Hearts of Iron 4 announced

It’s a math optimization thing that I didn’t bother to memorize, so I googled and copy&paste

Someone did a thing with 5inf a while back, it turns out org has diminishing returns the larger the division gets, ie 5inf has about 75% org of 10 and similar majorities in all other stats. Also once a big ass division runs dry it’s done; smaller ones prolong the fight which gives CAS carte blanche to butcher the other side.

The main thing is the ORG is currently average ORG. So if you take a 20-width division of INF, you can make it stay in the fight for twice as long by splitting it into two 10-width divs. This used to be counteracted by the concentrated firepower advantage of larger divisions, but they are (rightly) doing away with that.

Support companies complicate the picture, it’s true. It makes less sense to have say a supply company for a small division. On the other hand, support artillery and rocket artillery are really good bang-for-the-buck additions to any division, and two 10-width divs can bring twice as many as one 10-width div. Sure their per-division ORG would be a tad lower, but they start with twice the ORG, so they can easily afford more dilution.

You’re right. They have to fix it. It shouldn’t be better to have a smaller division be better than larger, although they went overboard with the bigger ones being so much better.

An AAR of a very changed game as a dev diary.

That certainly got me in the mood for HOI4. Looking forward to this expansion. Or rather, I’m looking forward to the first major hotfix after this expansion. :)

I thought the “sub” DD on mechanical changes was also interesting:

It seems like they’ve modified things from last week to once again give larger divisions an effective firepower concentration advantage, though it works a bit differently and depends on coordination and initiative. It looks to me that with high values for these, larger divisions could be even dominant, since two adjacent large divs can focus fire on the most-damaged div, overcome defences efficiently and drive it out of combat quickly.

He has a lot of interesting followup answers to questions raised. Will be interesting to see how this all shakes out at release.

On some level I think it’s interesting, a new angle to optimise! However, I’m not sure how much it does for the game. This expansion is set to introduce a lot more mechanics that are likely to push the AI further behind.

Beyond that, it just makes no sense that small divisions are willing to suffer more casualties in battle. Or really that large divisions have this large-scale targeting advantage, or that divisions have a certain number of “shots” they can deflect above which they take more casualties.

IIRC, the Darkest Hour version of HoI2 actually reversed that, and had excess firepower directed at a division actually reduced the per-shot damage. They argued that historically, beyond a certain threshold, throwing a larger attacking force at a defender didn’t help, and didn’t reduce attacking force casualties the way it does in HoI4.

Overall, I hope they completely overhaul the core combat model for HoI5.

All true, but I think calculating the hit probability of each rifle, mortar, machine gun and tank round is probably excessive so combat needs to be abstracted in some fashion.

I have to say I’m unsure,what happens in real life, if a 20,000 man division attacks or defends against two 10,000 man division are the combat result different than attacking another 20,000 man division. I really don’t know.

I’m looking forward to the expansion, but HOI IV is not EU IV, with decent strategic AI there is no way Poland should be able to take out Germany or Russia, much less both. I understand that is a fun challenge for some advanced players, but it really turns me off.

I remember War in the East actually did this: a battle was divisions firing all the weapons in their OOB at each other according to some fixed schedule with a bunch of modifiers on top. I don’t think it led to more accurate results. In some sense going super detailed on some aspects just makes it harder to find a decent abstraction for all of the stuff you leave out.

So I agree it needs to be abstracted. I just don’t like the abstraction they are using, where divisions have a number of shots they fire at other divisions, and each division has a number of shots it can partially deflect.

If divisions are independent combat elements, I think the optimal size of divisions should depend on the size and quality of your officer corps relative to your army, and maybe on communications tech. If you have plentiful officers and good doctrine and communications, then having smaller divisions should allow a degree of flexibility. I’m a poor student of this sort of thing, but I suspect there’s no shortage of military historians who would be happy to consult with them for free.

In other PDX news,

“Recap” of a change that wasn’t particularly covered, but, sure.

I like the officer changes, and AI improvements are always welcome.

No Step back announcement trailer.

Release date Nov 23

Looking forward to it.

Same! Bring on the DLC and the necessary hotfix or two. :)

I’m not exactly an art guy, but this was a really cool look at the process of making art for a computer game.

Now this is a Thanksgiving feast of patch notes here.


It’s way too much to copy and paste but I’ll grab the major features for the expansion and the free update.

Fellow revolutionaries might need this playlist

I have been following the development of the new expansion since I like playing as the USSR. However, after watching a few videos of the new systems they are implementing, I am not as enthusiastic. For one thing, the tank designer sounds like a cool idea in theory, but after watching the video it’s more than just selecting the best gun and chassis… it’s also about allocating individual armor points and increasing reliability through the addition of little things like “wet ammo storage.” This level of detail is too granular for me and will make designing tanks a chore. However, there is an “auto upgrade” button I am thankful for, but deciding when to change my production lines will be too much as well… should I switch over production because my tank design just got a 2-point armor increase? Should I wait until a 5-point armor increase? Ugh. Another overwhelming prospect is learning the new logistic model. Learning the current supply system was about as much as I wanted to learn about logistics. However, now I have to worry about building not only individual train tracks, but trains as well? Also, there is train artillery? How do I decide to build an artillery train instead of a light tank, or instead of mechanized vehicle? It seems to be a bit too much for me now.

They may have adjusted things, but I suspect the big hits to production efficiency only come when you change the base chassis. Armour upgrade should be the equivalent of an experience upgrade in the previous, which was a pretty minor efficiency hit.

I’m not really sold on that sort of thing. I’ve come to the conclusion over the years that “ship builder” type stuff just doesn’t really have a place in most high level strategy games. I’ve only found very few games where it added something to the game. Sword of the Stars is one game where I enjoyed it because tactical battles actually played out and each race had access to different levels of the various techs. For most 4X games, though, it usually just boils down to more micro than I want to deal with or pointless since there’s one or two meta configurations that everyone runs.

I’ll withhold judgement on the tank designer until I try it, but it’s not one of the features I have been looking forward to. New combat calculations and railways on the other hand I’m all kinds of looking forward to trying!

After this expansion, the only thing I’m really wanting is a rework of the Italian focus tree which I’ve been waiting for since the first DLC! It’s my favorite country to play so it’s been a long wait. Once I have that I’m all for sticking a fork in HOI4 and looking forward to what they come up with in HOI5.