Hard Empirical Data On The Differences Between Campaign Difficulties [Updated 9/11/13]

[kujirakira]

One thing that annoys me to no end on Legendary is that vassal states won't automatically accept trade. I think that should come with being my underling, "Unequal Treaties" and all. I made the mistake of settling peace w/Rhodes on the condition that they'd become my es ... and now I'm stuck missing out on 1 settlement of the Asia region, no trade, and according to the tooltip they give me 0 gold every turn on top of it all. It's a travesty. What's worse is they're sitting there with 2 stacks and I can't really afford to divert 2 of my own stacks from the front lines.
Other than that, it's not that bad... just a lot of ring-around-the-rosie. I think I captured Athens 5 times in the span of ~8 turns. First Athenians, then Epirus, then Macedon, then some Thracians (Tylis IIRC?) ... if it's not one faction, it's another.
Amazingly the two Illyrian factions have maintained an alliance and non-aggression pact for almost the entire duration of the game. And despite having 1 settlement in Italy, Rome seems entirely too distracted with Carthage and Libya to even notice (just went there to finish Epirus off once and for all -- didn't care 1 iota about the city and abandoned it on the same turn - yay forced march)

Edit: the other thing that annoys me about Legendary is that I can't fast-forward during battles.
After taking all the Capadocian cities, their leftover 3-unit Army decided to make a suicide attack against a 7-man garrison from the sea -- but they glitched out on the landing, halfway in the boats halfway out and un-targetable.
I had to leave the game running for 40min or something or forfeit my city to scrubs over a glitch (;><)
I get why we can't slow-motion or pause -- but see no reason 2x and 3x speed are disabled now. It worked fine in S2.

[Silven]

Bumping this thread again since modders now seem to have a considerable amount of access to the games files and values. Would anyone be kind enough to research this? I'd do it myself if I had any idea what to look for in the PFM.

[AusHaching]

Even on "Hard", the CAI must get a lot in terms of bonus income. Most minors with 1 region in my games are able to field 3 armies with up to 46 units (Treverorum). If they have access to ports, you can add two fleets with an additional 10 Units. Also, the building in their settlements are on at least the samt technological level as mine. I´d guess that a minor, 1-region faction gets at least 2000 - 3000 gold per turn (46 Units times 60 gold upkeep - and that´s a low assumption would already mean ca. 2760 gold per turn). Or maybe they get trade agreements with dozens of other factions.

[Antonius_Marcus]

I'd be interested in learning the differences between difficulties as well. Didn't earlier TW games have two seperate sliders for battle and campaign difficulty? Are they mashed into a single slider now?

[Medina]

You are looking for hard statistical data, not empirical, thats the data you have described above your question

[Silven]

You are looking for hard statistical data, not empirical, thats the data you have described above your question

em·pir·i·cal
emˈpirikəl/
adjective
adjective: empirical
1.
based on, concerned with, or verifiable by observation or experience rather than theory or pure logic.


No, I seek empirical data. Statistical information would just be a form of empirical data, which again, is what I seek. But I'm really not interested in arguing semantics.

I'd really just like to know what changes.

[Medina]

em·pir·i·cal
emˈpirikəl/
adjective
adjective: empirical
1.
based on, concerned with, or verifiable by observation or experience rather than theory or pure logic.


No, I seek empirical data. Statistical information would just be a form of empirical data, which again, is what I seek. But I'm really not interested in arguing semantics.

I'd really just like to know what changes.

The empirical data may be incorrect because the observation may or may not be completely correct. It would be alot easier if CA would just release the ins and outs.

[crzyrndm]

Since the forum wants to destroy the formatting this morning, this'll have to do for now: https://docs.google.com/document/d/1...it?usp=sharing

[Silven]

Since the forum wants to destroy the formatting this morning, this'll have to do for now: https://docs.google.com/document/d/1...it?usp=sharing

Wow! This is great Crzy, very nice work. Exactly what I was looking for, but you said there are inputs and values that are missing? Is there still more data to uncover? And when you say that the battle_variables is a place holder, do you mean to say that the db table hasnt been unlocked yet? Or that battle difficulty has no variables??? Surely this isn't the case?

The agent action inversion is definitely a puzzling one. I could see CA making actions more likely for the AI and less likely for the player as difficulty increases, but since the variables change parallel to each other for both the AI and the player, it could honestly go either way as far as inversion goes.

I'm going to try and format your findings and edit them into the first post of the thread. Well done mate!

[Silven]

I created a nice neat table using the advanced editor on TWC, but for some reason it won't actually display the table in the thread post. So I had to screenshot it and image host it. I hope the bandwidth doesn't die.

Anyway, crzyrndm I noted in my post that one of the biggest missing values is the diplomacy modifier. Any chance you could dig this up?

[crzyrndm]

When I said there were more values in the table, I meant there was an awful lot to do with siege deployables (18 entries per difficulty, for both AI and player, total of 180. These appear to control how many of each engine you get when besieging a walled settlement (so 4 ladders per ladder you built for example). None of these change with difficulty from a quick glance.

It basically amounts to: 2 rams, 4 galleries, 4 ladders, 2 siege towers, 1 tortoise (applies to all variations of each, light/med/heavy)

Battle_difficulty_modifiers is there but not used in the current build (it didn't exist for Shogun 2 either). From experiments I did with Shogun 2, it is likely that battle difficulty makes minor changes to AI unit morale, head on, you still get the same result although with a few more casualties on both sides.

To my knowledge, CAI vs. player aggression difficulty changes are handled internally. That is to say I can find no direct reference to it.

PS
This might be what you were after

[Silven]

When I said there were more values in the table, I meant there was an awful lot to do with siege deployables (18 entries per difficulty, for both AI and player, total of 180. These appear to control how many of each engine you get when besieging a walled settlement (so 4 ladders per ladder you built for example). None of these change with difficulty from a quick glance.

It basically amounts to: 2 rams, 4 galleries, 4 ladders, 2 siege towers, 1 tortoise (applies to all variations of each, light/med/heavy)

Battle_difficulty_modifiers is there but not used in the current build (it didn't exist for Shogun 2 either). From experiments I did with Shogun 2, it is likely that battle difficulty makes minor changes to AI unit morale, head on, you still get the same result although with a few more casualties on both sides.

To my knowledge, CAI vs. player aggression difficulty changes are handled internally. That is to say I can find no direct reference to it.

So basically what you're telling me, is that all variables that can be found, have been found? Is it possible that these values that you can't seem to find just haven't had their db tables unlocked and schema'd yet?

Did you not find any information for the Easy setting? True, hardly anyone plays on it, but it would be interesting information to have nonetheless.

And yes, that tables looks good. Pretty much the same layout, but much easier to read. (I should have just opened excel but I was feeling lazy and hoping I could put up a table that wouldn't require image hosting.)

By the way, in the table you made you swapped the players legendary values for "agent crit" and "agent to agent". They are in the incorrect places based off the sheet you originally posted.

[crzyrndm]

The empirical data may be incorrect because the observation may or may not be completely correct. It would be alot easier if CA would just release the ins and outs.

Observations? I'm pulling the controlling variables from the database of the game.

So basically what you're telling me, is that all variables that can be found, have been found? Is it possible that these values that you can't seem to find just haven't had their db tables unlocked and schema'd yet?

There is probably a few more variables to discover somewhere (the db is rather large), but I can find no more direct references to difficulty.

---

So I finally established what was going on with the agent chances. Some of the entries had been tagged as prologue only which makes that part much simpler.

Here's the updated table, adding easy difficulty, fixed agent rubbish, labels now more sensible (because having an AI column with player data at the bottom was bugging me).
There are two extra entries which only apply to the player on easy (hence why I didn't include them directly), these are: -30 resistance to occupation, and -30% corruption

[Silven]

Well done mate. I will update the first post again. Be sure to return with your findings if you discover any more, namely the diplomacy variables.

[auboy]

Well done mate. I will update the first post again. Be sure to return with your findings if you discover any more, namely the diplomacy variables.

This is great! Just what Ive been looking for!

Could we get the same thing for battle difficulty level?

Also maybe some tips where we can mod these variables in pak file manager?

[crzyrndm]

This is great! Just what Ive been looking for!

Could we get the same thing for battle difficulty level?

Also maybe some tips where we can mod these variables in pak file manager?

Battle difficulty just gives the AI small morale bonuses, and maybe changes a few variable relating to which tactics it will use in a given situation (assuming it actually works ofc ). There is a table which modders may be able to use for bonuses/maluses in battle, but in the official version it is only a placeholder and without the TWeak showing the table linking or some experimentation, I don't know what it can do yet.

If you want to change the difficulty bonuses/maluses

Campaign_difficulty_handicap_effects is the super huge one. It seems that the main reason for the increased size of the table is something to do with siege deployables, which I will be ignoring for now.
For anyone who wants to look into this themselves, the difficulties correspond to numbers as follows for the player: 1 (easy), 0 (normal), -1 (hard), -2 (V. hard), -3 (Legendary). The AI bonuses will be listed under <human difficulty> * -1 (there is a boolean column to tell whether this is a human or AI entry)

[auboy]

Battle difficulty just gives the AI small morale bonuses, and maybe changes a few variable relating to which tactics it will use in a given situation (assuming it actually works ofc ). There is a table which modders may be able to use for bonuses/maluses in battle, but in the official version it is only a placeholder and without the TWeak showing the table linking or some experimentation, I don't know what it can do yet.

If you want to change the difficulty bonuses/maluses

Thanks for the tips! I guess we will have to wait to mod battle difficulty?

Do you know what civil economy tax mod or the trade route sea cap does in the campaign difficulty?

[crzyrndm]

tax mod is the bonus tax rate percent (value of +20 means 20% extra tax rate). Trade route sea cap is not used at the value its at, but if you made it lower (1/2/3), my best guess is that would be how many trade routes each of your ports could handle.

Battle difficulty I might play with and see what options are viable. Likely the only things you can do are change unit stats which isn't particularly helpful.

[auboy]

tax mod is the bonus tax rate percent (value of +20 means 20% extra tax rate). Trade route sea cap is not used at the value its at, but if you made it lower (1/2/3), my best guess is that would be how many trade routes each of your ports could handle.

Battle difficulty I might play with and see what options are viable. Likely the only things you can do are change unit stats which isn't particularly helpful.

Thanks this is very helpful!

[Jackal2233]

So this explains why the AI is still unaggressive diplomatically on Legendary... no chances to their AI