Inn Unit Capacity

[Demon of Light]

This question is from Misplacedag
He didn't ask that it be put here but I thought I'd do it anyways since I noticed he was new and 9 out of 10 times that means he WISHES he could have put it here.
----------------------------------------------------
Wes chose to limit Mercenary Magnetism to 20 in his mod. I agree with his reason why, but I notice that when some Inns do attract Mercs, they can get a very large number of them, so one can still buy hugh Merc armies when they become available.

I was wondering if anyone has experimented with limiting Unit Capacity in the last field? It seems to me if Inn capacity was limited to 500 or less, then Magnetism could go well over 40, but one would still accomplish Wes's objective.

I did a search with Inn in the title only and zero hits. With it in the topic or title I received a hugh list which I tried to scan, but would take hours to read all.

Thanks.

[WesW]

I'm afraid you misunderstood what capacity referred to- it refers to the number of troops that can retreat inside a castle structure, not the number that show up on the review screen. And, remember that mercs cost twice as much to support as regular troops.

[Lord Of Storms]

Wes, I think what Misplacedag was trying to say is can you limit the Inn's capacity of mercs by editing the last field in the Crusader_build_prod_13 file ,last column entry unit capacity integer, as there is no default entry used I do not know if it is possible to edit, this would require testing but seeing as Wes accomplished this in his mod by reducing Merc.Magnetism integer from its default setting of 40 to 20 he achieved the same thing, so really if you want less mercs lower the integer to 10 or 5. http://www.totalwar.org/forum/non-cgi/emoticons/biggrin.gif

[MonkeyMan]

misplacedag

From the EntranceHall


Quote[/b] ]Wes is correct, I misinterpreted what "Unit Capacity" means in the field descriptors at the top of the unitprod file. I was assuming it would work for every building in the same manner it works for castle types. However, after testing, it does not appear to, so I must assume Wes is correct in saying that castle types are the only building to which the field applies.

The Lord of Storms is also correct in interpreting what I was trying to accomplish. Even with field 25, Mercenary Magnetism set to 20, the Inn can still attract a whole hoard of Mercs, (at least a stack and a half worth in the first 3 or 4 years). However, it has never made in logical sense to me that some Inns can attract 1500+ mercs, (does any castle ability go this high?) while the vast majority of Inns have none. What I really wanted to accomplish was having all Inns attract a few units, but none being able to attract almost 2 stacks worth.

I have tested field 25 with values of 01, 20, 40, and 90. I could see no statistical difference between 20, 40 and 90. Only significant dropoff came with 01, and then VERY few inns seemed to attract any at all, but I really don't have enough tests to make that a definite conclusion.

[WesW]

I had the Magnetism set to 10 or 15 back when the mod was in earlier development, and there is definitely a change as you adjust the settings.
The biggest factor in attracting mercs to a given Inn is war and conflict in nearby provinces. Also, castle structures use the magnetism column to cut down its value as you get bigger castles. Therefore provinces that are poorly developed attract more mercs.
I have eliminated most all elite units from the merc list, so while a lot of units show up, they are primarily light cav, missile, and spears. Therefore, they are not a very good way to hire an attacking army.

[MonkeyMan]

misplacedag

From the EntranceHall


[QUOTE=Quote ]I figured the drop off was somewhere between 01 and 20, just didn't feel like running the countless tests to determine exactly where.

"The biggest factor in attracting mercs to a given Inn is war and conflict in nearby provinces."

Do you have inside knowledge about the code to verify this. For testing, I created an Inn at the start for all factions, and then would start a new game for each region. No wars had broken out yet, and I only tested Venice and Wessex, but 2 - 5 units would appear in year 2, then anywhere from 6 - 15 in years 3 and 4. Those numbers stayed consistent at 20, 40 and 90. The reduction to 01, clearly reduced the numbers to 0 in year 2 and 3 - 5 in subsequent years. Plus, it seems that in the games I've played, certain regions were always Merc havens, whether there were wars in the region or not. Khazar, Arabia, Lithuania, Flanders come to mind. Inns in adjacent regions would have none. Of course Flanders always has a war around it, but the first two (especially Arabia) didn't.

Yea, I noticed the negative values for castles and a couple of other buildings (Constable and Marshal if I remember right). That's the main reason I tested at 90. I wanted a number much greater than 40 that should never be reduced to zero or less. (Note: I had noticed that other factors appear to be a triggering threshold level, rather than an actual amount level, and it looks to me like the Merc Mag field is as well. My guess is that its a 3 prong branch: Sub1>trigger, Sub20, None

[WesW]

Well, my opinions are based upon my time playing the game, and not on any formal tests, so if you have found some differences then I believe you. I do think that unit appearance is so random that it's best to just reduce the quality of troops eligible to appear and leave it at that.

[misplacedag]

I was wrong about Arabia being a merc haven. You are correct that being close to a war zone is a significant factor. I want to run a couple of more tests and format my results better, but I think the rule will become being adjacent to a country at war is the most significant factor. However, there still seems to be a third factor at play which I haven't figured out yet.

I fully agree with you that since we can't change the program logic, limiting quality is the best we can do right now. It still seems to me that it would have been easy for them to have used the unit capacity field to hold down the totals, especially if the MM field is nothing but a threshold switch.

In fact, if I were doing the programming, I would seriously consider making Inns a multi-level structure as well. i.e. Inn1 - one unit of Infrantry (Arc, Spr or Inf). Inn2 - One light Cav or two Inf. Etc.

[misplacedag]

Test 1. Started with MedMod 1.84, Early, Expert. Mods to Early.txt: Added Syria and Edessa to Egyptian owned regions, created Inns in every Egyptian region. Set Egypt treasury to 100000 for all levels. Only moves I would make each year were to move armies to all borders so I shouldn’t get attacked and to insure loyalty of each stack to prevent revolts. Every 3 or 4 years I would fill up the training stacks. I would hire all Mercs when becoming available. (Not sure if hiring Mercs helps keeps the flow coming, but its an untested opinion of mine that it might.)

1088 Byz/Tur and GRE/Itl wars started. 1089 Mercs started showing up and did so for every year thereafter. I ran the sim thru 1107 when I ran out of money. Frn & Eng went to war with GRE a couple of years later. Itl also started taking on the Pope and Sil in 1091. A couple of cease fires were made, must most of the above wars lasted the rest of the sim. In 1098, the Almos went to war with both Spa and Arg. Both wars lasted the rest of the way.

My opinion of Arabia being a Merc haven was false, neither it nor Sinai got a single Merc.

“Close to the War Zone as a factor”. Egypt, clearly closest to the Almo/Spa,Arg wars for 10 years never received a single Merc either. Antioch which was clearly closest to any battles only received 2 Mercs, one year. Syria and Tripoli were the first to receive any Mercs (2 each in 1089). Each would get 1 to 4 every 3 or 4 years thereafter, but 1089 was the only year they both got them the same year. Syria wound up getting a total of 11, Tripoli a total of 10. Edessa was the only big winner. Starting in 1091 and in every year but one thereafter, it received anywhere from 2 to 9 Mercs winding up with a total of 69. The other odd-ball was Palestine which received 2 one year. It’s the Syria and Tripoli results that makes me think there might be another factor involved.

Possible conclusion, but not statistically significant yet is that the closest region that shares a border with a country at war will be a Merc haven. Lesser Armenia and Libya remained in Rebel hands throughout the test so Egypt and Antioch never shared a border with a country at war. Changes I plan for test 2 are to give Lesser Armenia to the Turks, all of N.Af. to the Almos and up Egypt’s starting treasury.

[cugel]

Why worry about it? Just do what I do and don't build the inn (or at least don't ever hire Mercs). Since the AI doesn't hire them it's really only for the player's advantage. You really don't need them anyway Plus, since (post patch) you can't combine merc units, after 1 or 2 battles they tend to be severely understrength, which limits their usefulness.

[misplacedag]

I know. It's just that the early games I played I liked to hire Mercs to do the fighting and let them be the ones that got killed, while I used my own people to protect my borders. Silly, but sometimes hard to get over a given philosophy.

[WesW]

This is getting off topic, but I have seen units in AI armies before that surely had to be mercs. I don't recall any specific instances, but I have witnessed it several times.

[misplacedag]

Funny you should mention it. In testing my current CityState mod, I just found what had to be some Mercs in an Independent army. By independent, I mean, its year 3, and the region has never belonged to a real faction and never had a revolt.

I'm using the MedMod unit and prod files unchanged. Only changes I have made are to the Early.txt file.