Why do many players consider HoI4 to be easy/too easy?

[RikiBreeiki]

Neither DOTA nor Starcraft is "complex" games... they do require some tactical skill and knowledge of mechanics and good reflexes... but complex... NO.

How about attacking to pin an enemy down for a few hours or a day or two?!?

How do you teach an AI this without huge amount of training... what is loosing a single combat to the overall strategy or tactical situation?

How do you know that you are loosing more equipment than you opponent before a fight.... you don't know how air assets will impact the combat, reinforcements... weather or other factors will play into the final result. These are things very difficult for an AI to understand not to mention all the bonuses and penalties that can effect it by the tactics employee by the commanders in combat.

It is REALLY difficult for a machine learning AI to understand that something I do now will have HUGE impact in a few HOURS of game play down the line. In DOTA and Starcraft it is WAY easier to calculate what constitute a positive or negative feedback to the network, a nightmare in HoI.

You don't realize how different these games really are?!?

DOTA and Starcraft are NOT complex games, more complex than Chess but a game like HoI (Paradox games in general) play in a separate league in comparison... just accept that fact.

i heavily doubt you understand the games you are talking about , or you played them at a level that can be considered serious,

For hoi4 mechanics are simple, calculations are extremely simple, any person that knows what breakthrough defense attack and multipliers , reinforce rate means ,will easily predict if he will win combat or not places to defend are always in front of your eyes, you can simply know how much you will hit ,you can easily know all modifiers from cas air support-planning-force attack - general skill- terrain and guess results of the battle super easily, calculating if i will lose more equipment? lol you joking ? anyone with basic maths will have no issue understanding if he will win combat or not,, Ai will never have issue calculating its damage compared to enemy defence/breakthrough and will have no issue calculating modifiers, in fact these are super easy for ai, and movement is also limited , now let me count you few small stuff among the billion stuff ai has to cauclualte in dota2, you wont understand anything if you have not played the game, and even if you did you still may not understand ,

In dota2 you try to deny your own creeps, while killing enemy creeps and every decision you make while last hitting is a trade , in a sitaution where our ai is gonna deny , our ai sees that his allied creeps have 60 and 63 health , and enemy hero hits 50 and ai hits 50 too, which creep our ai will deny, how does he decide? let me count you some of th factors while our ai just have to last it
1-) calculates distances of both heroes to the creeps
2-) calculates projectile speed
3-) calculates attack animation
4-) calculates turning speed
5-) callculates backswing from attack animation in case someone goes for double hit on a creep ,
6-) calculates which creep enemy human is gonna hit to depending on factors like how humans choose their target ( eg human deciding to hit unit that is closer to unit it hit before so he moves his mouse for a shorter time ) ,
7-) calculates how much creeps are hitting to eachother
moves during backswing of attack animation to get closer to the creep in order to reduce the timeframe his attack projectile will fly
9-) at same time calculates bunch if x-y-z abilities from enemy hero are threat and damage-harrassment he takes will be worth it if he geos for deny
10-) calculates where enemy roaming heroes can be since it is a laning stage, our ai wont sucide to last hit a creep, will go back if heroe are mising

guess what humans try to do all i wrote above too! and they are just trying to last hit a creep, hoi4 players choose focus queue buldings and start smoking weed, ai would do same too


i can count 1000 stuff to perfectly last hit, but dota 2 is not a game about afk last hitting.

in hoi4 i can personally click on battle and look at enemy modifiers once and from that time know strenght of enemy general, which doctrine he is on, how entrenched he is ,affects of air superiroty and cas as modifiers, and cas as damage, and easily make my decision,

dota2 is one of the most complex games around, also far more complex than game it is being compared to named leauge of legends due to having less straight up calculated mechanics and being more based on human decisions map awarness , requiring more coordinated teamwork having a bigger varierty of skills, it is also choosen for a reason. eu4 is not a complex game either, there are bunch of non deep simple min maxing tabs that are designed to be simple easy to understand with not many trade offs, and is filled with insane rng, hoi4 on other hand have no significant rng

 

[Shinkuro Yukinari]

dota2 is one of the most complex games around, also far more complex than game it is being compared to named leauge of legends due to having less straight up calculated mechanics and being more based on human decisions map awarness

This statement varies from champion/hero to champion/hero. Ursa, Volibear, Invoker, Cassiopeia, for example

 

[agonistes]

Nothing wrong with that. I exploded into Russia with infantry only. In fact, 95%+ of my offense is infantry.

Tanks suck. Best use for them is throwing a light or two into the occasional infantry division. Otherwise you are just pissing away good factories.



Noob here...
I mowed down Russians on my way to Moscow while fighting a 3 front war. Russian divisions are fairly weak. Spainish have been much tougher. With the Russians, my 'skill factor' shot up once I started mixing up division types within each army. It wasn't intentional, I was rushing to close Romanian gaps and had to zerg troops out of my training camp and make hodgepodge armies.

Japan hates any divisions wider than 20 with a light tank in it.

So far, the toughest ai has been Germany, when I play as Russia. You need to get piercing up fast, and you need to make that line nice and thick and well fed.

Spain is second. They have very determined divisions and shred the Italian ai.

Japan gets honorable mention, 'cause they were nasty until I realized they were just using bigger divisions.

I wouldn't bother too much with tanks. If your infantry isn't doing it, try adding a couple more brigades (christ don't go to 40x, that's insane). Mix up the brigades, but probably at least 1 should always be infantry.

And don't neglect manpower, supply, reinforcements, and training. Screw planning. While you are planning, the enemy is positioning and entrenching. If you continously stop to plan, you will be fighting divisions the Soviets brought all the way from Kamchaka before you reach Moscow. Once you push, keep pushing until you see your supplies/HP getting low, then switch out that army with one of your reserve armies (obv send reserve army forward first), then start pushing again.

Once any country cracks, they all open up like eggs.

Tanks. Spare me.

If you *are* using tanks, please tell me you aren'l wasting rubber on motorized.

You will conquer with efficiency.

Yes, I expected maybe I would see a dogpile on me of 'respectfully disagree.'

Happens to me alot in the civ forums too - I never seem to be playing the same dang game as everyone else. But I'm not a troll. If it wasn't working for me, I wouldn't post it.

Efficiency wins. Same as in Civ. People who are building 40x, or using numerous tank divisions, are being inefficient. They are winning, but they have in a sense already won because they have factories (and resources) to spare.

But if you are in that margin where victory is uncertain, I humbly suggest making your divisions no wider than necessary to guarantee victory. Which tends to be 2-3 brigades wider than your opponent's divisions.

40x is nice where stats stack, poor where they spread. Nice where 1 support is more efficient, poor where they are not (towed guns).

1 40x has stacked HP, breakthrough, ORG, and defense, but its also more likely to be drawing 2x the attempted shots than 2 20x. It also can only be in 1/2 the provinces that 2 20x can be in.

If you have the factories to mix 40x and 20x and 10x, and throw armor around, the outcome shouldn't be in doubt, and 40x aren't needed. See? At the point when you can get away with using them, they are already a luxury, not a necessity. SP.

Before that point, I'm not going to hold ground or break lines with a 40x if I can do it with less than 40x.

I repeat, I am a noob.

 

[RikiBreeiki]

This statement varies from champion/hero to champion/hero. Ursa, Volibear, Invoker, Cassiopeia, for example

for humans i think they are close but for ai dota requere more humane calculations while lol requiere more mathematical calculations

 

[Secret Master]

Efficiency wins.

Do you mean efficiency in terms of equipment spent to take X number of provinces?

Or do you mean efficiency in terms of more days occupying more factories?

These are two very different kinds of efficiency. Spending less equipment to take enemy real estate is certainly valuable. No question there. But time is a resource, too. Occupying 100% of Poland's IC in 16 days instead of 36 days nets you more production sooner.

It gets more profitable when looking at Benelux and France. Occupying 100% of their IC within 60 days of starting the war, as opposed to 100 days of starting the war, makes a production difference. Given the generosity of HOI4's occupation laws, seizing all of France's European factories is a huge boost to Germany. The CIC in particular is the gift that keeps on giving in the build up to Barbarossa. Sooner is better once the shooting starts.

And when it comes to attacking the Soviet Union, time is once again a factor to consider. The faster the Soviets lose factories, and Germany gains them, the less damage the Soviets can do. Then there's the oil to consider; denying the Soviets oil

This also leads to the issue of encirclements against the AI. You can bag a ton of enemy divisions without heavy losses through judicial use of panzer forces and planes. Because of how breakthrough works on panzer divisions (especially with proper generals with the right traits), they can often just brush aside enemy forces with almost no losses. Once the enemy is cut off from supply, they can be defeated in detail easily and with fewer losses among all forces, infantry as well as panzer, since units out of supply function worse. Or you can just let them rot on the vine.

I'm sure some clever person has probably encircled half the Red Army at some point and left it alone while occupying all the VPs needed to force the Soviets to capitulate.

 

[TheMeInTeam]

I'm sure some clever person has probably encircled half the Red Army at some point and left it alone while occupying all the VPs needed to force the Soviets to capitulate.

Maybe someone has done it, but I'm not sure it's clever. Doing that leaves an enormous quantity of general XP on the table, and the casualties long-term out of supply divisions can inflict are almost nothing. Whether you want an extra guy instantly up to speed on bonuses or to keep pushing an amazing general, it's hard to justify not slaughtering such an encirclement wholesale in game terms because it costs so little.

 

[mursolini]

Nothing wrong with that. I exploded into Russia with infantry only. In fact, 95%+ of my offense is infantry.

Tanks suck. Best use for them is throwing a light or two into the occasional infantry division. Otherwise you are just pissing away good factories.



Noob here...
I mowed down Russians on my way to Moscow while fighting a 3 front war. Russian divisions are fairly weak. Spainish have been much tougher. With the Russians, my 'skill factor' shot up once I started mixing up division types within each army. It wasn't intentional, I was rushing to close Romanian gaps and had to zerg troops out of my training camp and make hodgepodge armies.

Japan hates any divisions wider than 20 with a light tank in it.

So far, the toughest ai has been Germany, when I play as Russia. You need to get piercing up fast, and you need to make that line nice and thick and well fed.

Spain is second. They have very determined divisions and shred the Italian ai.

Japan gets honorable mention, 'cause they were nasty until I realized they were just using bigger divisions.

I wouldn't bother too much with tanks. If your infantry isn't doing it, try adding a couple more brigades (christ don't go to 40x, that's insane). Mix up the brigades, but probably at least 1 should always be infantry.

And don't neglect manpower, supply, reinforcements, and training. Screw planning. While you are planning, the enemy is positioning and entrenching. If you continously stop to plan, you will be fighting divisions the Soviets brought all the way from Kamchaka before you reach Moscow. Once you push, keep pushing until you see your supplies/HP getting low, then switch out that army with one of your reserve armies (obv send reserve army forward first), then start pushing again.

Once any country cracks, they all open up like eggs.

Tanks. Spare me.

If you *are* using tanks, please tell me you aren'l wasting rubber on motorized.

You will conquer with efficiency.

Most efficient way to conquer is with planes and Larm+motorised. This division is ~4 times more expencive than infantry division, but it takes far less loses in long run. They also take objectives far faster, meaning they need to initiate combat far fewer times.

Yes, I expected maybe I would see a dogpile on me of 'respectfully disagree.'

Happens to me alot in the civ forums too - I never seem to be playing the same dang game as everyone else. But I'm not a troll. If it wasn't working for me, I wouldn't post it.

Efficiency wins. Same as in Civ. People who are building 40x, or using numerous tank divisions, are being inefficient. They are winning, but they have in a sense already won because they have factories (and resources) to spare.

But if you are in that margin where victory is uncertain, I humbly suggest making your divisions no wider than necessary to guarantee victory. Which tends to be 2-3 brigades wider than your opponent's divisions.

40x is nice where stats stack, poor where they spread. Nice where 1 support is more efficient, poor where they are not (towed guns).

1 40x has stacked HP, breakthrough, ORG, and defense, but its also more likely to be drawing 2x the attempted shots than 2 20x. It also can only be in 1/2 the provinces that 2 20x can be in.

If you have the factories to mix 40x and 20x and 10x, and throw armor around, the outcome shouldn't be in doubt, and 40x aren't needed. See? At the point when you can get away with using them, they are already a luxury, not a necessity. SP.

Before that point, I'm not going to hold ground or break lines with a 40x if I can do it with less than 40x.

I repeat, I am a noob.

Just out of curioucity, what do you optimise, when claiming about "efficiency"?
Production, I assume?

The issue is, the most valuable resouce in game, is manpower. You can conquer production and resources just fine, but "your" manpower is fairly set in stone. Hence, a lot of people have a habit of optimising manpower, not production. Anything that is attacking and isn`t tank based will take enourmous loses vs any decent opposition. Which is a critical problem, unless you are one of US, USSR, China or India.

It is extremely easy to just fill out the entire frontline in infantry up till supply cap, and then your opponent makes same, and you`re either at WW1 scenario, or you both invest into airforce and tanks.

 

[TheMeInTeam]

You can conquer production and resources just fine, but "your" manpower is fairly set in stone.

Not if you have the "together for manpower" DLC. Australia can trivially field and ultimately annex their way into having more active manpower (with lots in reserve) than their entire recruitable core population.

If manpower is set in stone this kind of picture is fishy:



Czechoslovakia has a core population of 15 million. Which means I have > 50% of core manpower sitting in a recruitable pool before even factoring the currently fielded armies.

With together for manpower, ANYBODY can get the bodies of USA/India/China/USSR in SP. Failing one of those, Brazil and Dutch East Indies each have a ton of pop to use too (which is why Australia can get so much).

Quite a few lesser countries can get you ~500,000 manpower at extensive conscription too. Argentina, Iran, Turkey, and Ethiopia are examples of countries like this (in fact several releasable African nations are, but they're usually under majors or Belgium, in which case you'd just hit NED if doing SP manpower plays).

 

[Alspego]

For hoi4 mechanics are simple, calculations are extremely simple, any person that knows what breakthrough defense attack and multipliers , reinforce rate means ,will easily predict if he will win combat or not places to defend are always in front of your eyes, you can simply know how much you will hit ,you can easily know all modifiers from cas air support-planning-force attack - general skill- terrain and guess results of the battle super easily, calculating if i will lose more equipment? lol you joking ? anyone with basic maths will have no issue understanding if he will win combat or not,, Ai will never have issue calculating its damage compared to enemy defence/breakthrough and will have no issue calculating modifiers, in fact these are super easy for ai, and movement is also limited , now let me count you few small stuff among the billion stuff ai has to cauclualte in dota2, you wont understand anything if you have not played the game, and even if you did you still may not understand ,

dota2 is one of the most complex games around, also far more complex than game it is being compared to named leauge of legends due to having less straight up calculated mechanics and being more based on human decisions map awarness , requiring more coordinated teamwork having a bigger varierty of skills, it is also choosen for a reason. eu4 is not a complex game either, there are bunch of non deep simple min maxing tabs that are designed to be simple easy to understand with not many trade offs, and is filled with insane rng, hoi4 on other hand have no significant rng

I think, your have a misconception of complexity and depth. Complexity represents the amount of different possibilities. Depth represents the amount of meaningful choices. In general, you want a lot of depth and not meaningless complexity. I would agree, that a game like dota2 has a lot depth in combination with high requirement in individual skill making it a good Esport title, butHoi4 has way more complexity (which is not necessary a good thing). Sure you can calculate the result of a single battle quiet easy, but you might have to perform thousand of battles to win a war, involving different terrain, different templates, different amount of fighting units, changing level of technologies and equipment, hundreds of units, multiple fronts, multiple nations, which are or could be involved etc. Furthermore, it might be worth to attack, even though you lose a single battle. You might want not to attack, even if you would win the battle. The number of possible "states" and possible actions is orders of magnitudes higher compared to a game like dota2.

In general we humans are very good in dealing with complexity, because we are strong in pattern recognition allowing us to differentiate between important and meaningless things, but it is extremely hard for a computer program. In contrast to that, a (machine learning) AI has way less problems to deal with depth, since it is way easier to test a "few" meaningful decisions.

 

[mursolini]

Not if you have the "together for manpower" DLC. Australia can trivially field and ultimately annex their way into having more active manpower (with lots in reserve) than their entire recruitable core population.

If manpower is set in stone this kind of picture is fishy:



Czechoslovakia has a core population of 15 million. Which means I have > 50% of core manpower sitting in a recruitable pool before even factoring the currently fielded armies.

With together for manpower, ANYBODY can get the bodies of USA/India/China/USSR in SP. Failing one of those, Brazil and Dutch East Indies each have a ton of pop to use too (which is why Australia can get so much).

Quite a few lesser countries can get you ~500,000 manpower at extensive conscription too. Argentina, Iran, Turkey, and Ethiopia are examples of countries like this (in fact several releasable African nations are, but they're usually under majors or Belgium, in which case you'd just hit NED if doing SP manpower plays).

I didn't play TFW countries, so I would appreciate more detailed description.

 

[TheMeInTeam]

I didn't play TFW countries, so I would appreciate more detailed description.

You can annex puppets with together for manpower DLC no matter what country you are playing. Recruit as much as possible from templates using their manpower, then annex them and those troops are yours...as is all the manpower comprising them.

 

[agonistes]

By efficiency I met all those areas you can be more efficient in. This would include manpower.

Specifically regarding the 40x, that's a lot of manpower and equipment, and it is less versatile.

I've been tinkering with 26x. It is just short of the 80 cap, and 120, but will win in most situations. If I have a defensive line with a point where the numbers may work against me, I fort it up. But generally you are pushing through, and the enemy won't have time to bring in max numbers. And it lets me cover more ground.

 

[Foxman8472]

a game like dota2 has a lot depth in combination with high requirement in individual skill making it a good Esport title, but Hoi4 has way more complexity (which is not necessary a good thing).

HoI4 is without a doubt the easiest to grasp of the Paradox games. You're right on the money, it doesn't have much depth, because nobody cares how many possible division template combinations you can get if you'll be railroading towards the same minmaxed ones. All the inferior branches of certain Land Doctrines are just wasted resources, because people won't play them unless they want to RP for some reason.

However, I must disagree, HoI4 is not more complex than DotA. DotA takes about 2 years to learn the heroes and their item combinations, specific effects and roles, to become decent enough to play a multiplayer game without your teammates committing Sudoku over your ineptitude. HoI4 can be learned in less, way less; People who disparage DotA clearly haven't played it (enough) and consider it to be one of those loser Russian games for kids on Krokodil. What HoI4 has is an abnormally large number of suboptimal strategies. It has junk tech as much as Diablo 2 has junk drops.

 

[TheMeInTeam]

^ You can get very good at either DOTA 2 or HOI 4 inside a month or two. Not professional levels of good in DOTA 2, but very good. Most people will not approach these games in a fashion required to attain that ability so quickly, but it's possible to for example study hero abilities like you'd study for an exam and practice 4 hours/day. Is that fun? Not for most people, but some people actually do stuff like this and presumably at least some percentage of those players actually do enjoy it.

 

[Ninaran]

dota2 is one of the most complex games around

Wait, are you being serious?

 

[permanently_afk]

As someone who has - some -experience in the fields of data modeling and machine learning, I feel that maybe some things should be mentioned and explained. First off, data modeling is an integral part of any "AI" effort. Interfaces must be provided, and data either enriched or stripped down to the relevant (that's gonna be important) metadata for the system. Most complex projects already face difficulties at this stage, since the "relevant" most often amounts to "educated guess" - there is a reason why data engineers are well paid. This adresses one of the points of @Shinkuro Yukinari :

Some notes. While the movement in HoI4 is much simpler than say Starcraft due to it being from tile to tile as opposed to free movement in any direction, it can also be more complicated to implement for the OpenAI due to the amount of tile connections present that need to be accounted for.

This is a false positive, because it can be modeled far easier. On the top of my head (ok, in the time it took me to make some tea):

• Start-Point
• End-Point
  
• Travel type (land/air)
  
• Travel time
• Route-weight (e.g. "previously encountered hostiles"), assigned by the "AI"
• If travel-type is "land", also the elevation

From this you can pre-compute "travel maps" for each game map and then fill them with learning data from games. And yes, this accounts for areas only accessible by transport. These can then by refined into "node maps" by the algorithm, representing spots of interest and so on. This is - relatively - easy, and more importantly precomputable. In HOI4 you'd have in addition (at least, again, tea-time):

• Type of each intervening "tile"
• "Most probable" travel-time (due to mud etc)
  
• Is (faster) sea-transfer available?
  
• Air-superiority modifies travel time
  
• Attrition vs. earlier arrival

But, worst of all, the HOI4 map is not precomputable, since the point-to-point can and will change during a game. And this is just the basic map for traveling. Not for combat.

But I am getting carried away. When talking about the topic of "AI" it is always important to remember that the computer will beat a human in the following fields - and for some games this is all it needs:

• Reaction time (It does not have that ~0.5 of IPO lag time of the human brain)
• Reaction reliability (It does not suffer from lapses of concentration or "bad days" - nor will it ever misclick)
• Data storage (All of its memory is rote and perfect. Meaning, yes, it can recall all action combinations. And it will never mis-remember either. This is why being able to precompute stuff is so important)
• Data processing (The algorithm can simply spend 900 game-years to precompute (there is that word again) its solutions. For illustration, for something I recently saw 900 000 000 games were considered a bit on the low end for training data.)

All hail our robot overlords, eh? But fear not, humans are vastly (and I mean vastly) superior in the following fields:

• Uncertainty deduction (any sentence containing the word "probably" - and yes, this is a lot. And keeps my gf employed ;-) )
• (Complex) pattern recognition (movement in the trees -> Tiger or kitten?, understanding system interactions - see Rube Goldberg)
  
• Out-of-context deduction/adaptation (For a system, a orange-painted ball can be an orange and not a ball)
• Self-improvement (Everything the machine does, must have been instructed to it. And yes, the instruction can be "learn this game". And no, this is not "Get better at the game")

Of course, all of these points can be circumvented or worked at (and indeed, there are efforts being made), but they tend to be - computationally or money-wise (those FPGAs are not cheap) - expensive. E.g. solving those orange-balls problem would require an additional sorter (because you need additional data points for modeling) and a system to discern fruits from balls.

If you are a good dev and making a game in a genre that allows for it (say, FPS, MOBA) you'll read Sun Tzu and use the first points to cover for the latter. Problem is, HOI is by its nature more of the latter than the former and so, the "AI" sucks and is exploitable. But then, on to the gentleman wo incited me to write this post. But first a last diversion:

I agree that it will be time intensive to implement some mechanics, and that it will take a decent amount of time to pull off a good AI.
I disagree that it is highly unlikely in the current state of things and with the achievements already made in the area. Dota and Starcraft aren't Chess.

While I agree with you in principle, I fear that any achievements in strategy games are far off since the genre is rather less flashy and unpopular in esports. And yes, I file SC under RTT.

i heavily doubt you understand the games you are talking about , or you played them at a level that can be considered serious,

Great start, with a True Scotsman fallacy. For my part, I *know* that several of the OpenAI et. al. developers don't play "their" games with anything resembling professional skill. But then again, understanding and modeling the mechanics is divorced from clicking fast, if you'll forgive me that aside.

For hoi4 mechanics are simple, calculations are extremely simple, any person that knows what breakthrough defense attack and multipliers , reinforce rate means ,will easily predict if he will win combat or not places to defend are always in front of your eyes, you can simply know how much you will hit ,you can easily know all modifiers from cas air support-planning-force attack - general skill- terrain and guess results of the battle super easily, calculating if i will lose more equipment? lol you joking ? anyone with basic maths will have no issue understanding if he will win combat or not,, Ai will never have issue calculating its damage compared to enemy defence/breakthrough and will have no issue calculating modifiers, in fact these are super easy for ai, and movement is also limited ,

You list quite a few and then claim its "simple"? That statement isn't even wrong. Furthermore, the interesting bit is not the amount of data points, it is whether or not the are deterministic and/or atomic. In HOI4, they are not.

now let me count you few small stuff among the billion stuff ai has to cauclualte in dota2, you wont understand anything if you have not played the game, and even if you did you still may not understand ,

In dota2 you try to deny your own creeps, while killing enemy creeps and every decision you make while last hitting is a trade , in a sitaution where our ai is gonna deny , our ai sees that his allied creeps have 60 and 63 health , and enemy hero hits 50 and ai hits 50 too, which creep our ai will deny, how does he decide? let me count you some of th factors while our ai just have to last it
1-) calculates distances of both heroes to the creeps
2-) calculates projectile speed
3-) calculates attack animation
4-) calculates turning speed
5-) callculates backswing from attack animation in case someone goes for double hit on a creep ,
6-) calculates which creep enemy human is gonna hit to depending on factors like how humans choose their target ( eg human deciding to hit unit that is closer to unit it hit before so he moves his mouse for a shorter time ) ,
7-) calculates how much creeps are hitting to eachother
moves during backswing of attack animation to get closer to the creep in order to reduce the timeframe his attack projectile will fly
9-) at same time calculates bunch if x-y-z abilities from enemy hero are threat and damage-harrassment he takes will be worth it if he geos for deny
10-) calculates where enemy roaming heroes can be since it is a laning stage, our ai wont sucide to last hit a creep, will go back if heroe are mising

guess what humans try to do all i wrote above too! and they are just trying to last hit a creep, hoi4 players choose focus queue buldings and start smoking weed, ai would do same too

All of these are deterministic, and therefore precaclulable. Also at least point #6 is something the "AI" does not have to contend with. All of these are mainly difficult for a human, not for an "AI" do do, since they play to its strengths of reaction time, reaction reliability and data processing. This entire action you describe is both deterministic (there is no differing in any of the values) and atomic. Therefore, it is many things. But certainly not complex.

dota2 is one of the most complex games around, also far more complex than game it is being compared to named leauge of legends due to having less straight up calculated mechanics and being more based on human decisions map awarness , requiring more coordinated teamwork having a bigger varierty of skills, it is also choosen for a reason. eu4 is not a complex game either, there are bunch of non deep simple min maxing tabs that are designed to be simple easy to understand with not many trade offs, and is filled with insane rng, hoi4 on other hand have no significant rng

Now, if you'll forgive me: Yu srs?

 

[Achul]

I'm a bit late to the story and would generally like an update.
Tired of playing Darkest hour and its variations but I just simply can't do anything but laugh at a game where this is possible on normal diffiuclty:

Luxembourg conquering Germany and Spain and France and fucking China.
It's not just the bad AI combat. It's the complete lack of a reactive AI in terms of diplomatic actions (he seems to leave the Axis in the middle of war, justify a war goal on Italy and then re-join because he forgot something).
Not to mention that Luxembourg never should be able to get all that manpower from occupied territories of subjects that probably wouldn't want to volounteer for Luxembourgs mad government and die in its random invasions.

I've complained about the direction of Paradox Studios in releasing unfinished games and somehow the fanbase accepts that one should wait 3 + years for all the expected content to come out. Well now it's about 3 years later and I thank those who have done so, so that I didn't need to.


I'd like to join you if there's any improvement? I know there are some mods, how good are they?
And how's multiplayer, is it active stil?

Thank you kindly.


PS: I gotta ask, in that video it looks like annexing a country gives you its stored manpower?
He goes from 8k manpower to 25k when taking the Netherlands. Why and how?

 

[Secret Master]

Thread necromancy, and especially thread necromancy that includes pointless profanity about China, is generally not acceptable.

We should know better. Thread closed.