Home Trending Latest threads New posts Developer diaries

Historical Accuracy vs Historical Precision

Jump to latest Follow Reply
  • 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.
Show only dev responses
U

unmerged(31587)

Sergeant
Jul 7, 2004
94
0
I used to play EU2 vanilla and having discovered AGCEEP not long ago, I decided to try it out based on its premise of being a more "historically accurate" mod. But the more I play it the more I realize that while AGCEEP is far more "historically precise", the question of whether it is more "historically accurate" is harder to say. In some cases yes but in some cases no.

Now I should define these terms. By "historical precision", I mean the "exactness" of the historic setup, details and events. In this regard, AGCEEP is more "historically precise" than vanilla. A great example of course is the entire HYW setup which was fairly well done.

But what I mean by "historical accuracy" is not the exactness or correctness of details but the "overall" historical "truthfulness" of the historical results in a "big picture" kind of way. In other words, how the game's results resemble real life historical outcomes overall (not necessarily every detail but the general basic outcome).

One thing I have noticed is that it SEEMS to me that in some cases, AGCEEP decided to trade-off "historical accuracy" for "historical precision". The way the game is setup in AGCEEP, I have noticed many regular historical outcomes that are not "accurate" and in some cases less accurate than in vanilla!

The problem, as we all know, is that the basic game engine and design of EU2 is too "primitive" to make it a true historical simulator which is what makes the game interesting after all. I mean, that is why we almost never play "Fantasia" which is just a gloried RISK game. The problem is that the game engine treats the game as RISK that is why you see "random" wars and alliances and weird provinces being conquered and such. So no matter how "historically precise" you make it, without "guidance" the game will just degenerate into nonsense and a glorified version of RISK, in effect.

Take a look at the HYW setup of France. Now it is very "historically precise". But I noticed that there is a lot of "guidance" (in the form of lots of determistic events that favor French victory in HYW) to make it also "historically accurate". But if this "guidance" wasn't there, meaning you just let all the nations determine the outcome thru the normal game mechanics, obviously 9/10 times the historical flow will degenerate into nonsense.

I guess what I am trying to say is that increasing "historical precision" may not make the game more "historically accurate". In some cases it may even make it LESS accurate because the AI and the EU2 game engine simply cannot cope with that level of precision and because greater precision also gives the AIs more opportunities and degrees of freedom to deviate historically into nonsense (if no "guidance" is involved). And if precision leads to less accuracy and more "nonsense" then it seems all that hard work is for naught. As someone new to AGCEEP, this is how I see it IMHO.
 
Yes, but EU2 is a game, not a historical simulation. This is the same sentiment that is expressed over and over again, in the forums of each and every Paradox game. There was another thread like this just last week in the AGCEEP subforum. I think everyone making these sort of criticisms ought to get together and make their own mod. It can be the Hands-Off Historical Simulation Mod. You just fire it up, sit back, and watch history unfold accurately. I, however, would rather continue to play AGCEEP.
 
Garbon said:
I hate to sound rude, but I'm not exactly sure what one should do with this information? Even if what you say may be the case, we are going to just suddenly stop.
Click to expand...

I likewise don't want to sound rude, but the obvious answer to Garbon's first question would be to take these opinions seriously rather than dismissing what is clearly a large constituency for AGCEEP with flippant replies like the one above.

polypheus' analysis is sound and in my view correct, and I thank him for providing a useful distinction. The facts appear to be that

1) in its present form AGCEEP is, as polypheus has said, far more historically precise than any other mod. The initial 1419 setup and the first century or so play out in an interesting and historically plausible way, generally speaking. This is the fruit of the labour of many people, myself included, who have devoted considerable time and expertise.

2) OTOH, past the 15th c. the mod plays generally without much historical accuracy, despite various rather sporadic efforts at fine-tuning.

What is to be done? Individually, one answer to the question would be to give up on AGCEEP and play IDLF's mod Europa Portugalis (EP), which sets out precisely to achieve an historically-accurate game, and uses AGCEEP as a base to capitalize on its historical precision. I have tried EP; I'm impressed at what it achieves. But (as I've said to IDLF himself), I'm unwilling to abandon AGCEEP for a variety of reasons.

Collectively, then, what is to be done? The first and obvious answer is to review the established policy which forbids (except in certain notable cases where they have been used anyway) so-called "AI cheats", or in other words, use of the moddable game mechanics behind the scenes to direct the AI along historically-accurate paths: AI files, AI-only events, etc.

This is the only feasible way to make what is already the most historically-precise mod historically accurate as well, and both the HC and AGCEEP players need to think seriously about incorporating such measures. The work is already well underway, it's not a question of asking who will do it - considerable parts of it are already done. It's simply a matter of deciding whether AGCEEP will live up to its claims.
 
If I play 100 hands-off-games it's more important that I get 100 very different outcomes than 100 outcomes that closely resemble what really happened.
Making history is extremely chaotic process, and as such a perfect simulation would never give the same result twice.
That's why I'd rather have a an accurate starting-setup than a precise finish-setup.
 
I think Polypheus has some good points, and they are known to all long time modders. I think it's a serious problem that ought to be high on our agenda. AGCEEP is/was put together from many small parts often without the necessary effort to make it fit together without unwanted consequences.
CountSix points to one solution of the problem, that is playing EP instead, that has seen a lot more work and attention to over-all and long-term historical accurate development of the major nations. (The history of the small ones is sometimes sacrificed in it.)

I don't think we would be successful in copying it, and we would probably spend a lot of time to modify stuff already made by IDLF, and we would end up with a mod played by fewer people than we have now, and for those that will like it they would probably prefer IDLF's original.
The solution for AGCEEP, as I see it, is to urge modders to spend more time in playtesting and fixing major nations. That is a road that I think we have hardly not started to explore. Until that is done I won't refute it, even if the final result will never be as close to real history as something forced of course.
 
Last edited:
Count Six said:
2) OTOH, past the 15th c. the mod plays generally without much historical accuracy, despite various rather sporadic efforts at fine-tuning.
Click to expand...

Precisely, sporadic. Not much effort has really been given to make sure things are progressing appropriately past the the 15th century. Hell, people are still trying to balance that century.

Count Six said:
Collectively, then, what is to be done? The first and obvious answer is to review the established policy which forbids (except in certain notable cases where they have been used anyway) so-called "AI cheats", or in other words, use of the moddable game mechanics behind the scenes to direct the AI along historically-accurate paths: AI files, AI-only events, etc.

This is the only feasible way to make what is already the most historically-precise mod historically accurate as well, and both the HC and AGCEEP players need to think seriously about incorporating such measures. The work is already well underway, it's not a question of asking who will do it - considerable parts of it are already done. It's simply a matter of deciding whether AGCEEP will live up to its claims.
Click to expand...

Umm...when have AI files ever been considered "AI cheats"? Regardless I'm curious as to see if opinions on "ai cheats" have really changed.

And no, the difference is that except for the slight mention of "guidance" near the end, polypheus's post offers up nothing but a criticism wrapped in a few compliments. Critiques without suggestions of what can be done have very limited utility value.
 
Garbon said:
Critiques without suggestions of what can be done have very limited utility value.
Click to expand...

I read and write quite a few book reviews, and while I often read opinions about a book's strengths and weaknesses, I rarely find someone telling the author how he should have wrote it. On the contrary, thoughtful critique is often very useful in and of itself, especially when it comes from an unbiased and fresh outsider. Rushing to silence useful criticism is a certain sign of stultification.

What is AGCEEP for? If it's meant to be a sort of online private club for old-timers or the HC, it's pointless. If it's meant, as it advertises itself, "to make the game more historically accurate", then its time to move on to Phase II - as Norrefeld says, modders need to be encouraged to make existing event-trees work more accurately, and discouraged from producing ever-more elaborate new event series. However, this will not be enough - there must be a clear understanding that if a measure is required for a country to perform more accurately, then that measure will be accepted by the HC, without cavil, complaint, or dogmatic adherence to some theory about "determinism".

In other words, if someone can fix an historical accuracy problem, let them do it. And if the fix requires so-called "AI cheats", then so be it.
 
I hate to be a party pooper, but everyone in this thread is guilty of exactly what they complain about: being all talk and no action.

Forget getting the last word, or being "right", or whatever. Find the problem and fix it. State specific instances. State SPECIFICS.
 
Count Six said:
I read and write quite a few book reviews, and while I often read opinions about a book's strengths and weaknesses, I rarely find someone telling the author how he should have wrote it. On the contrary, thoughtful critique is often very useful in and of itself, especially when it comes from an unbiased and fresh outsider. Rushing to silence useful criticism is a certain sign of stultification.
Click to expand...

If you look back my intial post, was wondering how such criticism is useful especially as its really not new criticism.

Count Six said:
What is AGCEEP for? If it's meant to be a sort of online private club for old-timers or the HC, it's pointless. If it's meant, as it advertises itself, "to make the game more historically accurate", then its time to move on to Phase II - as Norrefeld says, modders need to be encouraged to make existing event-trees work more accurately, and discouraged from producing ever-more elaborate new event series. However, this will not be enough - there must be a clear understanding that if a measure is required for a country to perform more accurately, then that measure will be accepted by the HC, without cavil, complaint, or dogmatic adherence to some theory about "determinism".

In other words, if someone can fix an historical accuracy problem, let them do it. And if the fix requires so-called "AI cheats", then so be it.
Click to expand...

Of course I agree that the AGCEEP should now be focusing on fixing nations and making sure they work properly. I mean I for one was always attempting to do so, but if its needed then lets by all means give that as our official new course of direction.

However, I agree with Norrefeldt that this road of fixing nations has not really been explored and as a result, and thus data is quite insufficient to be postulating that ai cheats are the only viable solution. And the current moment, it is unknow that there will be a correct 'measure' for fixing a given nation. Thus a retraction of the current policy on ai cheats at this time would premature.
 
Mad King James said:
I hate to be a party pooper, but everyone in this thread is guilty of exactly what they complain about: being all talk and no action.

Forget getting the last word, or being "right", or whatever. Find the problem and fix it. State specific instances. State SPECIFICS.
Click to expand...

In a thread on policy, it would be off topic to be discussing specific circumstances. Thats why we have regional threads.
 
I didn't mean to sound flippant, but I really do not understand these sorts of criticisms. Or rather, I understand them, but I too would like to know specifically what you have in mind. It is like the poster in this thread who wanted the entire to GC to play out historically "accurate". Should every country have a set of AI events that strip out all non-historical choices? But even that wouldn't insure the game would proceed "accurately," since it couldn't take into account alliances, peace settlements, etc. When a historical war ends, should an event fire to cede the corrent provinces, etc. to the correct participants, regardless of the actual game outcome? Should alliances among AI countries (hell, even AI and human players, if we are going for "accuracy") be forced to insure historical outcomes? These be would the logical solutions to this sort of criticism, and it goes a lot further than AI cheats. AI cheats are about game balance, and what he is talking about is creating a simulation.
 
Garbon said:
In a thread on policy, it would be off topic to be discussing specific circumstances. Thats why we have regional threads.
Click to expand...

That's just it though, we have these long, involved policy debates where nothing ends up being decided on, no consensus is ever reached, and nothing ever gets done.
 
I wonder why so many assume that the historical outcome was allways the most likely outcome... Why should the historical outcome be the most common outcome in a game that encompasses 400 years?
 
Garbon said:
If you look back my intial post, was wondering how such criticism is useful especially as its really not new criticism.
Click to expand...

It is useful IMO particularly because it reflects a fresh view - you'll note polypheus says he began to play AGCEEP "not long ago". In other words, this appears to me to be the cogent opinion of someone who has just encountered AGCEEP, as opposed to those of us who've been tinkering with it before there even was an AGCEEP.


Of course I agree that the AGCEEP should now be focusing on fixing nations and making sure they work properly. I mean I for one was always attempting to do so, but if its needed then lets by all means give that as our official new course of direction.

However, I agree with Norrefeldt that this road of fixing nations has not really been explored and as a result, and thus data is quite insufficient to be postulating that ai cheats are the only viable solution. And the current moment, it is unknow that there will be a correct 'measure' for fixing a given nation. Thus a retraction of the current policy on ai cheats at this time would premature.
Click to expand...


No, this is simply a recipe for inaction, and if we follow this course nothing will ever be done toward a Phase II AGCEEP. The door has to be opened, and experimenting with so-called "AI cheats", which have already been proven effective in another mod to achieve the purpose you endorse, must be allowed as part of the official package.
 
Mad King James said:
That's just it though, we have these long, involved policy debates where nothing ends up being decided on, no consensus is ever reached, and nothing ever gets done.
Click to expand...

Maybe thats because things are being done in the background that aren't being posted. Besides of course the Burgundy thread that polypheus has recently put up.
 
There's a lot of ground for me to cover in explaining my POV and insight (for whatever its worth) so I'll try to do it with more detail here.

PERSONALLY, I suspect most people are attracted to EU2 precisely because it is to some extent a historical simulation rather than a pure strategy game (in which the initial setup is precise and accurate but then after that anything goes). TO ME, although there is a great deal of "determinism", that is what is attractive about the game. If we took the "anything goes" approach, then what's the point of deterministic events? If I want to play a game of total make believe history of anything goes and total lack of determinism, I play something like Civilization for that.

Now if it were possible (I don't think it is actually) I would first create a completely pure historical simulation mod stripped of all alternate historical choices and outcomes so that in hands-off, that mod would strongly resemble real-life history 90% (if not 100%) of the time. Then once this "golden", "pure" mod was done, I'd start to slowly "loosen" it up, add alternative choices, consider alternative outcomes, etc. I don't expect this to ever be done but it is simply a way to explain my philosophy because this approach would result in a much more coherent, accurate and tight mod. Then you loosen it as much as it makes sense to do so and have great ability to account for and handle any major historical deviation (because you have controlled how much to loosen up from that "pure" mod).

Now as for "historical accuracy", I see that in AGCEEP the "focus" APPEARS to me to be tinkering and adding "precision" without consdering the "big picture" when the "big picture" TO ME matters a whole lot more.

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?

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.

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).

4. Prussia - 'nuff said

I could go on and on. Now I know some things are easier to fix than others. (Prussia would be hard without some sort of Netherlands-like birth approach that allows it to emerge "out of nowhere" but I actually would favor that and it wouldn't be THAT ahistorical. And it is supposed to be a MAJOR nation!)

Now I am not saying every little detail needs to be right (probably impossible). But I feel that some of the major details and outcomes should be modelled and reflected on a consistent basis a "majority" of the time. The key is to identify what those major details and outcomes are and make them happen on a more consistent basis IMHO and if necessary use "guidance".

I mean that is exactly what was done with the whole HYW setup right? You guys didn't just setup the scenario and let the nations determine the outcome using only EU2 mechanics as that would have resulted in nonsense most of the time with France rarely forming. So the same approach could be used elsewhere also.

If I had a choice between the more "precise" AGCEEP HYW setup (but without "guidance" to make the result accurate most of the time thus resulting in lots of nonsense many times) as opposed to the vanilla setup which is far less precise in its details but which more consistently results in France winning HYW and forming properly, guess what I would PREFER the vanilla setup in that case! The vanilla would be less PRECISE but far more ACCURATE. I hope this example explains the bottom-line point.
 
...I suppose I'm the only one who doesn't understand what distinction Poly's trying to make between historical accuracy and historical precision. The closest I can make out is that one indicates that the starting positions are correct and the other indicates that events reflect historically plausible outcomes. To my mind, that just means it's partially historically accurate/precise and partially not.

I'll be honest, I've been working on new events for BYZ more than playing lately, and when I play, a lot of the time I'm satisfied with what I've managed to do and stop playing around 1650-1700 (esp. with BYZ which so far hasn't had much interesting to happen to it after that.) For those who are playing further, I think we should start a sticky or use this thread to keep track of historically screwy things that keep happening. They indicate that the mod's current setup/developing setup is flawed.

The AKK situation is a great example of this - if AKK regularly gets swallowed, either a) it needs to start out in a stronger position and/or be helped out along the way or b) other events need to adjust to AKK being gone. The Shia rebellion would have happened regardless of whether AKK or QAR or IRA for that matter won out.

Scripting what if events for every aspect of history over 400 years is obviously ridiculous and too much work, but we can and should be fixing major implausible outcomes - AKK and SPA in America being great examples. It's not empty criticism or some fringe idea: it's just a matter of fine-tuning and perfecting what's already a great mod for a fantastic game.

jay,
who doesn't want pure determinism, but thinks that the ahistorical x-factor should be the players and their actions, not errors or lack of event scripting.
 
I will join Garbon in thinking that the best way of solving these problems is to roll up your sleeves and help out in the regional threads though. :)

jay.
 
You must log in or register to reply here.
Top Bottom
Back