Accurate army size & location?

[lampros69]

I studied most of the data on SSHIP's descr.strat file, and I was impressed as to how superior it was over the vanilla SS in terms of accuracy of settlement size. (Further, as someone who is further modifying SSHIP's descr. strat file, it was a useful for me that you went on to explain how you arrived at those numbers.)

Nonetheless, I also wondered whether you guys attempted to be accurate in regard to the size and location of each faction's military as well. And if such an attempt was made, what sources did you look at? Obviously, there is a ready and easily accessible way to determine settlement size numbers via the OpenHistory site, as well as various Wikipedia entries on historical world/European settlement sizes. But it's harder to nail the troop situation with any modicum of accuracy, as that information seems a bit harder to come by at least on my first online foray. (And I was also surprised at the numbers I did see; for instance, I knew medieval military was smaller, but I was shocked that there may have been no more than 50,000 troops defending the Byzantine empire even at the height of the Komnenian resurgence! Perhaps even more incredible, even the Ottomans appear to have had less than 100k men under arms around the time of the fall of Constantinople.) So any direction to which you can point would be very helpful - if you did make an earnest effort here as well!

[MWY]

Thanks for your compliments mate!

I think that in this matter, the gameplay would break too hard if you tried to get actual historical numbers. I mean 50k troops, that would be atleast 20 full stacks! That would mean pure chaos on the campaign map.
So I don't think it's possible to represent the military numbers historically sadly. We focused on reducing stack spam, making units more expensive but therefore decreasing numbers. You can still field around 10-20k units as the byzantine empire by the way, depending on the units you choose. So I think it's perhaps around half the historical numbers.

[lampros69]

I think that in this matter, the gameplay would break too hard if you tried to get actual historical numbers. I mean 50k troops, that would be atleast 20 full stacks!

Well, at the time when SS starts, Alexios' military was a lot smaller, since the Komnenian resurgence was just starting. So apparently around 1100 the Byzantine military was supposedly around 20k men. This is probably manageable, especially since I've modded the game so that all infantry units are 250 men. The bigger difficulty seems to be paying for these troops: For instance, there were apparently 5000 Varangians under Alexios and paying that and other troops would be impossible even with Byzantine's massive early economy!

I was unclear about what you meant when you said "the game would break" and that there would be "pure chaos on the campaign map." Do you mean that there would be too many battles to fight on the campaign map? (I did somewhat address by scripting AI to only attack if it has a clear advantage.) Or did you mean that the computer performance would slow too much with that many stacks? (Relatedly, I am trying to figure out if it's possible to mod the game so that the AI moves in larger stacks).

Finally, you did not answer my question on the other thread: If I wanted to simply move SSHIP's money script to SS vanilla, would simply copying & pasting it (with the faction names changed) do the trick? Or do I have to copy & paste other files or scripts as well?

Thanks in advance!

[Larkin]

I was unclear about what you meant when you said "the game would break" and that there would be "pure chaos on the campaign map." Do you mean that there would be too many battles to fight on the campaign map? (I did somewhat address by scripting AI to only attack if it has a clear advantage.) Or did you mean that the computer performance would slow too much with that many stacks? (Relatedly, I am trying to figure out if it's possible to mod the game so that the AI moves in larger stacks).

Both. If it were made so that units were both of large size and cheap to build/upkeep, the AI would do nothing except make stack after stack after stack. You see this in SS vanilla in the late late game - the faction that dominates its starting area will have dozens of full stacks just sitting around.

In SSHIP, the AI is better, smarter. So these stacks would be put to use. And with the extra factions, that's a lot more of them. A hotly contested area like Iberia, for example, you could see literally 50 stacks next to one another. Which may be historically accurate for a large military campaign with tens of thousands of men, but it's far more than the engine is built to deal with, so they'd bug out around one another, not knowing what to do, where to attack, etc.

Plus it would greatly increase turn length as there'd be dozens of AI stacks moving around (relatively) pointlessly. It would probably drain on performance a bit too.

[lampros69]

Both. If it were made so that units were both of large size and cheap to build/upkeep, the AI would do nothing except make stack after stack after stack. You see this in SS vanilla in the late late game - the faction that dominates its starting area will have dozens of full stacks just sitting around.

Actually, I am still not persuaded that there would be as many stacks running around as you and MWY envision with historically realistic troop strengths - provided that some of the adjustments I made to the game are also implemented (primarily more men per unit).

Let's run some actual numbers. With 250 men per infantry unit and 50 men per cavalry/artillery, a full, balanced stack would be about 4000 men - which is coincidentally exactly how large a standard Byzantine tagma was (a tagma was composed of exactly 20 bandons as well, so both tagma and bandon translate flawlessly into the "stack" and "unit" parlance of the game). Now, Alexios' Byzantine army was around 20,000 men at around 1100, so we'd need around 5 full stacks or 100 units for historically accurate Byzantine's start military - which is merely 37 more units than SSHIP's start number for Byzantine units. So we are not talking about exponentially higher number of stacks/units. Overall, I suspect that at most some factions would have about twice larger number of starting units than SSHIP's vanilla numbers. (I suspect the Byzantine numbers are close to the max for historical starting faction troop strengths at 1100, given that even the numbers I looked at toward the end of the medieval era for some of the larger states do not exceed this by much: e.g. Holy Roman Empire at around 1500 had 24,000.)

Nonetheless, it may be the case that even somewhere less than 100 percent more units would create such a logjam that you guys fear. I simply don't know, as I have not really experienced the end game for Medieval II, either vanilla or SS. The latest I have gotten to is around 1250. Instead, I am merely trying to emphasize that the number of stacks and units running around would not be exponentially higher as you guys seem to think.

In SSHIP, the AI is better, smarter. So these stacks would be put to use. And with the extra factions, that's a lot more of them. A hotly contested area like Iberia, for example, you could see literally 50 stacks next to one another. Which may be historically accurate for a large military campaign with tens of thousands of men, but it's far more than the engine is built to deal with, so they'd bug out around one another, not knowing what to do, where to attack, etc.

Could you elaborate how SSHIP's AI is superior to SS vanilla? MWY has explained how the money script is superior, but I did not realize that AI is generally superior as well. My understanding was that SSHIP relies on the same Germanicus [sp.] battle AI and Lusted/Gracul/Savage campaign AI that SS uses. Am I incorrect?

[Larkin]

Actually, I am still not persuaded that there would be as many stacks running around as you and MWY envision with historically realistic troop strengths - provided that some of the adjustments I made to the game are also implemented (primarily more men per unit).

Let's run some actual numbers. With 250 men per infantry unit and 50 men per cavalry/artillery, a full, balanced stack would be about 4000 men - which is coincidentally exactly how large a standard Byzantine tagma was (a tagma was composed of exactly 20 bandons as well, so both tagma and bandon translate flawlessly into the "stack" and "unit" parlance of the game). Now, Alexios' Byzantine army was around 20,000 men at around 1100, so we'd need around 5 full stacks or 100 units for historically accurate Byzantine's start military - which is merely 37 more units than SSHIP's start number for Byzantine units. So we are not talking about exponentially higher number of stacks/units. Overall, I suspect that at most some factions would have about twice larger number of starting units than SSHIP's vanilla numbers. (I suspect the Byzantine numbers are close to the max for historical starting faction troop strengths at 1100, given that even the numbers I looked at toward the end of the medieval era for some of the larger states do not exceed this by much: e.g. Holy Roman Empire at around 1500 had 24,000.)

Nonetheless, it may be the case that even somewhere less than 100 percent more units would create such a logjam that you guys fear. I simply don't know, as I have not really experienced the end game for Medieval II, either vanilla or SS. The latest I have gotten to is around 1250. Instead, I am merely trying to emphasize that the number of stacks and units running around would not be exponentially higher as you guys seem to think.

In the beginning, it wouldn't be too problematic in terms of realism. However, you'd see the ERE simply running over their neighbours with an army of that size. They're usually pretty dominating anyhow, unless they're affected by the player. You have game balance to consider as well as realism. What's the point in having Serbia as a faction if it gets rolled over by the ERE 95% of the time?

I think you need to see what late game looks like. Once you pass 450 turns you'll see what I mean, when superpowers exist and can churn out thousands of units. SSHIP alleviates this problem - the 'total war' that you get at end stage SS vanilla is highly unrealistic, as it is just factions with huge banks making 2-3 full stacks of units every couple turns. Not to mention that these stacks are often fully profiessional units, which is unrealistic on a different level.

Could you elaborate how SSHIP's AI is superior to SS vanilla? MWY has explained how the money script is superior, but I did not realize that AI is generally superior as well. My understanding was that SSHIP relies on the same Germanicus [sp.] battle AI and Lusted/Gracul/Savage campaign AI that SS uses. Am I incorrect?

I'm sure MWY or Lifth has something somewhere about why/how the AI is different, but yes, you are incorrect. It isn't just the same campaign AI as SS vanilla.

[Lifthrasir]

As MWY is gthe one who made the AI, I'll let him answer to that question.

Regarding the army size, we have to deal with 2 factors: the game balance and the historical accuracy.
Taking ERE as an example is not right. What about France who relied mostly on the Feudal system? What about England who relied on both, Feudal system and a important use of mercenaries?
If we manage to represent ERE as you said, I'm pretty sure that the AI wouldn't be able to handle it in a proper way. I'm convinced the the gameplay would suffer significantly.

[MWY]

Pretty much what larkin said concerning the stacks.

For the unit sizes, the thing that bugs me the most is actually the siege battles. Unit placement for walls as well as the choke point at the gate are issues here. Field battles should be fine with a higher amount of units I guess. Still not sure if it's worth it.

Yes, you could use the money script from SSHIP. But you would still get the AI boni described in the EDB money-wise. And you would have to adjust the increase of the kingspurse in the campaignscript aswell.

Concerning the AI: Well, it's a completely new one. I basically rebuild everything from scratch and tried to made the AI act as close to a human as possible.
Short notes:
- less alliances, no huge blocks which last forever
- Naval invasions working.
- Better allocation of troops, not only on one frontier.
- Hopefully smarter invasion choices.
- Somewhat active rebels.
- Reworked global standing. (it actually has a purpose now)
- Reworked faction standings
- Better Faction-Balance

There are some bugs with stuck armies in foreign territories after a crusade, some issues with reinforcements being put to far away to actually reinforce armies, or factions like england or the almoravids having problems holding both sides of their empire, but thats it I guess..

[lampros69]

In the beginning, it wouldn't be too problematic in terms of realism. However, you'd see the ERE simply running over their neighbours with an army of that size. They're usually pretty dominating anyhow, unless they're affected by the player. You have game balance to consider as well as realism. What's the point in having Serbia as a faction if it gets rolled over by the ERE 95% of the time?

Provided that information is available out there, we can scale up other factions' armies as well. So perhaps the perceived Byzantine dominance won't occur. Besides, if historical accuracy in the end turns out too damaging for game balance, things can be adjusted later. As I said before, there is no reason to simply assume the worst case scenario and give up when we are likely talking about 50-75 percent more units/stacks, not 500-700 percent more. But I guess ultimately we have to agree to disagree. In the meantime I will experiment and see if it works.

I think you need to see what late game looks like. Once you pass 450 turns you'll see what I mean, when superpowers exist and can churn out thousands of units. SSHIP alleviates this problem - the 'total war' that you get at end stage SS vanilla is highly unrealistic, as it is just factions with huge banks making 2-3 full stacks of units every couple turns. Not to mention that these stacks are often fully profiessional units, which is unrealistic on a different level.

A fair point; overall, I do not know the game as well as you or other folks who are skeptical of my proposal on this thread. But yet again, it's not as if there are not ways to address potential side effects. For instance, I tend to think unit recruitment pool replenishment is too rapid in this game, both vanilla and SS. So that can be adjusted as well to alleviate late-game unit spam. And other potential solutions abound.

Regarding the army size, we have to deal with 2 factors: the game balance and the historical accuracy.

Of course; but I assume that already.

Taking ERE as an example is not right. What about France who relied mostly on the Feudal system? What about England who relied on both, Feudal system and a important use of mercenaries?

Determining manpower for feudal armies might be more difficult, and there may not be any information out there. But it is still worth a try, and those numbers could also turn out to be "balanced" for the purpose of the game as well. It's too premature to conclude that the numbers you will find - if you find them - will prove to be unwieldy, I think.

If we manage to represent ERE as you said, I'm pretty sure that the AI wouldn't be able to handle it in a proper way. I'm convinced the the gameplay would suffer significantly.

I responded to this objection already, assuming you simply agree with Larkin's rationale for this particular objection.

Pretty much what larkin said concerning the stacks.

For the unit sizes, the thing that bugs me the most is actually the siege battles. Unit placement for walls as well as the choke point at the gate are issues here. Field battles should be fine with a higher amount of units I guess. Still not sure if it's worth it.

Yeah, I've found sieges to be problematic as well in general. As for the particulars, I don't find unit placement in the walls to be so bad, unless you try to squeeze your unit into smaller parapets above the gate. It's the narrow areas, as you said, that pose the biggest hurdles. But this too can be somewhat managed by simply moving units rapidly during the actual combat phase, as you simply have to get troops in place before the gate breaks.

Also, I am going to try to increase the threshhold for the AI to siege to possibly as high as 2:1 to eliminate frivolous sieges.

Yes, you could use the money script from SSHIP. But you would still get the AI boni described in the EDB money-wise.

I am confused; what is EDB? And what kind of boni?

Concerning the AI: Well, it's a completely new one. I basically rebuild everything from scratch and tried to made the AI act as close to a human as possible.
Short notes:
- less alliances, no huge blocks which last forever
- Naval invasions working.
- Better allocation of troops, not only on one frontier.
- Hopefully smarter invasion choices.
- Somewhat active rebels.
- Reworked global standing. (it actually has a purpose now)
- Reworked faction standings
- Better Faction-Balance

There are some bugs with stuck armies in foreign territories after a crusade, some issues with reinforcements being put to far away to actually reinforce armies, or factions like england or the almoravids having problems holding both sides of their empire, but thats it I guess..

Ah, ok. Thanks for the detailed list. And these should all be found in the campaign script file?

[MWY]

EDB is the export_description_buildings file. There, the AI gets trade/money boni from buildings per settlement.

No, thats not the campaign script thats the descr_campaign_ai_db file mostly. Other files affected are descr_campaign_db, battle_config, config_ai_battle, descr_faction_standing, descr_sm_factions, etc.

Unit replenishment rates are cut down pretty heavily in SSHIP. Perhaps not until the point of accuracy, but until the point were users would complain that they have to wait 20 turns for a unit.

Perhaps I will try how more stacks would actually change the game. But it's always a difficult path between buffing empires or small factions. If you only change f.e. the income in the descr_settlement_mechanics file, then small factions would be overrun. If you increase the kings purse, big factions would get overrun.. Right now we have a pretty good balance I'd say, and that's not that easy to keep..

[Gercog]

And when we can see all this splendor?