-
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.
Sorry, it never occured to me you couldn´t understand Pang´s intentions. I apologize.
I am only just starting to understand it. And now that I do, I will move to commenting on the game design. At least I know that my recent installation of Set 34 is now obsolete. No need to apologize - but accepted. Thank you.
Last edited:
I understand that there are two things which will affect whether or not Germany will do Barbarossa:
a) there is a ratio between the opposing units - set as you see fit. For sake of discussion, let's use 100 German mobiles against 300 SU units.
b) Next there is a "threshold"... that being a separate value which works very much like; the chance of it happening is so many percent".
For example, with a certain threshold setting, it might be that 50% of the time - when the unit ratio is 100 to 300 - Germany will cancel Barbarossa. Obviously, you would adjust the threshold so that possibly 90% of the time the chances are that Germany will attack when there exists a 100 to 300 ratio.
If I have the idea correct, may I suggest to you that having any threshold factor is simply building into the game for some chance that it will not run properly? Why have any threshold at all? Why simply not let Germany make its decision based only on ratio of units - this being something that can be more controlled.
The point is, if the ratio meets what you decide is appropriate for Germany to attack... then there should be zero chance that it will cancel. Agree?
Hi,
Lothos designed this with a weird emphasis on a combination of the historical date for Barbarossa and a certain random element. The AI never really "cancels" Barbarossa. If it meets the minimum conditions then there is a random element. If it fails to meet the random element it will just postpone and check again later. These random elements are set pretty high. So they normally fire the first time.
mm
Lothos designed this with a weird emphasis on a combination of the historical date for Barbarossa and a certain random element. The AI never really "cancels" Barbarossa. If it meets the minimum conditions then there is a random element. If it fails to meet the random element it will just postpone and check again later. These random elements are set pretty high. So they normally fire the first time.
mm
I think that Germany not doing historic Barbarossa can only be an embarrassment to the game designer. But that is not my point. My point is looking at the reasons why Germany cancels - and correcting that. I suspect that might be too much infra construction.
Next, a randomness in the decision is just that - randomness. It is not a game design fault if game designed to work on chance. But there should not be any chance as an extra consideration. As long as chance is a valid condition, we are blocked from searching for the real fault - inadequate attacker/defender ratio.
Finally, while I realize that - as done - the German AI will keep checking, it is rather unsatisfactory to start Barbarossa in September because ratio has now been attained and the event fires. Or ratio was OK in the first place, but randomness got a "no Barbarossa", but on following check randomness got a "yes".
I think I understand that Pang realizes he did not have threshold set right, and corrected it. Doesn't matter. If I do 100 hands off and Germany cancels once only because of the chance as per threshold setting, it is still needlessly wrong. Germany should only cancel when the ratio is inadequate.
EDIT: The "randomness" should never exist as a decision to attack when ratio suffices. However, there should be a different factor of "time delay" or "lack of predictability" as to exact date ONLY to keep players in the fog of war. Example: Japan doing Pearl Harbour Dec 2 (5 days early) is perfect to prevent players from capitalizing on known dates. I would suggest same for the most fixed "Danzig or War" event.
Next, a randomness in the decision is just that - randomness. It is not a game design fault if game designed to work on chance. But there should not be any chance as an extra consideration. As long as chance is a valid condition, we are blocked from searching for the real fault - inadequate attacker/defender ratio.
Finally, while I realize that - as done - the German AI will keep checking, it is rather unsatisfactory to start Barbarossa in September because ratio has now been attained and the event fires. Or ratio was OK in the first place, but randomness got a "no Barbarossa", but on following check randomness got a "yes".
I think I understand that Pang realizes he did not have threshold set right, and corrected it. Doesn't matter. If I do 100 hands off and Germany cancels once only because of the chance as per threshold setting, it is still needlessly wrong. Germany should only cancel when the ratio is inadequate.
EDIT: The "randomness" should never exist as a decision to attack when ratio suffices. However, there should be a different factor of "time delay" or "lack of predictability" as to exact date ONLY to keep players in the fog of war. Example: Japan doing Pearl Harbour Dec 2 (5 days early) is perfect to prevent players from capitalizing on known dates. I would suggest same for the most fixed "Danzig or War" event.
Last edited:
I have come to the conclusion that no matter what we say, no matter what kind of tests we do on 1.09, Pang will not remove a single infra serial. I will just stick with Jarski's brilliant creation that in my mind is undeniably better then 1.09.
So you have stopped being an optimist?
Well, I patched to v1.09 only because the "Lord" incorporated some of it into his mod. See you there. Have fun with your "no new DDs Royal Navy" when I catch up to you.
Well, I patched to v1.09 only because the "Lord" incorporated some of it into his mod. See you there. Have fun with your "no new DDs Royal Navy" when I catch up to you.
Last edited:
I always doubted whether we could persuade Pang to calm down with all the infra, and those doubts have been repeatedly reinforced.
I look forward to playing Jarski's marvellous mod against you!
I look forward to playing Jarski's marvellous mod against you!
Unfortunately I share your conclusions. And now that I learned that threshold may be blocking finding the real reason for Germany cancelling Barbarossa, I fear I may have little further to contribute. But Pang has not replied yet.
Yah, I look forward to it to - once I have done an SP on the mod. I'm very excited by both his mod because of new events, etc.... and our future game. But also I fear going up against what must probably be one of the best MPers in AoD.
Yah, I look forward to it to - once I have done an SP on the mod. I'm very excited by both his mod because of new events, etc.... and our future game. But also I fear going up against what must probably be one of the best MPers in AoD.
You are being funny. But you are also right, i remain rather opposed to avoiding Infra unless something substancial is gained by it, so simply deleting Infra will remain the exception. Reducing the serial lenght when needed is the way to go. So probably this will still not suit your preference and you are better off with Jarskis Mod.
That's what I said all along - at least in the sense that is a step in the right direction:
----------------
Unless something substantial is gained by it? How about an AI that actually challenges the player by building military units at a more pragmatic time? How about an AI that works best against a human player, not one endlessly tested in AI v. AI tests that very few AoDers actually enjoy playing.
At least we can agree there!
Last edited:
It's somewhat ironic Pang has left in a completely ahistorical event, whereby Germany will not perform Barbarossa because it has full knowledge of the entire size of the Red Army. Ignoring the fact the AI cannot use the espionage system to any rational degree, IRL there is little chance the Abwehr could've obtained such accurate information.
I highly recommend your tests! I'm sure you'll enjoy every single one. Aha, not so sure I'm one of the best MPers in AoD, just one of the most outspoken!
Well, there the difficulty setting applies. Therefore mainly AI vs. AI balance counts aswell as challenging mediocre players and giving them a challenge that tends to grow instead of the opposite. For motivation its better to start relativly easy and become more chalanging later than the other way around.
Is this game actually broken?
I had stated that I would present resource figures for USA at time of Pearl Harbour. However, the strange behavior with the passing of the Pearl Harbour event begs the question of, "Is this game broken?'
To explain, I was doing a hands off as Costa Rica, and all along I have been able to read USA's daily resource figures in the Trade Tab. The resource figures given below for Dec 1, 1941 (just before Pearl Harbour) are extremely representative of what's been happening for past several months. Oil is +600/day, energy +715, metal +383, rares +133.
But right after Pearl Harbour on Dec 2, 1941, the USA resources figures (according to read gotten in Costa Rica trade tab) change majorly.
Now it is oil +0.00/day (read “zero”) .... and it stays like that for all of following week (as far as I've run). The other resources take a considerable jump and change to energy +1022/day, metal +526, and rares +184.
NEXT, I loaded as USA for Dec 1st (auto save) which will not give me any resource figures until I run past midnight as is normal. While war starts Dec 2nd, the resource figures are of date before war. Problem is they differ considerably from what I read as Costa Rica. The new resource figures obtained opening as USA BEFORE WAR now are oil +647.4/day, energy +801, metal 429, energy 149. All figures are much higher than they were when read in the Coata Rica trade tab. That should not be, I think.
ANYWAY, to make matters even worse, continuing to run as USA after the start of war, it majorly changes yet again. Now daily oil accumulating jumps up to +798/day, energy +1172, metal +604 and rares +225. I assume it is the removal of peacetime penalty that increased all daily surpluses because all production increased. Similar happens watching as Costa Rica (except that the oil surplus disappeared). Basically, the Coatsa Rica read showing no oil must be broken, while the differences in readings as to how gotten might be whatever. Clearly, resources increase dramatically when war starts - except for the "broken oil reading" playing Costa Rica).
To summarize, looked at as played via the USA - with removal of peace time modifier, USA is stockpiling +800 oil per day in early December 1941.
BUT two problems exist:
a) Why any difference in resources with how game is viewed (thru Costa Rica or the USA)? And the differences are significant.
b) What the heck happened with the oil? As Costa Rica, we see USA suddenly "evaporating " 600/day glut of oil, and as it stays like that for days later (exactly 0.00 oil/day) it feels that the read out is broken.
But looking at the USA, we see the oil glut going from +600/day accumulating to +790 per day accumulating when war occurs. It is not on account of cancelled trades because there are only 3 oil trades, so making them trackable. Clearly this increase fits the other resource increases of same time; and is just peacetime modifier removed.
ANYWAY, I think I have made my point: the daily resources accumulating because of over-abundant resources is nuts. Looked at differently (total daily produced) then on Dec 4/41 (2 days after Pearl Harbour) the USA is producing 984.68 oil/per day... for sake of argument 1000/oil per day.
As regards the stockpiles, they are easiest measured in millions (or fractions of a million) because they are so huge.
As regards infra, the USA is completing the final build in the last serial. BUT IT HAS JUST RECENTLY LAID DOWN 3 NEW INFRA SERIALS even though those provinces are already at 200%. What can I say? Pang has brainwashed the game so that now the countries just build infra.... and build infra.... and build more infra - and when at 200% - just damn well build still more infra.
As regards the four factories constructing at accelerated speed, they will continue to consume 45 IC/day until 1948.
Nice game design, Pang.
I had stated that I would present resource figures for USA at time of Pearl Harbour. However, the strange behavior with the passing of the Pearl Harbour event begs the question of, "Is this game broken?'
To explain, I was doing a hands off as Costa Rica, and all along I have been able to read USA's daily resource figures in the Trade Tab. The resource figures given below for Dec 1, 1941 (just before Pearl Harbour) are extremely representative of what's been happening for past several months. Oil is +600/day, energy +715, metal +383, rares +133.
But right after Pearl Harbour on Dec 2, 1941, the USA resources figures (according to read gotten in Costa Rica trade tab) change majorly.
Now it is oil +0.00/day (read “zero”) .... and it stays like that for all of following week (as far as I've run). The other resources take a considerable jump and change to energy +1022/day, metal +526, and rares +184.
NEXT, I loaded as USA for Dec 1st (auto save) which will not give me any resource figures until I run past midnight as is normal. While war starts Dec 2nd, the resource figures are of date before war. Problem is they differ considerably from what I read as Costa Rica. The new resource figures obtained opening as USA BEFORE WAR now are oil +647.4/day, energy +801, metal 429, energy 149. All figures are much higher than they were when read in the Coata Rica trade tab. That should not be, I think.
ANYWAY, to make matters even worse, continuing to run as USA after the start of war, it majorly changes yet again. Now daily oil accumulating jumps up to +798/day, energy +1172, metal +604 and rares +225. I assume it is the removal of peacetime penalty that increased all daily surpluses because all production increased. Similar happens watching as Costa Rica (except that the oil surplus disappeared). Basically, the Coatsa Rica read showing no oil must be broken, while the differences in readings as to how gotten might be whatever. Clearly, resources increase dramatically when war starts - except for the "broken oil reading" playing Costa Rica).
To summarize, looked at as played via the USA - with removal of peace time modifier, USA is stockpiling +800 oil per day in early December 1941.
BUT two problems exist:
a) Why any difference in resources with how game is viewed (thru Costa Rica or the USA)? And the differences are significant.
b) What the heck happened with the oil? As Costa Rica, we see USA suddenly "evaporating " 600/day glut of oil, and as it stays like that for days later (exactly 0.00 oil/day) it feels that the read out is broken.
But looking at the USA, we see the oil glut going from +600/day accumulating to +790 per day accumulating when war occurs. It is not on account of cancelled trades because there are only 3 oil trades, so making them trackable. Clearly this increase fits the other resource increases of same time; and is just peacetime modifier removed.
ANYWAY, I think I have made my point: the daily resources accumulating because of over-abundant resources is nuts. Looked at differently (total daily produced) then on Dec 4/41 (2 days after Pearl Harbour) the USA is producing 984.68 oil/per day... for sake of argument 1000/oil per day.
As regards the stockpiles, they are easiest measured in millions (or fractions of a million) because they are so huge.
As regards infra, the USA is completing the final build in the last serial. BUT IT HAS JUST RECENTLY LAID DOWN 3 NEW INFRA SERIALS even though those provinces are already at 200%. What can I say? Pang has brainwashed the game so that now the countries just build infra.... and build infra.... and build more infra - and when at 200% - just damn well build still more infra.
As regards the four factories constructing at accelerated speed, they will continue to consume 45 IC/day until 1948.
Nice game design, Pang.
Last edited:
You should try the same advice for DEVs that haven't got a clue as to what all the implications are for what they are doing. You should roll back v1.09 to v1.07 BUT WITH ALL THE GOOD WORKS of Poinier et al somehow preserved.
Since you yourself stated - when some unexpected consequences appeared with the launching of v1.08 - that the best fix would be to go back to v1.07 - my honest suggestion should not be any radical idea.
There is much more to consider in this debate that just the mechanics of the game.
Last edited:
Regarding the resource 'bug' where you don't see the correct resource levels of other nations, that's typically caused by loading up more then one save. I.e. It's always best to restart the game when coming out of one save and loading up another.
The AI trying to build infra even after it's at 200% is a known bug that unfortunately we can't fix as we'd need a C++ programmer. Maybe a quick fix is to tell the AI to end IC builds at 1939/40 - meaning it should not add anymore factory or infra builds, but will maintain production on the current builds (I.e. The ones placed in the starting scenario).
But yes, the amount of resources the USA stockpiles is ridiculous. As is spending 45IC until 1948 on 3 factory serials.
The AI trying to build infra even after it's at 200% is a known bug that unfortunately we can't fix as we'd need a C++ programmer. Maybe a quick fix is to tell the AI to end IC builds at 1939/40 - meaning it should not add anymore factory or infra builds, but will maintain production on the current builds (I.e. The ones placed in the starting scenario).
But yes, the amount of resources the USA stockpiles is ridiculous. As is spending 45IC until 1948 on 3 factory serials.
Thanks... but I am rather sick of this boring "hands off" - and glad I have fulfilled my promised report so I can end it.
As regards now making a case of:
"USA over abundant resources result in over abundant Lend Lease; and all countries being much more resource self-sufficient combine to majorly reduce world-wide trading which then negatively affects game strategy for the Battle of the Atlantic" is much more difficult.
Not sure it is worth it given the hardness of the coconut shell that really needs cracking first before new flavor can be instilled to create a nicer cocktail... better enjoyed by all.
Last edited:
I did not have more than one save running anytime. I always surrendered as Costa Rica, "quit game" and hit "single Player" to access file save which I then opened as USA. I think that is correct methodology, or do I need to EXIT game?
But does the AI actually spend IC trying to build those serials past 200%? If so, that might be serious.