Investigation of AI reassessment upon reload

[Kraxis]

Nah, I rather think that the AI goes "Hey what is going on here... Ok I better get in a position where I can formulate a good strategy (hence sieges broken off) and I need to secure myself (hence acceptance of Protectorates)."
Basically the AI goes into 'start'-mode. It starts over again as if the game starts at that date, which seems to be a fitting explaination as the info to make the AI realize it was actually in a running game could possibly be rather substantial, and an obvious place to cut if size was demanded smaller.

[HarunTaiwan]

I agree with you. I think the results show if you play 6-7 turns it's not so bad...so I will play less now since sometimes I only have 20-30 minutes...shouldn't have had that kid, should I?

[roguebolo]

Notice Bromley confirms that's it's all rebel territory many without walls that you are seeing being taken, and not AI territory.

Right. I've done the same tests. I've watched what happens in Sicily without load games. I've also watched what happens in midgame with them. Even then the AI will relieve sieges, but the net effect is different. It continues to threaten the same territories. Like I said, don't take my word for it. I'll send you a number of different save game positions that show it to you, from the same game even.

The truth is that the AI has never caught me with my pants down so that waiting out a siege would ever work anyway...so it boils down to a matter of, "Should I sally or meet them in the field?"

The fact of the matter is that in the first 20 moves, without load games, it's primarily rebel territories that fall.

I doubt alliance does much at all, since the AI breaks those at will.

Here, I agree with The Shogun in that there is (in his words) "a misunderstanding of the AI." In midgame, the AI definitely does not easily break or form alliances. There's a very clear logic on what it considers acceptable. However, in the early phases of the game it will change them willy-nilly whenever there's a reload. It doesn't seem to merely reassess the situation and roll some dice for 50% odds -- it seems to intentionally reverse its previous decisions.

Play through the test scenario one more time with load games, but this time watch what happens in the diplomacy Pane with Egypt/Seleucids/Armenia or the factions in Sicily. Make note of ceasefires as well. You'll soon see a pattern of ... they removed the siege ... a ceasefire is in effect ... now they've formed an alliance ... etc. I should have recorded that information and posted it because it gives some good insight as to what is going on.

At any rate, if you doubt that alliances have an effect on the AIs willingness to follow through with sieges, I can illustrate with specific save games in which that is the one and only determining factor that modifies the behavior of the AI, despite load games. Isn't there an upload area for sharing data?

the last thing you want in a wargame is stability. You need to feel threatened as well as threatening.

Now, that's taking everything I've said out of context. Stability is a very generic word which can mean a lot of different things. I think you know perfectly well that I was not referring to a "comfortable, easy-going game where everything is under my control." I was referring to a stable political landscape which causes the AI to behave in a more stable, reasonable manner. You certainly don't want the kind of instability that it exhibits early in the game with numerous save/loads to continue throughout the game.

Sometimes I wonder how many people here are actually playing games beyond the first twenty moves. I can definitely see that there has not been much testing beyond that period of extrema. But to truly understand the phenomenon, you need to come up with instances in which it is NOT exhibited, and determine why. You need both a positive and a negative.

While the statistical data supporting the load game behavior is very good, I have seen very few suggestions that explain this behavior -- nothing that illustrates a truly good understanding of it.

It's a hard proposition to make projections about the overall difficulty of the AI factions in midgame on the grounds of its early behavior resulting from load games. Generalized statements like "it's less aggressive" don't carry much weight in the light of actual savegames which illustrate the AI's midgame behavior to be substantially different. They tend to be semantical statements arguing a foregone conclusion of an empirically observed and repeatable fact, and that's simply not good enough. You not only need to be able to make projections about the future behavior of the game, but you need to back them up by observing and measuring the resultant behavior as well.

A similar semantical argument (and similarly illogical) would be, "It sounds to me like you want one of the AI factions to beat all of the others for you, so that you can take it on after it's overextended itself." When the Scipii, your allies at that stage of the game, grow stronger, Carthage and the Greek Cities grow weaker. But the sum total of the AI factions is the same, and must all eventually be defeated.

I really feel that some serious thought should be given to proven instances in which this phenomenon does not seem to exist or does not seem to exhibit the projected results. I don't think it will be fully understood unless those are understood as well. I have a number of such save games, which is why I've asked several times about sharing them with people. I'd like someone else to take a look at them and see what conclusions they come to. If anyone is interested, please PM me and I can send them by email.

[Bromley]

How does that fit with my last test? That seemed to show, to me, that the AI broke off all sieges (note that I did not visually confirm this, it's just that they ceased to expand into each other and only took a few Rebels). That's even with a human playing and after the diplomatic scene had been given 20 turns to mature (although the human was passive in those first 20 turns).

Also, although without save/loads the Rebel territories are a larger part of the total territories taken in the first 20 turns, there were more provinces swapped in my 270-260, no save/load test than taken from Rebels. The proportion of Rebel territories to all territories taken approaches 100% as save/load interval is reduced. I personally believe that this is because the Rebel territories are easier to take (fewer defences, smaller defending armies, no supporting armies), but may be wrong there.

[HarunTaiwan]

Not only easier to take, possibly hardcoded into diplomacy for the AI to ignore rebels as a negotiating partner, etc. If you can't negotiate with them, lifting a siege won't result in any diplomatic benefit, so you might as well see it through.

Also, Roguebolo, your point noted on "stability" but yes, I do think having AI's expand and compete along with you, to the point where you face factions of similar sizes makes a much better game.

[roguebolo]

Bromley, yours is some of the statistical data that I said was referring to as very good. I guess the point is that although you've been able to create scenarios which duplicate the phenomenon, I've been able to dig up some scenarios which duplicate explicitly contrary behavior. In some cases they illustrate the very kinds of AI behavior that The Shogun mentioned in his post, but that does not necessarily redeem them. Aren't both worth some study if you wish to understand that behavior?

I actually do visually confirm the results of the first 20 moves with an "End Year" without save/loads, and there are typically four territories under dispute that are not rebel. All of the remaining acquisitions are rebel, and if you ask me that's some pretty sensible playing. It's also easier to debate the stance and prove the projection that rebel territories being occupied at a later point in the game will be less developed than if they were occupied earlier. I hope no one disagrees on that point.

It's much more difficult to debate the idea that a vague concept like "the AI is less aggressive to other factions" has any real impact on the gameplay, particularly in midgame -- then you'll need to face the entire slew of possibilities that The Shogun mentioned in his post and show that each and every one was not a reasonable assumption.

I honestly do consider the inactivity on the part of the human player to be a weakness of these tests. I'm not going to debate the reasons why but eliminating that lack of inactivity is what yielded the contrary results that I consider worth studying. In one case, even something as simple as forming three alliances in the space of two turns completely changed the behavior of the AI and made it more aggressive.

I don't really agree that a single large faction makes a better opponent; I would rather see smaller factions allying against me as there is more diversity in the gameplay because of the different unit types they can create. The only case in which I would like to see other factions grow at an equal pace is when I'm playing a Roman faction and eventually have to face the other Roman factions after civil war begins. When I first played I thought this is the way things would turn out, and I was surprised to find my Urban Cohorts facing large bands of Hastati and Velites.

This is a little off-topic, but the thing that reduces the challenge the most for me is that the AI does a poor job of building the more powerful units and a balanced army, and that it does a poor job of upgrading the arms and armor for its units. (There was actually a minor improvement in the latest patch, but not enough to be significant.) If it did a better job at these issues, it would make a considerably superior opponent.

[roguebolo]

Apparently my posts on this forum have become somewhat controversial, but I guess I'm getting used to that. Recently on a different forum regarding a different game some of my posts were originally considered controversial, eventually got me flamed by a few individuals, and ultimately won me an award as their best strategy columnist. Bromley has asked me to post a summary of my position because apparently some people have been referring to my posts as evidence that there is no loadgame bug.

In some respects, they are correct. The problem here is the word "bug". To a programmer, a bug is a clear and undetected error that produces incorrect or unintended results or an interruptive abberation in the behavior of a program, such as divide-by-zero errors or integer overflows or a deadlock due to improper use of synchronization objects. I have encountered, tracked down, and eliminated many such bugs over the last 20+ years, and this does not truly qualify as a "bug." I assure you that if you were to refer to it in those terms to the developer(s) responsible for the AI that they would take at least a small amount of offense to it. In fact, as a fellow programmer, I take a small amount of offense to it myself on their behalf. I do wish everyone would adopt a more accurate description such as the "load game reassessment behavior" or "load game reassessment anomaly".

What we are dealing with here is a peculiar behavior of the reassessment strategy under conditions of polar extrema -- beginning from the opening move of the game, without player interaction except for the refusal of all alliances which serves to reinforce the extrema. I'm not undermining this test, because it quite possibly gives us some insight into the viability of the the algorithms and heurisms used by the AI throughout the course of the game, including midgame and endgame, in the absence of such extrema.

There is a clear tendency for the reassessment that occurs after a load game to often (not always) reverse diplomatic decisions and to relieve sieges. It's also clear that this behavior tends to diminish or manifest itself in a different way as the game moves away from a position of extrema. Unfortunately, I'm one of the few people who has apparently studied the phenomenon with that possibility in mind, although Bromley is moving in the same direction. It helps to have a large database of saved games to draw from.

I think there are some weaknesses with the research that's been undertaken here. It seems that many people have reached "obvious", foregone conclusions from their observations of this fairly limited test, projecting those "absolutes" to the conclusion of the game even when the extrema is removed. There is no support for such conclusions except inasmuch as they can be supported semantically. There is virtually no empirical evidence as to the resulting strength of the AI opposition as a whole due to the AI's reassessment behavior.

There are also a few conclusions that are entirely accurate, although unsupported empirically, such as the fact that the load game behavior will allow the human player more time to acquire rebel territories and more time to develop them than the AI factions, which definitely gives the human player an advantage. I think this could easily be supported empirically, although the mathematical analysis would be quite challenging.

There has been very little attempt to understand the reasoning behind the AI's behavior, which I consider to be a key factor in suggesting how it might be changed. To simply say, "The AI needs to stop relieving sieges" could very well introduce another undesirable behavior in the AI. In order to understand the decision-making process of the AI, and how it might be improved, it is important to analyze those scenarios where the AI is behaving in a different manner than that which was observed under conditions of extrema.

In other words, I think the research here is a good start, but incomplete. It is not conclusive because it does not exhibit empirical evidence of the projected conclusions.

[Puzz3D]

Basically the AI goes into 'start'-mode. It starts over again as if the game starts at that date, which seems to be a fitting explaination as the info to make the AI realize it was actually in a running game could possibly be rather substantial, and an obvious place to cut if size was demanded smaller.

I think that start-mode is a good word to describe what's happening. The AI has to rebuild some memory resident data structures.