantinople itself).
And if baron's castles could be manned, in fact, with something like 50-60 men, big castles (like the Crac des Chevaliers, which was garrisonned with more than 2000 men) and, even more, cities, weren't so small.
Additionnally, it's the number of men left for the WHOLE PROVINCE, so a few hundred wouldn't feel exagerated, I think.
Still, this isn't something I'm bent-on about. You may make it 250 men per level of fortification and I would still find it fine. Or even 100 men per level, if we can rationalize it by them only being put in key points, and people from the province just changing their allegiance.
But still, I think that 500 men per level is big enough to make an impact (and make the player think twice about taking everything in his path), and small enough to not cripple any army going through.
Here, the argument for the sake of the argument is done.
Now...
2) The Facts
Regardless of the argument and explanations we give, the facts stay, that attrition (except in exceptionnally dire circumstances, which are, well, exceptionnals) DIDN'T reach the levels we see in the game, by whole scales. We can rationnalize all the day long, it's only theorycraft if the reality shows differently, and, well, the reality DOES show differently.
Rome, Greece, Persia, Byzantine Empire, Arabs, Mongols, Chinese and so on WERE able to field hundred-thousands-like armies and to keep them working (and not vanishing out of attrition in a few weeks). It was, in fact, only in Western Europe that it didn't happen, up to the Renaissance, when precisely Europe grew rich again and could afford again big armies (while technologically, supplies were the same than in Middle-ages).
As such, regardless of the arguments we can throw, there is this simple fact we can't go against : history denies such level of attrition. They are unrealistic.
1) The sake of the argument :
Yes, which is why, as I said above, only wealthy and organized kingdom could afford such a thing. Which is also why there was usually larger armies in the East, and larger armies in the Antiquity, while medieval Western Europe, still quite poor, couldn't afford that and had essentially pillaging armies.Woz Early said:
Yes, and such tactics were still used under Napoléon, who was able to march armies of up to 100 000 soldiers without significant attrition, as I already pointed it out already.
That's the best argument so far, considering all the others were already answered in previous posts
Indeed, going bankrupt should have a serious effect. Hopefully, it's something that is very easy to alter, even for an amateurish modder like me : just put an event that fires often once you're in at 0 gold, and which randomly kills a regiment.
Why ?
WRONG !
Totally wrong.
First, you have to remember that hunger isn't some hard-defined treshold. You can eat a little less food and not be worse. You can drastically reduce your food consumption and be quite worse and much weaker, but still not starving nor dead. As such, saying that only 10 % more people would suddendly make the whole army enter into hunger-madness is absurd : they would simply, as you said, tighten their belt. Take a morale loss. Perhaps a few (not 15 % as in the game
) would desert, but that's all.Second, as Jaok pointed, the population of a province is still quite large compared to anything but a huge army, and the province is able to feed them for the WHOLE YEAR, not just a few monthes. So there is more food, even in semi-desert province, than you may think.
Third, there is supplies arriving. They may be largely not sufficient, but they are still here, and compensate partially for the lack of food.
Fourth, if your army starts to be in really dangerous conditions regarding supplies, then people will die and desert, but they won't do it proportionnally higher if they are more numerous. Again, as I said it previously, you seem to forget that a 15 % of losses ALREADY kills more people in a big army than in a small. Making the percentage skyrocket is useless, considering how fast an army diminish in size with such an attrition. And additionnally, the very point of this thread is precisely that attrition gets enormous too fast. It is the main point to correct (a cap is good, but as again I said previously, if the curve is not flattened, then a cap simply means we'll always be at the maximum attrition, until the army has been reduced in to the little size that suffer no attrition, which will be VERY fast).
Hu, well, that's probably why I said "500 men by level of fortification", which isn't several thousands until you reach with the highest-level ones
Anyway, I consider that "castles" don't simply represent a single castle, but the whole network of fortifications, including cities (it seems to me a bit absurd to think that "huge castle" in Byzantion refers to some lone castle somewhere, and not Constantinople itself
).And if baron's castles could be manned, in fact, with something like 50-60 men, big castles (like the Crac des Chevaliers, which was garrisonned with more than 2000 men) and, even more, cities, weren't so small.
Additionnally, it's the number of men left for the WHOLE PROVINCE, so a few hundred wouldn't feel exagerated, I think.
Still, this isn't something I'm bent-on about. You may make it 250 men per level of fortification and I would still find it fine. Or even 100 men per level, if we can rationalize it by them only being put in key points, and people from the province just changing their allegiance.
But still, I think that 500 men per level is big enough to make an impact (and make the player think twice about taking everything in his path), and small enough to not cripple any army going through.
Well, CK2 is a long way ahead, for now I'm just giving opinion on the next path(s).In these times of swaying loyalties and pre-nationalism, most peasants weren't over-concerned with who ruled them. At least, not at the moment that they'd been conquered. There would be more pressing worries, like not starving. That action is better represented by leaving a regiment behind, rather than deducting troops from the total. At present, that isn't necessary...but who knows what CK2 will bring.
Here, the argument for the sake of the argument is done.
Now...
2) The Facts
Regardless of the argument and explanations we give, the facts stay, that attrition (except in exceptionnally dire circumstances, which are, well, exceptionnals) DIDN'T reach the levels we see in the game, by whole scales. We can rationnalize all the day long, it's only theorycraft if the reality shows differently, and, well, the reality DOES show differently.
Rome, Greece, Persia, Byzantine Empire, Arabs, Mongols, Chinese and so on WERE able to field hundred-thousands-like armies and to keep them working (and not vanishing out of attrition in a few weeks). It was, in fact, only in Western Europe that it didn't happen, up to the Renaissance, when precisely Europe grew rich again and could afford again big armies (while technologically, supplies were the same than in Middle-ages).
As such, regardless of the arguments we can throw, there is this simple fact we can't go against : history denies such level of attrition. They are unrealistic.
Ah, and another "nice" effect of the insane attrition that I had just today...
I've some provinces around Venezia, which is a VERY wealthy province. I've got a VERY high stewardship. There is a war against Germany. The two provinces just north of Venezia are besieged by the enemy. I raise the army of Venezia => about 50 000 soldiers (yep, that's a lot). I do it on the 1 of the month not to suffer the attrition (around 20-30 % in the home province fully equiped ), break the siege, then go to lift the other siege in the neighbouring province, thinking that even if one third ( ) of my soldiers die (in the middle of one of the most lenient and richest province of the game, fully supplied and with excellent roads and support building...), the 35 000 left will be enough to fight the 16 000 that are actually besieging.
The only little problem is that I didn't paid attention that the ennemy army I just routed was still not out of the province... Not a very big army, something like 8-10 000, but it made the attrition worse... And also, I didn't paid attention that the 20-30 % attrition was in Venezia, which is richer than the liberated province just north-west of it...
Result, I see too late that there is whopping 87 % attrition ( ), and take it. The army then finish its 10-days trip to the other beseiged province, and, of course, being reduced to less than 8000 men, is crushed...
40 000 deads in one month due to the horrible fact of being at home in very good conditions. That's just so sensible...
I've some provinces around Venezia, which is a VERY wealthy province. I've got a VERY high stewardship. There is a war against Germany. The two provinces just north of Venezia are besieged by the enemy. I raise the army of Venezia => about 50 000 soldiers (yep, that's a lot). I do it on the 1 of the month not to suffer the attrition (around 20-30 % in the home province fully equiped
), break the siege, then go to lift the other siege in the neighbouring province, thinking that even if one third ( ) of my soldiers die (in the middle of one of the most lenient and richest province of the game, fully supplied and with excellent roads and support building...), the 35 000 left will be enough to fight the 16 000 that are actually besieging.The only little problem is that I didn't paid attention that the ennemy army I just routed was still not out of the province... Not a very big army, something like 8-10 000, but it made the attrition worse... And also, I didn't paid attention that the 20-30 % attrition was in Venezia, which is richer than the liberated province just north-west of it...
Result, I see too late that there is whopping 87 % attrition (
), and take it. The army then finish its 10-days trip to the other beseiged province, and, of course, being reduced to less than 8000 men, is crushed...40 000 deads in one month due to the horrible fact of being at home in very good conditions. That's just so sensible...
I agree, joak, it would be very rare for the army to be larger than the number of inhabitants in the province, but there's more to it than that...
If inhabitants knew an army was coming, they wouldn't think "Oh well, never mind" and leave the fields alone. Food would be rapidly harvested, especially by the castles and major towns, which would try to stockpile and then seal themselves off from the invaders outside. Naturally, not all the food or wealth in the province could be secreted away, but a fair amount could be. If the defenders were expecting the possibility of a lengthy siege, they might well burn many fields and slaughter any livestock they found to deny their attackers supplies of food.
That said, the current level of max support across provinces is, IMO, low if you compare it historically...since testing, I've not disagreed with that at any point, you might note.
Akka, whilst I understand the examples of attrition, etc. from Napoleonic and Roman times being comparable, I think if wealth-related issues are to be assessed accurately and fairly then we need to stay within the period. If the overwhelming majority of Europe was incapable of raising the wealth required to supply an army in the field, then it seems nonsensical to implement the ability for them to do so. Its different if it was a matter of choice, that Kings could afford to pay for supplies but chose not to, and there are good examples of Kings that *did* pay for regular supplies in the period, but AFAIK that isn't the case.
Psst...and if you do implement a bankruptcy event, I'd have it trigger starting at -1 gold, or the AI will be crippled in every battle. Plus, it should do more than randomly kill a regiment. Bankruptcy should destroy prestige, cause rampant chaos and disorder in the province as the liege and his court scratch around for any money they can, trigger high likelihood of forced acquisition of the Temperate trait or general health loss and higher chance of illness amongst all characters in the court, force changes to elective law as vassals grow tired of their liege's financial ineptitude, etc. Bankruptcy should be terrifically damning if you're going to make military expertise almost entirely dependent on wealth (given that wealth is already what defines army size).
This is along the lines of what I meant by creating more issues than the slider solution would solve. I haven't completely abandoned the idea as idiotic, but the more I think about it the more I think you'd really need to implement a lot of other features to stop it being a nice, easy exploit for the player.
And finally, I think you're grossly mistaken about the attrition curve, and full caps won't make me change my mind.
Hunger isn't a hard-defined threshold, but starvation is. Plus, by the time a good number of people are beyond the starvation threshold. Starving to death will be only one effect, coupled with disease, desertion and rioting. Those factors don't increase gently in a nice linear fashion (oh, two people died, so we'd better have a small fight this evening and draw straws to decide which one of us deserts...oh, four people died? Sorry, we'd better make sure six people get beaten up and draw straws twice). At the lower levels, the odd man or two dying of starvation or deserting are just background noise. These things happen in armies. Fifty men dying or deserting would cause rumours. A thousand could easily cause panic. Morale would be through the floor and given that morale affects desertion and the ability to keep order in the army, visible attrition would rise sharply and people deserted, fought over food and, to boot, began catching diseases from the dead and dying. That's a non-linear relationship.
In terms of food supply, the same applies. Consider that you have the food for 10,000 men. Not food for 10,000 men (and if we share it round a bit it could stretch to 15,000), but having already done that - 10,000 being the limit of people you can feed and keep healthy enough to march and fight. The 10,001th man is screwed. You simply can't afford to feed him, or else at least two of your men are going to collapse from starvation. So he dies - oh well, that's 0.01% attrition, who cares.
Assume for simplicity that those affected die within a month (with no food...go figure) and scale the figures up: 10,000 = 0%, 20,000 = 50%, 30,000 = 66%, 40,000 = 75%, 50,000 = 80%. 0-50-66-75-80 for equal manpower increments: a non-linear relationship.
I don't think that the supply limit in CK is supposed to represent that absolute threshold, hence why initial attrition is fairly light (go a couple of hundred over the threshold and it rarely clears 0.5%), but as you get further over it the problems magnify. Not just starvation, but all of the other effects that we've been mentioning too. I fail to see how that is supposed to make for a linear relationship.
The key issue is where at what height that curve should be pegged, as I've said several times already. People have yet to come up with any figures. Right now, its pegged pretty much at the supply limits displayed on the provinces (unless they have extensive roadnet, in which case roughly double the value), since that's where the lowest part kicks in. When you're at double the supportable limit with no roads, IIRC then you run at about 10-15% attrition.
I have a strong feeling that its pegged too low, and armies are suffering disproportionate amounts of attrition because they are further up the curve than they should be for their size, NOT because the curve itself is too steep.
Please, PLEASE can someone give their thoughts on this:
==========================================================
a) How many men should Plains, Hills, Mountains, etc. be able to support before ANY attrition is suffered at the base level (ie, no improvements)
b) What should the maximum possible figure be for each with all improvements that the terrain allows?
c) What is a good 'working' number of men for each terrain type to support with no attrition for the majority of the CK time period (ie, assume small castle and roadnet/extensive roadnet).
a) and c) really need two different figures - one for friendly territory and one for enemy territory.
Get a few values that we agree on, and we can test with them - and if examples are being used from the time period to justify figures, PLEASE bear in mind the dates and any major factors that need to be considered.
If inhabitants knew an army was coming, they wouldn't think "Oh well, never mind" and leave the fields alone. Food would be rapidly harvested, especially by the castles and major towns, which would try to stockpile and then seal themselves off from the invaders outside. Naturally, not all the food or wealth in the province could be secreted away, but a fair amount could be. If the defenders were expecting the possibility of a lengthy siege, they might well burn many fields and slaughter any livestock they found to deny their attackers supplies of food.
That said, the current level of max support across provinces is, IMO, low if you compare it historically...since testing, I've not disagreed with that at any point, you might note.
Akka, whilst I understand the examples of attrition, etc. from Napoleonic and Roman times being comparable, I think if wealth-related issues are to be assessed accurately and fairly then we need to stay within the period. If the overwhelming majority of Europe was incapable of raising the wealth required to supply an army in the field, then it seems nonsensical to implement the ability for them to do so. Its different if it was a matter of choice, that Kings could afford to pay for supplies but chose not to, and there are good examples of Kings that *did* pay for regular supplies in the period, but AFAIK that isn't the case.
Psst...and if you do implement a bankruptcy event, I'd have it trigger starting at -1 gold, or the AI will be crippled in every battle. Plus, it should do more than randomly kill a regiment. Bankruptcy should destroy prestige, cause rampant chaos and disorder in the province as the liege and his court scratch around for any money they can, trigger high likelihood of forced acquisition of the Temperate trait or general health loss and higher chance of illness amongst all characters in the court, force changes to elective law as vassals grow tired of their liege's financial ineptitude, etc. Bankruptcy should be terrifically damning if you're going to make military expertise almost entirely dependent on wealth (given that wealth is already what defines army size).
This is along the lines of what I meant by creating more issues than the slider solution would solve. I haven't completely abandoned the idea as idiotic, but the more I think about it the more I think you'd really need to implement a lot of other features to stop it being a nice, easy exploit for the player.
And finally, I think you're grossly mistaken about the attrition curve, and full caps won't make me change my mind.
Hunger isn't a hard-defined threshold, but starvation is. Plus, by the time a good number of people are beyond the starvation threshold. Starving to death will be only one effect, coupled with disease, desertion and rioting. Those factors don't increase gently in a nice linear fashion (oh, two people died, so we'd better have a small fight this evening and draw straws to decide which one of us deserts...oh, four people died? Sorry, we'd better make sure six people get beaten up and draw straws twice). At the lower levels, the odd man or two dying of starvation or deserting are just background noise. These things happen in armies. Fifty men dying or deserting would cause rumours. A thousand could easily cause panic. Morale would be through the floor and given that morale affects desertion and the ability to keep order in the army, visible attrition would rise sharply and people deserted, fought over food and, to boot, began catching diseases from the dead and dying. That's a non-linear relationship.
In terms of food supply, the same applies. Consider that you have the food for 10,000 men. Not food for 10,000 men (and if we share it round a bit it could stretch to 15,000), but having already done that - 10,000 being the limit of people you can feed and keep healthy enough to march and fight. The 10,001th man is screwed. You simply can't afford to feed him, or else at least two of your men are going to collapse from starvation. So he dies - oh well, that's 0.01% attrition, who cares.
Assume for simplicity that those affected die within a month (with no food...go figure) and scale the figures up: 10,000 = 0%, 20,000 = 50%, 30,000 = 66%, 40,000 = 75%, 50,000 = 80%. 0-50-66-75-80 for equal manpower increments: a non-linear relationship.
I don't think that the supply limit in CK is supposed to represent that absolute threshold, hence why initial attrition is fairly light (go a couple of hundred over the threshold and it rarely clears 0.5%), but as you get further over it the problems magnify. Not just starvation, but all of the other effects that we've been mentioning too. I fail to see how that is supposed to make for a linear relationship.
The key issue is where at what height that curve should be pegged, as I've said several times already. People have yet to come up with any figures. Right now, its pegged pretty much at the supply limits displayed on the provinces (unless they have extensive roadnet, in which case roughly double the value), since that's where the lowest part kicks in. When you're at double the supportable limit with no roads, IIRC then you run at about 10-15% attrition.
I have a strong feeling that its pegged too low, and armies are suffering disproportionate amounts of attrition because they are further up the curve than they should be for their size, NOT because the curve itself is too steep.
Please, PLEASE can someone give their thoughts on this:
==========================================================
a) How many men should Plains, Hills, Mountains, etc. be able to support before ANY attrition is suffered at the base level (ie, no improvements)
b) What should the maximum possible figure be for each with all improvements that the terrain allows?
c) What is a good 'working' number of men for each terrain type to support with no attrition for the majority of the CK time period (ie, assume small castle and roadnet/extensive roadnet).
a) and c) really need two different figures - one for friendly territory and one for enemy territory.
Get a few values that we agree on, and we can test with them - and if examples are being used from the time period to justify figures, PLEASE bear in mind the dates and any major factors that need to be considered.
I like Woz Early's reasoning. Increasing supply limits for realism and establishing a hard cap at some point (say, 20%) for gameplay sound reasonable to me and not even things that are exceedingly difficult to code.
Speaking of attrition... I know this is the exact opposite of what people are asking for in this thread (i.e., lower attrition), but would it perhaps be possible to increase attrition during the winter months, or would that involve too much coding? That was a great feature in EU2, and I don't understand why invading Finland in January would have been easier in 1066 than in 1466.
Akka le Vil said:
I'm sorry, I have to completely disagree here--despite having some issue with the CK model (mostly on the way the curve ramps up.)
I'm not an expert on ancient armies, but the great field armies ones I know of Rome, Greece or Byzantium I know of wouldn't exceed 50k. Which is irrelevant anyway, since the test for CK "realism" is what medeival Europeans did, not what historians or poets report or exaggerate about other times and places. Can you give examples of a European state keeping a field army of even 20k in the field indefinitely?
Even a brilliant general like Henry V lost perhaps 50% of his army between landing at Honfleur and Agincourt (~3 months), despite campaigning with a smallish force in wealthy France. This sort of effect was typical. English campaigns against the Scots--even defensive ones, fought on English soil--would see portions of the army get frustrated and wander back home after a few months.
Even the crusades, after the first, essentially consisted of forces moving in multiple groups moving slowly to the holy land, joining up, then petering out amid supply problems and desertions (except when the exhausted force was defeated more dramatically, as in Egypt or at Nicephorus).
Like I've said, I have some issues with CK's handling of it (exacerbated by campaigns last "too long" and armies that have to march too far), but the levels strike me as pretty reasonable.
Woz Early said:
Actually, I suspect there are very few instances of peasants turning over food to townspeople or nobles for "safekeeping", for obvious reasons.
Actually, after fiddling around more last night, I'm yet more disturbed by the curve and less by the overall level, as per my last post. Losing half a 50k army over 6 months isn't inherently ridiculous.
Except, again, you continue to assume a constant level of food available to the army, as if they all went to a single vending machine then starved when they ran out of the powdered donuts they like. Armies supplied themselves. The bigger the army, the more men you have gathering food. The first 10k men might be able to gather food for 10,000, but the next 10,000 you add aren't going to sit around on their hands. They'll spend more time gathering food, and find enough to feed 80% of them, the next regiment of 10,000 will find enough to feed 60%, and so on.
Beyond that, the reason I don't think the overall levels are too high are because of desertions and disease. You can't ignore these. Which means a 10k army would face desertions at a certain rate, and under a competent leader a 20k one wouldn't be losing them much faster. This flattens out the curve even more.
Ah, well, plenty of posts to answer to ^^
Byzantium and Arabic powers did field a lot of men. Byzantium, particularly, was able not only to field lots of them, but to keep them quite well supplied and battle-ready (particularly before Mantzikert). Arabs were much more irregular, sometimes (like the Mameluks or the Almohad at their beginning) being very militaristics with professionnals large armies, sometimes much more "western-like" (like during a good part of the Crusades).
I thought about putting at -1 rather than 0, to reflect that you aren't strictly in debt at 0, but kept it at that PRECISELY because the AI wouldn't be affected.
The Ai is already immune to debts and lots of nasty effects. Perhaps it's much dumber than a human and does need help, but after a while, it's just becoming tiring to fight an ennemy not obeying the same rules. Ruining a foe by managing the war, limiting your own expenses and waiting his bankrupcy, is a smart move, and if it's to make the AI immune to everything but raw military power, I can just go back to Warcraft, which at least has a decent fighting AI and nicer graphics.
When the castles were besieged, for example, it wasn't "everyone is fine until the point where everyone starve, and then everyone dies". What happened is that the food rations were reduced... then reduced more... then reduced more... and people starved progressively, and some died, then some more, etc...
It's a progressive thing, without such a clear-cut idea as "everyone past the limit dies".
A modern example can be found in the siege of Leningrad during WW2 (not the same technology at all, of course, but starvation by lack of food works the same today than in these days ^^). The city slowly ran out of food, and people died progressively, not suddendly, and even when food was very scarce, they didn't all died.
The siege happened for 900 days, a bit less than 3 years, and the population dwindled from three millions to about 900 000. They had, at some points, supplies coming from a lake. Unsufficient supplies for the size of the city, but sufficient to lift a bit the pressure.
Hey, it's reality, here. Not theory
As for the steepness of the course, anything that can make the attrition goes above 100 % has OBVIOUSLY a problem in the formula.
Samely, look at my example around Venezia : attrition that reduce my army under half the numbers where it would not suffer any attrition, is a clear proof that something is seriouslyflawed.
That's why I have a problem answering your next question : it's far too "treshold-oriented", while the problem with attrition is not when it starts, but how quickly|/i] it grows.
As such, I have a very hard time answering it, because I reason not in "how many men until attrition kicks in ?", but "for how many men does the attrition becomes 1 % bigger ?".
If you're interested in these answers, then I would gladly give them. But sorry, the "treshold" seems far too unnatural to me.
Morevoer, putting up the treshold is easily done by modding roadnet in the game (which I did after the ridiculous fiasco around Venezia this morning -_-), while I think we're trying to obtain something a bit more sophisticated here
Yep. I have not disagreed with attrition grinding armies neither, just the level of it ^^Woz Early said:
Ah, but well, it WAS the case. But rather in occasions than on regular basis.
Byzantium and Arabic powers did field a lot of men. Byzantium, particularly, was able not only to field lots of them, but to keep them quite well supplied and battle-ready (particularly before Mantzikert). Arabs were much more irregular, sometimes (like the Mameluks or the Almohad at their beginning) being very militaristics with professionnals large armies, sometimes much more "western-like" (like during a good part of the Crusades).
No.
I thought about putting at -1 rather than 0, to reflect that you aren't strictly in debt at 0, but kept it at that PRECISELY because the AI wouldn't be affected.
The Ai is already immune to debts and lots of nasty effects. Perhaps it's much dumber than a human and does need help, but after a while, it's just becoming tiring to fight an ennemy not obeying the same rules. Ruining a foe by managing the war, limiting your own expenses and waiting his bankrupcy, is a smart move, and if it's to make the AI immune to everything but raw military power, I can just go back to Warcraft, which at least has a decent fighting AI and nicer graphics.
Yep, all this could be interesting. It's actually not that hard to mod, in fact (though I'm afflicted with the dreaded "lazy" trait and I probably won't do it myself, except for the most basic "regiment killed" ^^).
Yep, that's the risk. But I prefer to have a fun game and restrain myself from exploits, than an irritating game without exploits, as I can control the level of exploits, but not the level of fun ^^
Well, because it's how it works in reality, simple as thatAnd finally, I think you're grossly mistaken about the attrition curve, and full caps won't make me change my mind.
Hunger isn't a hard-defined threshold, but starvation is. Plus, by the time a good number of people are beyond the starvation threshold. Starving to death will be only one effect, coupled with disease, desertion and rioting. Those factors don't increase gently in a nice linear fashion (oh, two people died, so we'd better have a small fight this evening and draw straws to decide which one of us deserts...oh, four people died? Sorry, we'd better make sure six people get beaten up and draw straws twice). At the lower levels, the odd man or two dying of starvation or deserting are just background noise. These things happen in armies. Fifty men dying or deserting would cause rumours. A thousand could easily cause panic. Morale would be through the floor and given that morale affects desertion and the ability to keep order in the army, visible attrition would rise sharply and people deserted, fought over food and, to boot, began catching diseases from the dead and dying. That's a non-linear relationship.
In terms of food supply, the same applies. Consider that you have the food for 10,000 men. Not food for 10,000 men (and if we share it round a bit it could stretch to 15,000), but having already done that - 10,000 being the limit of people you can feed and keep healthy enough to march and fight. The 10,001th man is screwed. You simply can't afford to feed him, or else at least two of your men are going to collapse from starvation. So he dies - oh well, that's 0.01% attrition, who cares.
Assume for simplicity that those affected die within a month (with no food...go figure) and scale the figures up: 10,000 = 0%, 20,000 = 50%, 30,000 = 66%, 40,000 = 75%, 50,000 = 80%. 0-50-66-75-80 for equal manpower increments: a non-linear relationship.
I don't think that the supply limit in CK is supposed to represent that absolute threshold, hence why initial attrition is fairly light (go a couple of hundred over the threshold and it rarely clears 0.5%), but as you get further over it the problems magnify. Not just starvation, but all of the other effects that we've been mentioning too. I fail to see how that is supposed to make for a linear relationship.
When the castles were besieged, for example, it wasn't "everyone is fine until the point where everyone starve, and then everyone dies". What happened is that the food rations were reduced... then reduced more... then reduced more... and people starved progressively, and some died, then some more, etc...
It's a progressive thing, without such a clear-cut idea as "everyone past the limit dies".
A modern example can be found in the siege of Leningrad during WW2 (not the same technology at all, of course, but starvation by lack of food works the same today than in these days ^^). The city slowly ran out of food, and people died progressively, not suddendly, and even when food was very scarce, they didn't all died.
The siege happened for 900 days, a bit less than 3 years, and the population dwindled from three millions to about 900 000. They had, at some points, supplies coming from a lake. Unsufficient supplies for the size of the city, but sufficient to lift a bit the pressure.
Hey, it's reality, here. Not theory
As for the steepness of the course, anything that can make the attrition goes above 100 % has OBVIOUSLY a problem in the formula.
Samely, look at my example around Venezia : attrition that reduce my army under half the numbers where it would not suffer any attrition, is a clear proof that something is seriouslyflawed.
On the contrary. Realism shows that attrition starts quite low, probably even lower than in the game (in fact, it should starts right from the start, with something like 0,5 %), but increase MUCH slower.
That's why I have a problem answering your next question : it's far too "treshold-oriented", while the problem with attrition is not when it starts, but how quickly|/i] it grows.
As such, I have a very hard time answering it, because I reason not in "how many men until attrition kicks in ?", but "for how many men does the attrition becomes 1 % bigger ?".
If you're interested in these answers, then I would gladly give them. But sorry, the "treshold" seems far too unnatural to me.
Morevoer, putting up the treshold is easily done by modding roadnet in the game (which I did after the ridiculous fiasco around Venezia this morning -_-), while I think we're trying to obtain something a bit more sophisticated here
Well, I would like a much lower attrition, but it doesn't mean I disagree with what you say.Quarto said:
Not that I want the attrition to become ever worst, but I agree with the principle of having higher attrition in frozen plains in winter than in summer (and samely, higher attrition in desert in summer than in winter).
Sorry, they didjoak said:
Rome :
The overall standing professionnals armies of Rome were above 500 000 men. Not all in the same province, obviously, but they didn't lived off the land as they were carefully supplied.
As for armies been on the same battlefield, the battle between Constantine and Lucinus at Hadrianopolis, involved, in total, about 300 000 men. It would be something like 150 % attrition with the CK engine (and perhaps even more ^^).
Byzantium :
Mantzikert saw more than 100 000 men together. The emperor part of the army alone was above 60 000, to which you had to add the ones that never fought, left the field and lead to the defeat.
And it was after the bureaucracy had reduced the military power of Byzantium.
Greece :
Alexander led an army of 40-45 000 footmen and about 3-5000 cavalrymen through Persia, Egypt and India. Granted, it was an army living mainly "off the land".
Persia had a standing army of several hundred of thousands men (though scattered through the whole empire).
As I said, Western Europe wasn't very wealthy, and moreover used the feudal system, which was precisely based on "no standing army, call it when you need it".
But Byzantium, for example, had such an army. The famous Varangian guard alone, which was an elite troop, the personnal bodyguards of the Basileus, not just regular troops, numbered 6000 soldiers, and was always in service.
Akka le Vil said:
Yes, I knew the size of the total Roman armies, but it's kind of meaningless in the attrition context. You could easily support 500k in CK, scattered across the old Roman territory.
Not quite the several hundred thousand I objected to.
Somewhat more to the point
OTOH, the Manzikert campaign saw forces separated from the main one, which contributed both to the battle and the defeat. My memory, based on Treadgold's histories, is that total regular soldiers available for non-garrison duty was around 40,000. 100k was an exceptional force for the Byzantine and proved more than they could efficiently handle. They were defeated by a smaller force; the army, still intact after the defeat, then melted away.
But to the extent that there is a conflict, "Western" (including Northern and non-Byzantine Eastern) Europe has to be the standard, especially since it's Western politics that form the basis for the dynastic system. The Byzantine Empire covers about 15% of the CK map, and their professional system was useful for about a century of the CK period (by 1204, Byzantine elites recognized the clear superiority of the "Frankish" system).
In game, I should emphasize I don't think I actually disagree with you that much on what attrition should look like. IMHO it should ideally start lower, but ramp up more slowly and I'd agree with a cap at 20-30%.
Joak - heh. I wasn't implying that the peasants 'willingly' turned over their food to the local garrisons, or willingly burned their own fields. Those would be the actions of the ruling power in the province, figuring that its better to have the peasants outside the fort starve but make it far more difficult to siege the fort out - since when the fort falls, their heads might well be mounted on spikes, or they might be ransomed for bank-breaking sums of money.
Throughout what I've been sketching out I have been assuming that most underlying conditions are constants for the simple reason that it gives figures that are easy to look at and go "that looks about right" or "no, that's nothing like realistic". The problem with assuming that having 10% more people means that you have the potential for 10% more forage (or 8% more forage, or however you want to determine efficiency) is that you're assuming that the province has an effectively infinite supply of food that just needs to be harvested. On that point I disagree, but it doesn't really matter.
Specifically, the example I gave doesn't work perfectly because you need to factor into account the effect of time (ie, what happens in the following month - is there enough food for 10,000 again?). As you rightly point out, the liklihood is that yes, there is enough food for 10,000 men there, and it could be foraged for in the first month unless its seasonally dependent. However, this is basically mortgaging current attrition levels against future ones - excess foraging in early months would make foraging harder in later months with nearby and poorly defended areas exhausted.
Obviously, my example was supposed to be simple...the minute you move away from simple constants then you walk into a minefield. I'm not trying to perfectly model attrition. I was just giving a working mathematical example of why attrition is non-linear, based on one single factor (food supply). I thought it was only Akka that was disagreeing on that point?
Akka - Byzantium is a rather unique example, given its rather vast wealth, and the Arabic countries are non-playable. This is pretty much what I was thinking, that the possibility to support was there, but more of a theoretical possibility than an actual practice, and only the amazingly wealthy countries could afford to tinker with the idea. Given that the superpowers in CK don't need any more of a leg-up as it is, I'd go with not implementing the slider...putting it in would only serve to add an exploit for the player - and as you say yourself you'd rather not use exploits, there's a good reason not to do it.
As for events affecting the AI, the AI already needs all the help it can get. Speaking personally, I run rings round it 9 times out of 10, and the times I don't are normally blatant errors on my part rather than insightful strategy by the AI. The AI has little concept of how to fight a war properly in CK, so giving them events that destroy their already ineptly commanded regiments would just be kicking them while they're down.
Returning to attrition curves, even in the examples you just gave the relationship is still curved. I was just using an arbitrary threshold to show the curved relationship. However, demolishing the threshold doesn't remove the curve. What you've described by rations being reduced further and further and further is just the curve at the lower part of the scale, where the mass death and desertion is the higher part of the scale. My example wasn't "Everyone is fine until everyone dies" but "Those who can survive can only just survive...everyone else will die".
Remember that except in your own territory, attrition always runs at either 1% or 2%, so attrition already starts earlier than the threshold (which is correct). However, the curve is flattened until the army crosses the support threshold of the province, at which point attrition stops being linear background noise and moves into non-linear attrition which is more notable, but flattened at the top with a rate cap to prevent extreme rates of attrition. The idea is to try to reflect the reality of attrition without actually trying to model it.
If you want to talk in terms of marginal cost of attrition, then I'm happy to work with those figures. Personally, I'm not trying to come up with anything sophisticated. I'm just trying to draw up with something simple to implement and test which makes attrition follow more realistic levels and, if your initial assumption is right, will make the game more fun.
Throughout what I've been sketching out I have been assuming that most underlying conditions are constants for the simple reason that it gives figures that are easy to look at and go "that looks about right" or "no, that's nothing like realistic". The problem with assuming that having 10% more people means that you have the potential for 10% more forage (or 8% more forage, or however you want to determine efficiency) is that you're assuming that the province has an effectively infinite supply of food that just needs to be harvested. On that point I disagree, but it doesn't really matter.
Specifically, the example I gave doesn't work perfectly because you need to factor into account the effect of time (ie, what happens in the following month - is there enough food for 10,000 again?). As you rightly point out, the liklihood is that yes, there is enough food for 10,000 men there, and it could be foraged for in the first month unless its seasonally dependent. However, this is basically mortgaging current attrition levels against future ones - excess foraging in early months would make foraging harder in later months with nearby and poorly defended areas exhausted.
Obviously, my example was supposed to be simple...the minute you move away from simple constants then you walk into a minefield. I'm not trying to perfectly model attrition. I was just giving a working mathematical example of why attrition is non-linear, based on one single factor (food supply). I thought it was only Akka that was disagreeing on that point?
Akka - Byzantium is a rather unique example, given its rather vast wealth, and the Arabic countries are non-playable. This is pretty much what I was thinking, that the possibility to support was there, but more of a theoretical possibility than an actual practice, and only the amazingly wealthy countries could afford to tinker with the idea. Given that the superpowers in CK don't need any more of a leg-up as it is, I'd go with not implementing the slider...putting it in would only serve to add an exploit for the player - and as you say yourself you'd rather not use exploits, there's a good reason not to do it.
As for events affecting the AI, the AI already needs all the help it can get. Speaking personally, I run rings round it 9 times out of 10, and the times I don't are normally blatant errors on my part rather than insightful strategy by the AI. The AI has little concept of how to fight a war properly in CK, so giving them events that destroy their already ineptly commanded regiments would just be kicking them while they're down.
Returning to attrition curves, even in the examples you just gave the relationship is still curved. I was just using an arbitrary threshold to show the curved relationship. However, demolishing the threshold doesn't remove the curve. What you've described by rations being reduced further and further and further is just the curve at the lower part of the scale, where the mass death and desertion is the higher part of the scale. My example wasn't "Everyone is fine until everyone dies" but "Those who can survive can only just survive...everyone else will die".
Remember that except in your own territory, attrition always runs at either 1% or 2%, so attrition already starts earlier than the threshold (which is correct). However, the curve is flattened until the army crosses the support threshold of the province, at which point attrition stops being linear background noise and moves into non-linear attrition which is more notable, but flattened at the top with a rate cap to prevent extreme rates of attrition. The idea is to try to reflect the reality of attrition without actually trying to model it.
If you want to talk in terms of marginal cost of attrition, then I'm happy to work with those figures. Personally, I'm not trying to come up with anything sophisticated. I'm just trying to draw up with something simple to implement and test which makes attrition follow more realistic levels and, if your initial assumption is right, will make the game more fun.
Woz Early said:
From what I've read, burning fields was a universal tactic of the invading medeival army--in fact, the pre-eminent tactic, as a way of making your opponents suffer without taking on much risk to yourself. The defenders were not all that interested in destroying their next year's wealth.
The problem with assuming that having 10% more people means that you have the potential for 10% more forage (or 8% more forage, or however you want to determine efficiency) is that you're assuming that the province has an effectively infinite supply of food that just needs to be harvested. On that point I disagree, but it doesn't really matter.
It assumes finite resources that get progressively harder to gather, and is pretty much the stereotypical economic resource model.
But my arguing for a shallower curve is based on primarily on gameplay and "feel," not a passion for modelling production possibilities curves. Having an army that suffers 1% attrition one month, then marches to Adrianople and suddenly suffers 15% (Akka's example, but one I can reproduce in my game) changes attrition from a military challenge, to a boring pathfinding problem.
For me, as I've said, losing about half an army over four to six months of campaigning "feels" about right, and as a broad generalization is at least as realistic as any other. Tamping down on this a bit to make up for the ahistorically long campaigns and cross-country marches CK design encourages is reasonable.
Keeping the curve the way it is now simply makes all armies approach about the limit (say 15,000 men) quickly, but lets armies at that size survive essentially indefinitely. This cuts out the sense of urgency that led to many stupid feudal engagments--armies would not exist forever, and therefore needed to be used. This built-in lack of patience contributed to classic defeats at places like Najera, Nicephorus and Agincourt.