|
|
Skeggy
Promising
Famous Hero
|
posted February 18, 2017 07:11 PM |
|
|
Google translation from
http://ru.nival.com/forum/index.php?threads/30797/page-16
It has been the campaign Heroes of Might and Magic 5, beat all the original scripts and laptop win at the highest difficulty level without difficulty. You can somehow make it stronger and more intelligent? It is possible to adjust the difficulty of the passing game?
Answer:
Yes, you can. It will have to dig a little in the AI characteristics. You will find them in section AI DefaultStats.xdb file that is in the directory GameMechanics RPGStats (data.pak archive). Open the file in the "Notepad" and get Combat subsection, which spelled out the correlation of forces of the opposing sides during the battle. There EnemyDamageCoef option - damage coefficient caused by the troops of the computer (as a percentage). Enter here a value in the region of 110-120 (or even higher - initially worth 100). As a result, the computer will increase the army.
Attribute EnemyTroopsHealthCoef (coefficient of enemy troops health), assign a value of 50-60, and the characterization EnemyWarMachinesHealthCoef (ratio of strength of the enemy combat vehicles) - 1,5.
Go to Adventure. He describes a computer action game on the map, and is divided into three small blocks: DifficultyNormal (refers to the low level of difficulty), DifficultyHard (average complexity) and DifficultyHeroic (high complexity). If you are an experienced player, then you are only interested in the last block.
The first indicator that you need to edit - AttackHumanAgressivityCoef, it determines the aggressiveness of computer players to attack. Please note that all three levels of difficulty value for this parameter is 1. Increase this value to 1.2-1.5, then the computer will attack you as soon as possible.
The next attribute - AttackComputerAgressivityCoef, defining the aggressiveness of one race, a computer-controlled, with respect to the other. If you want your computer opponents were cut in full to each other, put here the value of 1-1.5. But note that this change does not always make AI more serious opponent. More often than opponents in the early rounds start civil wars and for a while forget about you. But then one of them wins, captures castles and beat it becomes extremely difficult. In our opinion, the value of the characteristics of the best reduced to 0.1-0.2.
All units have CowardiceVersusHuman setting. It identifies the computer cowardice to fight with you. Set the value to 0. The cowards have no place in our mini-modification!
Go ahead. ArmyPowerCoefBonusPerHeroLevel - army computer power factor depending on its level. At all levels, the complexity of its value is the same. In blocks DifficultyHard and DifficultyHeroic should prescribe that attribute values of 0.2 and 0.3 respectively. In this case, medium and high complexity significantly increase strength dependence of the computer player's level.
Indicators AvengerVersusHumanCoef AvengerVersusComputerCoef and revenge set against the computer player to you and the computer opponents, respectively. In the first case, the value is the same for all levels of difficulty - 1. Sections DifficultyHard and DifficultyHeroic recommended to register the number of 1.25 and 1.5. The computer will become very vindictive.
Definite interest also configure HireHeroesPerTown and HireHeroesGlobalLimit. The first specifies the maximum number of characters from the same town, which the computer can run simultaneously. The second - the maximum total number of characters that are subordinate to the AI. If you want to make enemies captivated a whole map, place the attribute value HireHeroesPerTown more. But note that in this case a really strong opponents you will discover yourself unlikely (all the characters are very weak), and the game will be very slow.
As for the index HireHeroesGlobalLimit, the corresponding value is best reduced to 12-15. The fact that AI is not able to normally operated with a large number of characters. He brings a lot of weak opponents.
By the way, if you are not a professional player, and just starting to explore the world HoMM 5, you can do a reverse operations described - play will be much easier.
|
|
magnomagus
Admirable
Legendary Hero
modding wizard
|
|
posted February 18, 2017 07:56 PM |
|
|
Thank you for digging up that post, I should add however that Quantomas warned me lots of values in defaultstats.xdb are no longer used, but I know "HireHeroes.." works so there may be more. I would like to hear
from anyone toying with these values if the can make an improvement so I can put the best values in the mod.
____________
MMH5.5 Downloads | MMH5.5 Translations | MMH5.5 FAQ
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 18, 2017 08:43 PM |
|
|
@Skeggy: Thank you very much for finding this! Well done!
After reading your description very carefully, and after finding a couple of other sources for various parameters I have made the following adjustments to my game. Note that I'm only interested in impossible difficulty now.
<DifficultyHeroic>
<AttackHumanAggressivityCoef>1.3</AttackHumanAggressivityCoef> <!-- def:1 -->
<AttackComputerAggressivityCoef>0.15</AttackComputerAggressivityCoef> <!-- def:0.25 -->
<CowardiceVersusHuman>0</CowardiceVersusHuman> <!-- def:0.3 -->
...
<AvengerVersusHumanCoef>1.5</AvengerVersusHumanCoef> <!-- def:1 -->
...
<CheckTownsInDanger>1</CheckTownsInDanger> <!-- considering to set this to 2 -->
...
<HireHeroesPerTown>3</HireHeroesPerTown> <!-- def:2 -->
<HireHeroesGlobalLimit>12</HireHeroesGlobalLimit> <!-- def:17 -->
...
</DifficultyHeroic>
I will put this to the test asap. Hopefully we get some good results.
Edit : Added another parameter i changed
|
|
magnomagus
Admirable
Legendary Hero
modding wizard
|
|
posted February 19, 2017 01:22 AM |
|
|
|
Skeggy
Promising
Famous Hero
|
|
posted February 19, 2017 03:08 AM |
|
|
cichy said:
Changed labels
Clearing "artifactIDs" node, when swapping artifacts.
Can enable scripts
Edit: repaired group ratio
In MMH55_Editor_64 there are three sections in the general tab: Quantity, mood and courage.
Mood section has four mutually exclusive options: friendly, hostile, aggressive, wild.
Courage section has three mutually exclusive options: always join, always fight, flee or join.
If some monster has defined courage as „Always join“ then mood section of the monster property window is disabled. Same thing for “Always fight” option. Only if “Flee or join” option in “Courage” section is selected, only then mood section is active.
I suppose that mapmakers used courage section more frequently then mood section because if some monster is defined by “always fight”, it will fight no matter what. Same thing goes for “always join” option.
So, is it possible that new creatures and new groups inherit courage and mood properties, and if not mood, then just courage?
|
|
Mario
Known Hero
|
|
posted February 19, 2017 06:49 AM |
|
|
CobraL0rd said:
I have found this website: http://www.bonddisc.com/ref/h5/
So, apparently Quantomas is working on a different game now. And from what i notice he is not sharing the source code for HoMM 5.5 AI. Also, i can't find a way to contact him..
You can try to contact with Quantomas on celestialheavens.com
This is his last post:
https://www.celestialheavens.com/forum/16/15952?start=60#p366630
|
|
cichy
Hired Hero
|
|
posted February 19, 2017 09:56 AM |
|
Edited by cichy at 09:58, 19 Feb 2017.
|
Skeggy said:
cichy said:
Changed labels
Clearing "artifactIDs" node, when swapping artifacts.
Can enable scripts
Edit: repaired group ratio
In MMH55_Editor_64 there are three sections in the general tab: Quantity, mood and courage.
Mood section has four mutually exclusive options: friendly, hostile, aggressive, wild.
Courage section has three mutually exclusive options: always join, always fight, flee or join.
If some monster has defined courage as „Always join“ then mood section of the monster property window is disabled. Same thing for “Always fight” option. Only if “Flee or join” option in “Courage” section is selected, only then mood section is active.
I suppose that mapmakers used courage section more frequently then mood section because if some monster is defined by “always fight”, it will fight no matter what. Same thing goes for “always join” option.
So, is it possible that new creatures and new groups inherit courage and mood properties, and if not mood, then just courage?
Only mood is changed, courage remains same. So if mapmaker create stack before town, that will always join, then it should join.
Mood section is there, because RMG creates creatures with mood "hostile" or "wild". That's almost like "always fight".
As I remember, then if you add creature in editor, then it has mood "aggressive".
So tool has default values "aggressive" and "hostile".
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 19, 2017 01:08 PM |
|
|
@Mario Thanks, I will keep that in mind.
So i completed a game (again) on impossible difficulty, very large map, 4 players with these new settings.
First and foremost by increasing HireHeroesPerTown to 3 turn time is greatly increased, i think it tripled for me and i could see that from the start of the game. I continued playing with for about an hour and i also realized that it bugged the game. The enemies made so many heroes after a while that the game bugged. So after wasting almost 2 hours i changed it back to 2 and started a new game again.
Now with all the other changes i mentioned above, except HireHeroesPerTown, turn time was a little bit longer, the game had become more difficult for sure. I felt the difference from the start. First of, the enemy wiped out an AI player at month 3, when under "normal" Impossible difficulty this would happen around month 5. Thus, we can deduce that, as mentioned, the AI has become more unpredictable and possibly vindictive.
And i realized that in many other parts of the game as well. In general the AI kept "pushing" and reaching to me closer than ever before. I felt afraid at times, and kept retreating my weak secondary heroes back to base, because the enemy would undoubtedly start a killing spree.
In short I used the same strategy i used before, but the game had become noticeably more difficult and somehow "aware" of my actions. Unfortunately it didn't become really more intelligent, but the AI now didn't fear me as much, it did manage to attack a somewhat strong secondary hero that i had outside of town and killed him easily (i did a silly move there and payed for it, which is good) but it still did not pursue an attack with my second best hero that i had inside the town. But the AI wasn't afraid to stay around that town and see whether i could make a mistake and leave the town or something.
When i brought my primary hero on the 1st enemy base, the enemy hero would finally not have left the town! What happened until now, is that all enemy heroes would leave their town empty when they saw me coming. This time it didn't happen. The "Threat" level was challenging. I did not attack at first, because i realized he was very strong and also he was in town. This "scared" me and i liked that. I wondered how much more the AI could push. So i needed to fill my high tier units. I did a "Summon creatures" spell and resupplied. When i did, i was close to the enemy town and was ready to attack it after that turn, but unfortunately the enemy hero saw that and left! It realized the match would be very risky so it didn't stay. But it is important to mention that he left when he could, i wasn't "blocking" the way. It was able to escape and it did, i dk what would happen if he couldn't escape within that turn.
I also noticed that AI were attacking and clearing dungeons relentlessly, because when i reached them with my main hero they were empty (it said "there's nothing here at this time" sth like that).
To conclude, yeah, the AI was definitely brutal now, more unexpected, more reckless, significantly less afraid of the player and / or other AIs, but nothing experienced strategy players will have a very hard time on. Overall the match took longer to complete, about an hour longer than most such games (which normally take around 7 hours for me). I will keep these settings, i believe if i push them even more the AI will "collapse". I'm happy with this and i believe there's not much more we can do at this point. I think the game is hard enough and it will do.
Finally, it would be interesting to know what this parameter does: CheckTownsInDanger . From its descriptive name, i understand it has something to do, with how many Towns the AI can "keep up with", how many they should be aware of. It's always set to 1, and this may be a simplistic explanation but i can't find another one. Anyway, i don't think it will matter a lot, but just leaving it here, in case anybody wants to try something.
All other parameters change life of enemy creatures, and the growth of AI, which are not strategic / intelligence factors per say. They are cheating factors and i don't like that.
Finally, one could try and push these: <AttackHumanAggressivityCoef> , <AttackComputerAggressivityCoef> to 1.5 and 1 respectively, and the AI would be even more reckless and whatnot. But i still don't expect miracles. The game AI apparently can't be pushed further, it needs complete redesign for that.
I do recommend these parameters be changed though to reflect more to the new ones (at least in impossible difficulty). Because obviously someone who pursues impossible difficulty really wants something exciting and "dangerous".
That's about all. I'm out.
|
|
magnomagus
Admirable
Legendary Hero
modding wizard
|
|
posted February 19, 2017 01:27 PM |
|
|
@Corbalord: Great!
Quantomas already disabled any cheating parameters.
You should leave all hireheroes settings as it was in H55, i already tested those to be optimal.
EDIT: I also wonder what is difference between cowardice and aggressiveness?
____________
MMH5.5 Downloads | MMH5.5 Translations | MMH5.5 FAQ
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 19, 2017 02:18 PM |
|
|
I've performed some more tests.
At first, i thought cowardice was the probability for the enemy to "Flee" games it's not going to win. I'm not sure of the difference still, but..
But cowardice along with Aggressivity parameters, as i mentioned, it's just that the enemy AI is being able to sort of "push" closer, more into "human" territory without being afraid of consequences. Only significantly stronger enemy Human / AI heroes will make AI heroes fall back into a safe zone (i consider safe zone a place where human hero can't reach enemy hero within the turn). Just change them both to be sure. I think they affect each other as well.
From further tests i did i believe the AI fears most in descending order of magnitude:
1. Army Numbers in general
2. Hero Stats
3. Tier greater than 3 army numbers
4. High Tier Caster army numbers
5. Shooter army numbers
6. Caster / Shooter Army numbers
7. Hero Level
8. Tier 7 numbers
9. Whether hero is inside town or not
10. Whether hero has received temporary "next match" bonuses from the map. I have reasons to believe this is true. The AI places empty hero outside the town just for my primary hero to lose his bonuses. This happened many times.
This should be accurate.
Also note that parameters AttackHumanAggressivityCoef and AvengerVersusHumanCoef don't have any effect above 2.
Overall the new changes modify AI recklessness and “awareness”. These are welcome changes.
By the way, note that the AI can see almost the entire map in impossible difficulty and acts accordingly (i have no doubt about that).
|
|
magnomagus
Admirable
Legendary Hero
modding wizard
|
|
posted February 19, 2017 02:23 PM |
|
|
Quote:
At first, i thought cowardice was the probability for the enemy to "Flee" games it's not going to win. I'm not sure of the difference still, but..
ok so when you had set cowardice to 0 you did not notice a difference in AI fleeing behavior?
I tend to think that the aggresivitycoefs are only relevant as the difference between vshuman and vscomputer, and cowardice regulates the actual courage of the Ai in attacking?
____________
MMH5.5 Downloads | MMH5.5 Translations | MMH5.5 FAQ
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 19, 2017 02:42 PM |
|
|
magnomagus said:
Quote:
At first, i thought cowardice was the probability for the enemy to "Flee" games it's not going to win. I'm not sure of the difference still, but..
ok so when you had set cowardice to 0 you did not notice a difference in AI fleeing behavior?
When i said fleeing, i meant while in a fight. You know, the probability they will "Flee" (with their forces intact and the hero left for hire - if they can afford it) In general this AI fleeing ability is very accurate in impossible difficulty. I wouldn't want to disturb it.
magnomagus said:
Quote:
I tend to think that the aggresivitycoefs are only relevant as the difference between vshuman and vscomputer, and cowardice regulates the actual courage of the Ai in attacking?
Since the cowardice towards computer was 0 before those changes were made, i would say that aggressivity coef was the crucial factor. This was the factor that made the AI want to expand more and much faster, be more reckless. And that was a great substantial difference. The AI wiped each other out much faster and had greater armies while recruiting from multiple towns faster.
Cowardice should have only made a difference towards human / me and it did. The AI had more heroes reaching closer to my base and capturing my outposts and mines, not trying to attack my main base (unless i extracted my strong hero out) but more to harass me etc.
I'm not sure what else i can say about that. But, cowardice should definitely be 0 for both human and ai @ impossible diff.
The crucial parameter was aggressivity though.
And don't forget AvengerVersusHumanCoef and AvengerVersusComputerCoef parameters. Those should be high as well. When i killed a few secondary AI heroes, i noticed the AI to be more "alert".
It's hard to extract more precise information out of these since we don't have the actual explanation of the parameters and we can only do experiments. But i think, this will do now. The game really was harder and AI displayed additional intelligence. Although it's not game changing, it's enough (for me at least).
|
|
magnomagus
Admirable
Legendary Hero
modding wizard
|
|
posted February 19, 2017 03:17 PM |
|
|
|
thGryphn
Promising
Famous Hero
|
|
posted February 19, 2017 03:35 PM |
|
|
magnomagus said:
If the AI becomes more aggressive vs AI then the game actually becomes easier ...
I believe this is true as well. In most, if not all cases, it's better (more challenging to play against) if all AI players were in full collusion/cooperation against the human player even if they're not technically teamed-up.
@magno, maybe devise a higher difficulty setting that increases the aggressiveness of AI against human (with aforementioned parameters) and make them cooperate with each other (or even team them up)? Call it "Masochistic" level 
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 19, 2017 04:04 PM |
|
|
magnomagus said:
If the AI becomes more aggresive vs AI then the game actually becomes easier so I disagree on vsComputer values.
No this isn't always true. I will explain to you why.
It could be true in some circumstances, but i don't think it applies here, or if it does then maybe a few times.
First and foremost, this isn't what i saw in the game, because clearly the AI was much stronger and had gathered bigger armies, than before and FASTER. How could this happen then? If the AI was fighting among each other all the time, they would have smaller armies.
Well this is because apparently in HoMM 5 aggressiveness doesn't imply that the greatest heroes of two AIs clash and as such, only one will win and the other will lose the game. The AI in general is cowardish, even against other AI. What most likely (i really believe that, it happens in at least half the games i've played) happens is that, say we have two AIs, one of them will bring its heroes towards the other, they will try to capture enemy positions, outposts etc. This aggressive stance happens by the strongest of the two. Because the AI is cowardish the weakling will fall back towards its castle and bring back all its heroes as back from the enemy as it can (it always does this!).
Now the defending AI will feed all its troops to the main hero who will most likely be inside town. Here's the scenario. If both sides are about equal (which is hard to happen, since they spawn in different locations in the map and are of different race, maps are usually very varied, tons of things could happen around etc, etc) then the defending hero will indeed stay at its castle and fight the incoming threat from the ai. As such both will lose a lot of units. As i explained this is hard to happen, because:
1. The AI is in general cowardish as I have noticed (really i saw it, ie. i have opened the fog of war in entire games, just to see what they do),
2. It's hard for the two ais to be of about equal strength. Too many factors are in play as i explained,
Now, regardless of all this, suppose this happens. This still doesn't guarantee that the human player will have it easier, since now that one AI has lost (not yet, it probably has other heroes with no armies in the back, but will lose in a week afterwards with no trouble) the other AI has two towns at his disposal and an entire area where he can in general do more stuff. Plus his unit growth is doubled.
[Also, if having the setting "casual game", the AI will convert the town, since it probably has many resources]
But again, i strongly believe this is unlikely. What the defending / weaker heroes will do is abandon their town, or leave just a very small army behind and leave to somewhere else on the map, since they realize that they can't win (again it all boils down to this: "The AI is in general cowardice").
What follows, is the most absurd thing: Since that AI has realized it can't win, it will venture to various areas back and forth in the map (note: now this AI has no other town), while the other AI will try to secure his town that he has just captured. The latter will try to hold it for at least 7 days and then (if we're not talking about a huge map) the former (the weaker / more cowardish AI) will just lose the game, becaue it won't attack at all! Thus the winner hero won't lose any units at all! This is the most stupid thing the AI does, as i mentioned multiple times. The AI would rather DIE than challenge the enemy hero, be it human or other ai. That's what's stupid. So now the winner ai, get another town, double growth, double outposts, more heroes to feed the primary hero with troops, expand, explore whatnot etc etc.
I'm almost certain this is what happened with my latest game now. The enemies had much greater armies! And i reached them on month ~ 4.5 and they have killed this other AI on month 3 as i said. Similar things were happening before i tell you. It's just that with these changes on impossible difficulty, things are harder and more unpredictable now.
It was a long explanation.
|
|
magnomagus
Admirable
Legendary Hero
modding wizard
|
|
posted February 19, 2017 05:00 PM |
|
|
Quote:
First and foremost, this isn't what i saw in the game, because clearly the AI was much stronger and had gathered bigger armies, than before and FASTER. How could this happen, since if the AI was fighting amongst each other all the time, they would have smaller armies?
you cannot gain more armies by fighting, unless it results in an extra native town, but the purpose of 4 player map is to have multiple good fights, not to have one AI kill the others and fight one huge boss. Then it is better to play vs 1 player and maximize army size cheating.
In some cases setting all AI to defensive and allied is the hardest, they won't attack, but now you have the problem to kill the first 2 so efficiently that you can still beat the last one.
____________
MMH5.5 Downloads | MMH5.5 Translations | MMH5.5 FAQ
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 19, 2017 05:07 PM |
|
|
|
Ah yes, i didn't mention that before. All the AIs i've played with are offensive.
|
|
CobraL0rd
Tavern Dweller
|
|
posted February 19, 2017 06:00 PM |
|
|
|
One more thing. The AttackComputerAggressivityCoef parameter is probably not just for the enemy AI's but for the neutral creatures, dungeons, towns etc. as well. Which explains some things I've said above, which i consider good.
|
|
Skeggy
Promising
Famous Hero
|
|
posted February 19, 2017 06:30 PM |
|
|
cichy said:
Skeggy said:
cichy said:
Changed labels
Clearing "artifactIDs" node, when swapping artifacts.
Can enable scripts
Edit: repaired group ratio
In MMH55_Editor_64 there are three sections in the general tab: Quantity, mood and courage.
Mood section has four mutually exclusive options: friendly, hostile, aggressive, wild.
Courage section has three mutually exclusive options: always join, always fight, flee or join.
If some monster has defined courage as „Always join“ then mood section of the monster property window is disabled. Same thing for “Always fight” option. Only if “Flee or join” option in “Courage” section is selected, only then mood section is active.
I suppose that mapmakers used courage section more frequently then mood section because if some monster is defined by “always fight”, it will fight no matter what. Same thing goes for “always join” option.
So, is it possible that new creatures and new groups inherit courage and mood properties, and if not mood, then just courage?
Only mood is changed, courage remains same. So if mapmaker create stack before town, that will always join, then it should join.
Mood section is there, because RMG creates creatures with mood "hostile" or "wild". That's almost like "always fight".
As I remember, then if you add creature in editor, then it has mood "aggressive".
So tool has default values "aggressive" and "hostile".
Combination 1: On mapmixer “swap” is enabled together with “random blocks”.
Result: any kind of courage settings that every defined (defined in a meaning non-random monster, it does not look like a box in map editor) monster had, is reset to “Flee or join”, resulting in mood distribution according to “creatures mood” section of the mapmixer.
In combination 1 any random monster (one that looks like a box in the map editor) has completely intact courage and mood settings.
Combination 2: On mapmixer “swap” is enabled together with “random blocks” and together with “only random blocks”.
Result: nothing really changes, everything stays the same, non random and random monsters looks the same and have same courage and mood settings. Everything is the same, doesn’t matter how “creatures mood” section is defined.
Combination 3: On mapmixer “swap” is enabled together with “only random blocks”.
Result: same as combination 2.
Combination 4: On mapmixer only “swap” is enabled.
Result: same as combination 1 with addition of groups. Percentage of groups depends of on the “mixed stacks” slider.
Groups are formed only from non-random monsters, they do not inherit courage, their courage settings is reset to “Flee or join”, resulting in mood distribution according to “creatures mood” section of the mapmixer.
|
|
Elessar
Adventuring Hero
|
|
posted February 19, 2017 06:44 PM |
|
Edited by Elessar at 18:59, 19 Feb 2017.
|
Very early in my game (for other reasons wanted to see how things would play out if there were only 8 towns total)-impossible sized map at the highest difficulty. Have cranked up all of the given aggression parameters mentioned above to 1.5, while lowering the cowardice ones down to zero. I have strong kingdom guards on this map so it will be awhile before I can get a feel for things. Even then note it would just be a sample size of one...
magnomagus said:
If the AI becomes more aggresive vs AI then the game actually becomes easier so I disagree on vsComputer values.
I disagree-when I read this recent discussion my ears really perked up. My biggest issue with the AI has been how the mains of each side would simply "ghost" and utterly ignore each other. On a map with a lot of towns towards the endgame they would thus suffer from "conversion fatigue"-not only is it 100 grand to convert a town, but there would also be a lot of attrition (as towns and external dwellings are taken and retaken).
The end result would typically be my main walking in to invade the other AI's-only to find a patchwork of towns of various factions, and mains which (after the 90 day cheating bonus ran out) would be no match for mine.
While, if they actually grew a pair and would attack when they had the advantage, I would likely come in-and see one faction (after eliminating another) with a solid block of like 6 towns, and a very powerful main which I might actually find to be an actual challenge. This is how I recall things from my Heroes 3 days.
magnomagus said:
you cannot gain more armies by fighting, unless it results in an extra native town, but the purpose of 4 player map is to have multiple good fights, not to have one AI kill the others and fight one huge boss. Then it is better to play vs 1 player and maximize army size cheating.
Translation: Yeah, the stronger of the two will lose some troops in the big battle, but will now have a much larger base from which to expand further. I simply have never seen the huge AI empires that I recall from playing H3, for this reason. And those would typically make the game very challenging, since I had then to build up my own empire to counter them.
If you really want 4 weaker AI's, vs. one huge and very dangerous one, and think the former is more of a challenge, I'm not sure what else to say...
Cheers to Cobralord, only saw his post after I had penned mine.
|
| |
|
|
This Popular Thread is pages long: 
