Summary of my Impressions in Feedback lately

[xsmilingbanditx]

Hey guys (and...maybe some gals who knows ),

I have been following the appearance and discussions about HoI4 in it's current iteration, because we intend to do a little "just for fun" CoOp game soon....y'all know, with beer and talk and stuff.
I already did two tests with diverse results (One where Austria denied Anschluss and I ended up in a war with Hungary, Austria and Italy with no chance to cross the Alps and the second one where everything went as predicted).

Now, I'm not gonna say everything is fine but when looking at most of the Complaints (except the few real issues i so agree upon) the consens looks like this:

- Will China ever win?
- Japan is always driven off the Mainland!
- Germany rolls over Europe!
- Why can't I conquer France as Germany?
- AI is too dumb, it cannot do naval invasions!
- Germany always naval invades Norway and rolls over scandinavia...
- USSR too strong, Germany gets bogged down
- Germany too strong, always annexes USSR in 1943
- (former AI is dumb, it suicides into forts!
- (now AI is so dumb! It does not attack my fortified defensive positions where I do have superior numbers AND Air Superiority...
- list goes on...

Now, while I do not want to make fun of people who do have valid Complaints (in their opinion at least) - and though the list sounds funny - everybody looking outside of his own information/opinion Bubble gets to the following conclusions: Things are alright...yeah there are issues but the "Complaints" are not cohesive mostly and so diverse that there seems to be no pattern.

People often seem to fail to understand what an "AI" is in its current state and possibilities (and vastly underestimate the complexity and amount of calculations necessary for a "system" like HoI).
While there are strange behaviors and edge-cases (like declaring war because the focus/Trigger tells it to while not attacking because too low Probability of success) the "AI" does have no understanding of what it does to whom and why...it does not know what Germany is or USSR, what a Tank is...or "War" for that matter. It does not even know it WILL declare war three years from now and why it should build Fighters...yes guys, that IS the current state of AI in nearly all Applications you may find. Some are just disguised way better and others are just "Big Data Analysis Algorithms".

I sometimes think "Man, imagine ALL HoI4 AIs currently playing would contribute to a central Server as a persistent knowledge Base they could Call upon, and talk to each other like 'Yo fella AI, my player here just rolls into China with Tanks, what worked best for you last Time and what did not?'"

Yes, that would be possible, no that still would not be an AI
And it would be hell for new Players...and everybody would stop playing after three games but still....cool nonetheless

Edit: Sorry for typos, I am on a mobile with German language Settings

 

[Deinhardt]

This man. He speaks sense.

It's refreshing really with all the negativity going around this forum lately.

Thanks!

 

[xsmilingbanditx]

This man. He speaks sense.

It's refreshing really with all the negativity going around this forum lately.

Thanks!

Hehe, thanks. Positivity and a grain of irony usually aren't received very well usually it seems

 

[Zauberelefant]

I think that with 1.5.1 the game is pretty much alright.
If Future patches follow that line, I am going to be happy.

 

[Serensius]

Great post. I think a lot of times on this forum, the words "EVERY TIME" or "ALWAYS" tend to mean "In the three games I've played..."

I think another aspect of this is that the more content the devs add, the more complex the system becomes. With almost a hundred countries that can take vastly different paths based in part on random number generation, the harder it becomes to produce a plausible outcome (i.e. a world that makes sense) - especially with wild curveballs like communist Japan.

So I think the devs deserve huge kudos for making everything and I hope they continue to focus on developing dynamic systems that are designed to provide outcomes that are both plausible and predictable, but not inevitable.