Yes. Do you remember AI in HoI1 1.0?
Horror. Pure horror.
Do you remember the EU2 AI in any version?
Incidentally, also horror. Pure horror.
But, clearly, that would have nothing to do with HoI AI being terrible.
-
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.
- Status
Do you remember the EU2 AI in any version?
Incidentally, also horror. Pure horror.
But, clearly, that would have nothing to do with HoI AI being terrible.
I've edited my post to explain more what I had in mind.
I assume you know that power of EU3/Rome AI is not based on its combat part (which is critical for HoI), but on the diplomacy and pre-war build-up. So no, I don't find references to EU3 AI as cure to HoI problems fitting.
I assume you know that power of EU3/Rome AI is not based on its combat part (which is critical for HoI), but on the diplomacy and pre-war build-up. So no, I don't find references to EU3 AI as cure to HoI problems fitting.
My, just because it's Intelligence people start talking about the Artificial aspect of it. I frankly haven't heard anything of substance about its Human aspect, especially in Multi-Player. After all, one could end up balancing it on single-player yet end up with it completely broken over the Net.
Steady on, chum.
The 1.09 AI isn't that bad. It's certainly better than 1.00 (as it should be) and arguably EU3 1.0's AI
MP player intelligence is often hard to decypher...
Im MP games the issues around having an AI country is that it is often irrational in its choices of build/research and easy to defeat as it responds to very predictable stimuli. For example, I played one game where landlocked Hungary was using both its research teams to research naval stuff. Ok, so they had an Admiral as Head of State but still...
An AI that works better in SP mode must be better in MP mode than what we have now, surely?
K
Im MP games the issues around having an AI country is that it is often irrational in its choices of build/research and easy to defeat as it responds to very predictable stimuli. For example, I played one game where landlocked Hungary was using both its research teams to research naval stuff. Ok, so they had an Admiral as Head of State but still...
An AI that works better in SP mode must be better in MP mode than what we have now, surely?
K
The overwhelming majority (95%, most likely) of all players never play a single MP game, so, in all honesty, balancing anything for MP is a waste of developer resources. No offense, but I'm telling it like it is.
EDIT: In any case, "balancing" a historical setup for MP is intrinsically a waste of time. Historical setups and balance do not in any way mesh. It's impossible. The Allies always have much more IC than Germany, so Germany can never win except through being played by a much more skillful player...which is pretty much the definition of unbalanced. If you want balance, you need a fantasy scenario, like one of those in Armageddon (never played them, but I like the idea in principle).
EDIT2: Plus, players in MP do not in any way play like historical leaders. If UK is seeking an understanding with USSR and funneling them supplies from 1936, your balance is already ruined.
EDIT: In any case, "balancing" a historical setup for MP is intrinsically a waste of time. Historical setups and balance do not in any way mesh. It's impossible. The Allies always have much more IC than Germany, so Germany can never win except through being played by a much more skillful player...which is pretty much the definition of unbalanced. If you want balance, you need a fantasy scenario, like one of those in Armageddon (never played them, but I like the idea in principle).
EDIT2: Plus, players in MP do not in any way play like historical leaders. If UK is seeking an understanding with USSR and funneling them supplies from 1936, your balance is already ruined.
I think the Finns during the winter did a few nice large ambushes.
You didn't name a division, so there's no way of telling if it had more than 10,000 troops. To clarify further, it should be an ambush in the sense that adjacent (in HoI3) large enemy formations were completely oblivious to that movement. It's not really an ambush if you know they're there. So I am holding on to my cookies for now
Battle of Changsa was a large ambush of an entire Japanese army. Granted you had a pretty dense Japanese commander who had nothing but contempt for the Chinese and did not even bother with even the rudimentary safeguards concerning his flanks or anything that resembled recon.
I think partisans and dissent should effect the amount of intel you gather. . One of the things partisans where used for was gathering Intelligence on the enemy and it was considered more valuable by the Western allies then shooting axis soldier or blowing things up.
And I for one would like to be able to build reconnaissance aircraft to send on patrol
And I for one would like to be able to build reconnaissance aircraft to send on patrol
Yes, I was thinking about that too. Antwerp needs to be a port ingame, looking at the thames it could be a solution for the Antwerp problem. Make a small strait through the netherlands towards Belgium with Antwerp as a 10/10 Port. This will also do wonders to the realism of the Netherlands, the delta will be portayed correct and that just leaves the belgium sea line.
I saw they improved it but just a quick glance on a map shows you it's still too small.
Keep up the good work though Paradox, I can't wait for this game!
This is the most awesome development I have seen so far. It seems like it will really add an element to the game that didn't exist prior - the ability for the AI to take both proactive and reactive actions.
I'm so very happy about this.
I'm so very happy about this.
It appears at a glance that terrain is impacting the radar/intelligence, which is a good thing. Hopefully this is the case.
I cant stand that, its a complete nonsense, what the hell is balancing for MP? It doesnt matter if it is MP or single player game, normal setup shoould be well balanced for each side. To make a well balanced game u need to have players testing posibilities cheats and gaps in the system and what is better then MP test with every player trying to maximaise its own country capabilities?
Human will play as human so ai should be learned to act as a human, or u will have silly engine beeing easy to be cheaten by even unskillfull player. And game that was never tested in short ot long MP run - HoI1 release version.
Also there is very little effort needed to get those 5% more clients
.
I cant stand that, its a complete nonsense, what the hell is balancing for MP? It doesnt matter if it is MP or single player game, normal setup shoould be well balanced for each side. To make a well balanced game u need to have players testing posibilities cheats and gaps in the system and what is better then MP test with every player trying to maximaise its own country capabilities?
Human will play as human so ai should be learned to act as a human, or u will have silly engine beeing easy to be cheaten by even unskillfull player. And game that was never tested in short ot long MP run - HoI1 release version.
Also there is very little effort needed to get those 5% more clients
If so little effort is necessary, maybe you could care to explain how to balance a historical scenario so that all sides are equally matched? And since it's so easy, how about taking on an even better challenge - balancing something like the 1944 scenario for MP. Unless we have some radically different ideas of what balance is, there's no way you can balance that and still stay within a reasonable distance from a more or less plausible depiction of the situation. In 1944, the Axis was beat. There's no way to make them "equally matched."
EDIT: Also, if we have AI actually have the same motivations as human players, the game will start with a declaration of war by Mexico on Guatemala. Every. Single. Time. Because that's how 90% of all players play the minors. (naturally, every single one of these players still expects the AI to act historically)
Sure, 1944 is a kind of scenario where one side is close to the edge. Despite that it still can be a great fun in MP game.
From the other hand u have 1936 and 1938 scenarios and even 1939 scenario that are all very interesting and more or less enough balanced for MP games.
Alexander Seil said:
Thats the point (although i never saw Mexico beeing played by human), AI is a thing from the other world. I will do that my way, but she will anyway act as regular and silly as alwayes.
Let's leave the multiplayer for another discussion.
It is hard to balance intrinsically, as others have pointed out. Also, just which countries are played by real players and which aren't is a major determinant of player behavior. Thus, not something to be modified by AI behavior and thus not suited to this discussion.
It is hard to balance intrinsically, as others have pointed out. Also, just which countries are played by real players and which aren't is a major determinant of player behavior. Thus, not something to be modified by AI behavior and thus not suited to this discussion.
The point raised is simply:
How do we balance the amount of actionable intelligence gathered during MP games?
This has consequences in SP because if done right, the AI should see and act on similar information as that consumed by a human player.
How do we balance the amount of actionable intelligence gathered during MP games?
This has consequences in SP because if done right, the AI should see and act on similar information as that consumed by a human player.
My concern is not what intelligence advantages the AI should or shouldn't have over humans...it's what the AI does with that information. If the AI doens't treat a large armour concentration any differently then a concentration of militias and garrisons, intelligence is meaningless to the AI. Basically the only thing to do is spending thousands of hours testing and tinkering with the AI. If only the AI files were written in C++ then I could help...
- Status