Historical Accuracy vs Historical Precision

[LlywelynII]

Now if you guys like, you can simply borrow my Persian formation events in EP and then adapt them and de-AI cheat them to AGCEEP.

Who cares about AI cheats when they're historically justified but too much of a gimme for players?
jay.

 

[unmerged(17856)]

The break alliance command doesn't work and it never will. The days of event command editing are over. We were told this not too long ago now. As has been said before, we have our tools. They aren't perfect, but they must do. Even if we adapted a top down priority, we'd still require people to take a long hard look at national and regional development across the AGCEEP globe. I am of the opinion that we don't need any more bureaucratic or administrative ideaology and rhetoric. We are in need of 'engineers' who are willing to put to practice the last year's worth of ideaology.

We only now need to make a clear definition which events are considered "regionally important" and thus should be made with all possible triggers and as little as possible outcome variations (mostly only one - historical) and which are not "regionally important" and therefore should stand as they are.

What's considered 'regionally important' is essentially an issue-specific situation to be resolved in whatever region concerns the issue being debated.

 

[unmerged(2456)]

What's considered 'regionally important' is essentially an issue-specific situation to be resolved in whatever region concerns the issue being debated.

Plus making sure it had triggers to fire as much as possible isn't good, unless the situation occured before the beginning of the game. Sometimes, it just would be too much of a stretch to have stuff fire every time, reguardless of how game-breaking it could be, but this mostly post 1520 stuff i'm talking about here.

Making sure that no matter what the situation is, you straight-jacket it so that a situation must occur is to many just as bad as letting all hell break loose and having denmark conquer constantinople.

 

[LlywelynII]

So to me some are confusing the scaffolding for the actual masonry. Backup events are scaffolding that should support the mod while we work to get things working properly. So I guess in general this A vs P approach sounds awfully similar to me with the whole ai cheat debate; which to me has always smelled as a cop-out to putting all of the work necessary to actually get something to work.

I completely agree with this - it is much better to have well-scripted event sequences, accurate triggers, and well-balanced gaming. This should be the goal towards which we all strive.

A province based sequence that leads to the formation of Persia doesn't really demonstrate an understanding of how the Safavid Empire came to be. It was specific conditions, a collapsing empire than lead to Ismail's successes. I don't see how, and this is talking purely in historical terms, how a strong centralized state would have fallen to the Safavids, seeing as how they had been causing problems for a good part of the 15th century before managing to accomplish anything.

That being said, this is a game, and as such its best that we try to keep things adhering to our timeline if possible. And as a result, one should be getting a Safavid Empire a majority of the time.

With due respect, though, none of the countries in the region were strong centralized states, with the somewhat exception of the Ottomans - and they stayed the hell out of Persia for a century because they didn't want anything to do with the Sufis. It was enough of a problem keeping the converts among the Janissaries in line. And the problem with having QAR instead of PER isn't that we all like Persia more - personally I think a buncha nomadic hooligans named after sheep is charming - it's that with no events, QAR never suffered any of the instability or limitations that Persia endures. Poly is a little rude, but his heart's in the right place.

Your events should start healing up the PER issue, but I'm all for bandaids for other major gameplay issues while we wait for someone to have the time and resourcefulness to script the proper events. At least until we have some guys under retainer who do nothing but script AGCEEP

jay.

 

[unmerged(7597)]

So I guess in general this A vs P approach sounds awfully similar to me with the whole ai cheat debate; which to me has always smelled as a cop-out to putting all of the work necessary to actually get something to work.

I mean this with every measure of respect, but you have a misplaced faith in the ability of this engine to achieve the desired results without significant guidance, Garbon. How long has EEP/AGC/AGCEEP been around - it was there when I discovered it more than three years ago. Vast amounts of work have been devoted to the project - my own contribution has been very small, and yet I know just how many hours I have spent on it.

No amount of work will achieve what we want unless it involves what you call "ai cheats", and I'd call simply necessary.

 

[unmerged(2456)]

No amount of work will achieve what we want unless it involves what you call "ai cheats", and I'd call simply necessary.

Some are okay and nessary, but wholesale lets just use ai cheats isn't, especially as long as we are giving nations unhistorically deserved cultures and cores that both human and ai can use, those same things should be usable by humans and ai, unless there is a specific reason, fe a different sequece do to game mechanics to make the same thing happen.

 

[Garbon]

Then I guess you can just call me a pollyana, as I've been here most of the past 3 years and have yet to fall into the camp of wholesale ai cheats.

 

[unmerged(31425)]

1 .For instance, in AGCEEP (and vanilla too), Spain 95% usually doesn't conquer the Aztecs and Incas. Its fine if it isn't 100% deterministic and Spain can sometimes fail or struggle but 95% struggling?

Spain cannot conquer the Incas without AI cheats. The engine is simply not designed to be able to handle that kind of a circumstance. It can conquer the Aztecs without AI cheats, but it is unlikely that it can be made to do it consistently and within a reasonable timeframe without AI cheats.

2. IRL history, Ak Koyunlu and Kara Koyunlu are both minors and Kara Koyunlu is supposed to be defeated by Ak Koyunlu then Ak Koyunlu is defeated by Persia in 1501. But in AGCEEP, 100% of the time Ak Koyunlu is destroyed, Kara Koyunlu 100% of the time takes over half of the Western Timurid Empire and becomes a "major" when that was supposed to be Persia.

This is extremely improbable without AI cheats.

3. Ottomans are nearly completely dismembered or destroyed around in the mid 18th century. Yes they are supposed to go into decline but consistently reduced to minor status all the time is a bit much. This would be historically accurate if this were the Polish-Lithuania Empire (which OTOH doesn't nearly decline and get dismembered as often as it should).

This cannot be corrected without AI cheats.

4. Prussia - 'nuff said

I'm not 100% sure on this one, as I do not know the history well, and thus don't know how event sequences could be historically justified to be arranged. Personally, I suspect that without AI cheats it's not going to happen.

1. HYW French win England loses France forms (1453)

This can be done without AI cheats but only with a certain degree of consistency, and of course a human English player can easily prevent it.

2. Rise and Fall of Ottoman Empire resembling history

There is no chance that this can happen without AI cheats. The Ottos run into way too many engine defects to be able to pull this off without AI cheats pulling him out of the rut situations.

3. Spain forms and by 1550 or so should have colonized/conquered Mexico and Peru

Spain cannot colonize well without being formed(meaning Aragon AND Castile are one country not one as a vassal of the other)about the time or slightly earlier than Columbus' appearance. In addition to this the current AGCEEP Spanish DP settings are fairly hostile to optimal colonization. With a single Spanish country and optimal DP settings, Spain can colonize fairly well, but it cannot colonize in the first half of the 16th century, with historic proportions without AI cheats, because he simply has too much ground to cover too quickly. Nonetheless, even without AI cheats he can be much improved in this area.

4. Formation of Safavid Persia starting in 1501

If it's done with a forced independence, forced province cessessions, and cash and troop stacks, it can be made to happen without AI cheats.

6. Timurid defeat/destruction -> Rise of Mughal in India starting in 1525

The Mughals cannot be made to conquer India without AI cheats. Sure it could still happen, but it would be extremely rare. Once again the engine dynamics are very hostile to this situation developing.

9, Thirty Years War (1600-1648) (including formation of Netherlands)

This can be modelled, but without AI cheats, the Dutch AI cannot perform historically. Without AI cheats it is also extremely difficult to stop human or even a large AI's exploitation of the emerging Netherlands AI.

11. English colonization of North America leading later to USA formation

The primary problem here is the DP settings of England and his horrible inflation due to too many forced wars early and cash and stab hits prior to his needing to so this. The AGCEEP English events and their hostility to the English AI are the bulk of the problem here. Asking the English AI to accomplish all of that colonizing without AI cheats, is a heavy task, but I think it could be made to happen if he has ideal circumstances leading up to that time period.

13. Partition/Destruction of Poland-Lithuania

This can likely be modelled, though it is awfully hard to make it consistent without AI cheats.

14. Napoleonic Wars

The engine is not well designed to be able to handle this. But obviously without AI cheats it can still be modelled to a limited extent.

Anyways, I've pointed some of these things out to you guys before, for example the English and Spanish colonization problems and how a good chunk of it can be solidly fixed. So far no one has done anything about those problems.

But the main point of my post here, is to point out to you the futility in even discussing fixing some problems, because they are literally impossible to fix with AI cheats. Being that some people here simply don't want AI cheats or to resort to them just yet, perhaps it is better to stick to the areas that can be fixed without them.

 

[unmerged(17856)]

Really, we can't expect much historical accuracy in regions where a human player is involved. For example, if I tried to address the ease of an ENG human player of conquering France, then AI England loses too early. If I make it harder for the Dauphin human player to smash the English, then the AI Dauphin finds it harder to form. The context of historical accuracy requires a hands-off groundwork. Place a human into a given region, and the best we can hope for is a ROTW historically accurate development.

I think IB and myself addressed ENG's DP sliders regarding colonization a while back. Perhaps those weren't implmented. I'd found that increasing the expansion values in MKJs AI files, and fine tuning the 'areas of exploration' did wonders for England's colonial efforts.

 

[unmerged(31425)]

I think IB and myself addressed ENG's DP sliders regarding colonization a while back. Perhaps those weren't implmented. I'd found that increasing the expansion values in MKJs AI files, and fine tuning the 'areas of exploration' did wonders for England's colonial efforts.

Was this before or after I started modding? Because if it was before, I can asure you that the DP settings are still plenty messed up.

 

[Garbon]

This is extremely improbable without AI cheats.

Funny that. I mean I introduced a few ceding events for everyone based on owning key provinces that triggered downfalls, and I haven't been having many issues accomplishing this.

 

[unmerged(31425)]

Funny that. I mean I introduced a few ceding events for everyone based on owning key provinces that triggered downfalls, and I haven't been having many issues accomplishing this.

So you have AKK surviving, QAR getting big early then getting beaten or assimilated by AKK and/or PER? Because if I understand the complaints correctly, AKK is getting smashed early and PER either isn't forming correctly or isn't forming at all.

 

[Garbon]

But then, all of those complaints are about the standard AGCEEP setup+events. No one has commented as of yet on my new downloadable.

 

[unmerged(31425)]

But then, all of those complaints are about the standard AGCEEP setup+events. No one has commented as of yet on my new downloadable.

How do you keep AKK alive? Doesn't Persia just get swallowed up by his neighbors, when forming?

 

[Garbon]

How do you keep AKK alive? Doesn't Persia just get swallowed up by his neighbors, when forming?

AKK is a vassal of the Timurids and often ends up allying with them. I also copied the dp settings that are in the QAR's 1419 scenario file and pasted them in the AK Koyunlu's file, which really changed the tide for battles between the two. Now more common is that the Kara Koyunlu get raped early on, but there are a few historical restoration events. AKK gets Tabriz ceded to it shortly after the death of Jahan Shah. And then if it or rebels control Hamadan (new capital of QAR), it annexes the Kara Koyunlu.

I went with forced independence for Persia on the part of the AKK, and two or so additional provinces are also ceded. I'm modeling the fall of the AKK by having it disintegrate into two states. One AKK in the north and Iraq in the south. The current weak point in the sequence is making sure the AKK owns Iraq or Kirkuk. Persia then gets most of IRA's provinces - all the way to Fars in 1503...and then in 1508, Persia annexes Iraq as its ruler flees. So sort of a forced ceding of provinces but only from certain parties. I think that I might want to make the events that give Persia, IRA be dependent on owning a province or two.

I need to work more on balance around 1500ish still as I was working mainly to get the 15th century working correctly. Hence Persia could still be falling quickly after its formation.

 

[unmerged(2456)]

Anyways, I've pointed some of these things out to you guys before, for example the English and Spanish colonization problems and how a good chunk of it can be solidly fixed. So far no one has done anything about those problems.

But the main point of my post here, is to point out to you the futility in even discussing fixing some problems, because they are literally impossible to fix without AI cheats. Being that some people here simply don't want AI cheats or to resort to them just yet, perhaps it is better to stick to the areas that can be fixed without them.

Well since we're not trying to follow history quite so stringently as EP, it not happening as often isn't a bad thing imo (several things you listed said could be done w/o such ai cheats).

3. Ottomans are nearly completely dismembered or destroyed around in the mid 18th century. Yes they are supposed to go into decline but consistently reduced to minor status all the time is a bit much. This would be historically accurate if this were the Polish-Lithuania Empire (which OTOH doesn't nearly decline and get dismembered as often as it should).

This is extremely improbable without AI cheats.

The main point is that persia should form. It would be better if it formed naturally, and their has been stuff done to remedy the problem seen, but the end result we are looking for is the emergance of Perisa in historically proportial size and location, give or take a bit.

2. Rise and Fall of Ottoman Empire resembling history

There is no chance that this can happen without AI cheats. The Ottos run into way too many engine defects to be able to pull this off without AI cheats pulling him out of the rut situations.

Not really. Too much help imo has been given to them, mosty with cores and cultures.

3. Spain forms and by 1550 or so should have colonized/conquered Mexico and Peru

Spain cannot colonize well without being formed(meaning Aragon AND Castile are one country not one as a vassal of the other)about the time or slightly earlier than Columbus' appearance. In addition to this the current AGCEEP Spanish DP settings are fairly hostile to optimal colonization. With a single Spanish country and optimal DP settings, Spain can colonize fairly well, but it cannot colonize in the first half of the 16th century, with historic proportions without AI cheats, because he simply has too much ground to cover too quickly. Nonetheless, even without AI cheats he can be much improved in this area.

I agree improvement could be done, especially with DP sliders moving more appropriate settings over time till they're more coreect at about 1485-1490, but straight-jacketing things that aragon and castile must unite is a bit much. Their colonization of the Americas to historic proportions isn't so vital that we need to resort to thus brutish methods.
[QUOTEE=polypheus]6. Timurid defeat/destruction -> Rise of Mughal in India starting in 1525

The Mughals cannot be made to conquer India without AI cheats. Sure it could still happen, but it would be extremely rare. Once again the engine dynamics are very hostile to this situation developing.

[/QUOTE]This has gotten better since it has multiple paths to form. What is mostly needed is troop stacks and possibly a war command because if its too small it'll never DoW.

11. English colonization of North America leading later to USA formation

The primary problem here is the DP settings of England and his horrible inflation due to too many forced wars early and cash and stab hits prior to his needing to so this. The AGCEEP English events and their hostility to the English AI are the bulk of the problem here. Asking the English AI to accomplish all of that colonizing without AI cheats, is a heavy task, but I think it could be made to happen if he has ideal circumstances leading up to that time period.

Actually the primary problem is not that. Its that only England can have the USA form from it really. Not much work has been done for alternate scenerios which even in the best circumstances with the best AI and AI cheats can still happen that England doesn't dominate.

 

[unmerged(7597)]

straight-jacketing things that aragon and castile must unite is a bit much. Their colonization of the Americas to historic proportions isn't so vital that we need to resort to thus brutish methods.

Would you mind expatiating on this point? You seem to be saying that you consider the union of the Spains and its colonization of the Americas something which a mod purporting to be the "most historically accurate" can do without. I'm wondering, because if you were to ask me to name the five most significant features of world history between 1400-1800, I'm certain the rise of Spain to hegemonic great-power status and its colonization of most of the New World would be on that list.

I don't mean to pick a semantic or pedantic quarrel, but there is a consensus that certain major events should happen with near certainty barring the player's intervention - and the rise of Spain is surely one of those. As I understand it, the question is how to achieve those major events - some arguing that the engine can be made to bring them about naturally, and others (myself included) arguing the AI must be led to water before it drinks. If on the other hand we're back to square one, debating whether such major events should occur at all, then I'd like to know.

 

[unmerged(31425)]

AKK is a vassal of the Timurids and often ends up allying with them. I also copied the dp settings that are in the QAR's 1419 scenario file and pasted them in the AK Koyunlu's file, which really changed the tide for battles between the two. Now more common is that the Kara Koyunlu get raped early on, but there are a few historical restoration events. AKK gets Tabriz ceded to it shortly after the death of Jahan Shah. And then if it or rebels control Hamadan (new capital of QAR), it annexes the Kara Koyunlu.

I went with forced independence for Persia on the part of the AKK, and two or so additional provinces are also ceded. I'm modeling the fall of the AKK by having it disintegrate into two states. One AKK in the north and Iraq in the south. The current weak point in the sequence is making sure the AKK owns Iraq or Kirkuk. Persia then gets most of IRA's provinces - all the way to Fars in 1503...and then in 1508, Persia annexes Iraq as its ruler flees. So sort of a forced ceding of provinces but only from certain parties. I think that I might want to make the events that give Persia, IRA be dependent on owning a province or two.

I need to work more on balance around 1500ish still as I was working mainly to get the 15th century working correctly. Hence Persia could still be falling quickly after its formation.

Well I'm a bit skeptical that this model is more efficient and consistent than the one I made, but I definitely will admit that this sounds alot better than how it is in the AGCEEP, curruntly or is it just prior to the latest release?

 

[unmerged(31425)]

Well since we're not trying to follow history quite so stringently as EP, it not happening as often isn't a bad thing imo (several things you listed said could be done w/o such ai cheats).

Often? I believe the problem with the issues raised, is that they either aren't happening at all or only seldomly.

I agree improvement could be done, especially with DP sliders moving more appropriate settings over time till they're more coreect at about 1485-1490, but straight-jacketing things that aragon and castile must unite is a bit much. Their colonization of the Americas to historic proportions isn't so vital that we need to resort to thus brutish methods.

I see this one the way Countsix does.

This has gotten better since it has multiple paths to form. What is mostly needed is troop stacks and possibly a war command because if its too small it'll never DoW.

Delhi forming the Mughals, which is how it usually happens, assuming it does in the AGCEEP, last time I checked, is pure fiction. It is not a historical event it is a fantasy one.

Mere troop stacks and forced DOWs aren't going to be enough. That Afghan/Hindustan border is very very different every game. The forced DOW often will not be against a neighboring country. Also those Indians don't just die once the Mughals DOW them. Very often their coalition is the stronger one, not the weaker one.

Actually the primary problem is not that. Its that only England can have the USA form from it really. Not much work has been done for alternate scenerios which even in the best circumstances with the best AI and AI cheats can still happen that England doesn't dominate.

None of the other AIs are colonizing the Eastern US heavily, so yes it is an English AI problem.

 

[unmerged(2456)]

Would you mind expatiating on this point? You seem to be saying that you consider the union of the Spains and its colonization of the Americas something which a mod purporting to be the "most historically accurate" can do without. I'm wondering, because if you were to ask me to name the five most significant features of world history between 1400-1800, I'm certain the rise of Spain to hegemonic great-power status and its colonization of most of the New World would be on that list.

Well why not then just read history? Why should it happen 95% for something like that?

These are key for Iberia and should happen most of the time (~95%):

1. Spain and Portugal both existing when ToT needs to fire.
2. Spain and Portugal colonizing historical regions
3. Spain and Portugal for atleast become descently powerful early on.
4. Spain discoering America when it should and first.
5. Portugal mapping Africa aproximatly when it should and first.

These things are important, but not key and should happen descent amount of the time (~70%):

1. Spain taking on American natives and winning.
2. Portugal being involved to some extent in Near East and India.
3. Portugal and Spain colonizing historical regions with correct numbers.
4. Catile and Aragon uniting.
5. Granada dying when it historically should aproximatly (not within the first few years nor lingering on indefinatly), perferably by Castile/Spain, not Portugal.
6. Atleast some involvement with Spain/Portugal in the far east.

These things because they happened should likely happen, but shouldn't be prssed (~51%)

1. Spain taking out american natives in one war apiece.
2. Portugal dominating India and Near East Trade for a time.
3. The eventual decline of Portugal and Spain.
4. Heavier involvemnt in Far East than currently.

These things should be possible, but not likely since they didn't historically happen (~49%):

1. Some ahisotirc revolters forming under certain conditions
2. Colonization in other surrounding historical (ie not just colonize only provinces they histroically colonized, but also surrounding ones, FE Spain going and colonizing all of SA, including all of Brazil and Portugal concentrating on India than SA
3. Spain or Portugal be able to out-compete Britian and others in some unhistroic areas (FE Inidia) (this is kida related to #2)

These should be possible, but not likely. However, they shouldn't be stopped from happening on occasion, just the norm.

1. Granada dying quickly
2. Granda surviving infinatly long (perhaps to 1820(
3. A unifitcation of Spain/Portugal after the ToT
4. Navarra becoming a part of Spain/Castile or France/Dauphaine.
5. Aragon, if it survives doing <i>some</i> colonizing later on. Perhaps also Navarra...and dare I say Granada, but not so much for either of them.
6. Unhistorical Colonization, such as Spain colonizing Eastern NA or Portugal colonzing CA.
7. Castile or Portugal going Reformed or Protestant

These few things should preety much be stamped out (~5%

1. Both Spain and Portugal not existing.
2. Resurgant Granda (except for <i<Frantasy</i> Resurgant Granada scenerio)
3. A powerhouse Navarra
4. Non-Iberians controlling huge chunks of Iberia.

These are all my opinions of course, some of which i'm sure you agree with and others you don't.

I don't mean to pick a semantic or pedantic quarrel, but there is a consensus that certain major events should happen with near certainty barring the player's intervention - and the rise of Spain is surely one of those. As I understand it, the question is how to achieve those major events - some arguing that the engine can be made to bring them about naturally, and others (myself included) arguing the AI must be led to water before it drinks. If on the other hand we're back to square one, debating whether such major events should occur at all, then I'd like to know.

Again, I'm not arguing against the rise of Spain, but merely the method and the extent.