This is the eighth in a series of reviews of Paradox’s empire management games. See the earlier reviews here:
Europa Universalis II: A Tale of Two Europas
Hearts of Iron: Europa of Iron
Victoria: Nineteenth Century Essay
Crusader Kings: A Dynastic Adventure
Hearts of Iron II: Return Engagement
Europa Universalis III: A Whole New World
Europa Universalis: Rome: Make a Desert and Call it a Game

Given the number of other projects that Paradox could have done, the four-year gap between Hearts of Iron II (2005) and III (2009) is fairly short (especially when considering that the final expansion to II didn’t come out until 2007). However, in that time Paradox had developed the Clausewitz engine for their empire management games, and with the HoI series being Paradox’s most successful one, it is no surprise that they hurried to get a new WWII-era (1936-48) empire management game out with the new engine.

Unfortunately, ‘hurried’ is all too apt a word. Paradox’s games always had a reputation for being buggy upon release, but would then be patched into a solid game. The release of HoI III was particularly bad in this regard, and after the complaints that caused, Paradox has made an (increasingly successful) effort to release their games in good shape.

After the initial release of HoI III for PC and Mac in late 2009, the first expansion, Semper Fi, was released in June 2010, and a second, For the Motherland, was released in June 2011. These were packaged together as the HoI III Collection, which is all I’ve played, and what this review will be about.


Originally, all the resources in the HoI economy were tied together, as they all fed into Industrial Capacity (IC), which was used for everything, including research. In HoI II, research was separated from the IC economy, but still required money to fund, which came from consumer goods (generated by IC), and the number of possible simultaneous projects was controlled by the amount of IC the country had.

HoI III introduces leadership as a second economy that does not generally interact with the resource/IC economy. Leadership is like IC in that it is generated by provinces the nation controls, but is a constant like resources, instead of being upgradeable like IC. Leadership is then split to research, espionage, diplomacy and officers.

Tripartite Threat

HoI II used policy sliders like in EU II to determine government type and which faction a country was close to. HoI III returns to the tripartite political model of the original game of the series, where every country is placed in a triangular diplomatic space with the Allies, Axis and Comintern represented by the points.

As usual, diplomacy has a number of options, including influencing another nation towards your point of view. This is only available if you are already in a faction, or else you are subject to influence from those countries that are (you can align yourself towards one particular point of view). All of these actions cost diplomatic points, generated from leadership. Most actions (negotiating a trade deal, instigating a coup) have a one-time cost, but influence has an ongoing cost for as long as it is kept up.

Most countries also start with a fairly high neutrality rating. This is a new mechanic to represent a country’s resistance to going to war. You can only go to war with a country with a threat rating higher than your neutrality rating. Threat is generated by military build ups, going to war with other countries and the like, and is also reduced by distance. So an aggressive Germany will generate a lot a threat against France, but not so much against America on the other side of the Atlantic.

Practical Engineering

In the first HoI game, research was split into theoretical advances that then allowed practical advance follow-ups. HoI II simplified the model down with nearly every advance having a direct impact, but each was made up a few sub-advances that had different demands, but those did not interact with each other at all.

HoI III introduces the idea of engineering knowledge (or theory), and practical knowledge. Every time you complete research in a field, you get a point in the theory for that field, which will slowly decay over time. Building units, or engaging in combat generates practical knowledge in the relevant field. Knowledge of both types affects the difficulty of research. So, the more you concentrate on one field, the better your theory is, and the easier it is to get more advances. The more experience you get with, say, air combat, the better you know what improvements your aircraft need, and it becomes easier to advance in that field.

In general, each unit type has a number of different advances (often about four) that each improve different unit statistics, and together add up to the next ‘level’ of that unit (Infantry I to Infantry II, etc.). As in HoI II, land and air units can be upgraded in the field when new equipment is available by allocating IC to the job. However, land units can also be upgraded by type, with, say, motorized infantry being converted into mechanized infantry, which places them into the general production queue (removing them from the map), but they will return to their previous unit if it still exists at that time.

The old problem of inappropriate detail returns, with many ship and tank types being given real-world names appropriate to the country. As these are often appended with asterisks to denote one or more differences from the ‘actual’ version, this is cosmetics without meaning. Worse, the ahistorical research sequence of light armor allowing medium armor, which then allows heavy armor returns from the first game.


HoI II: Doomsday introduced espionage to the series, and it is still here for this game with spies, counterespionage and the like. Like last time, I’ve generally found it too much effort for the payoff.

However, For the Motherland did introduce a new resistance… mini-game. Countries can try to put together resistance cells in a foreign country. This is the most likely in the occupied territory of the country putting together the resistance. A resistance network is generally invisible to the targeted country, though Police units do have an effect on them. Once it is in place, it can be turned into a normal militia unit, which seizes control of the province it is in.

A long-term problem with the HoI series is that partisans never really worked, despite attempts to include their effects. This much more direct approach actually solves the problem fairly well. Supplies are important in this game, and often get moved up to the front through a handful of province routes, so a few Soviet partisan units can have a dramatic effect. One of the goals of this was to give a human government-in-exile player something (potentially) useful to do for his allies in a multiplayer game, though it would still be a somewhat lackluster experience.


Production works pretty much the same as with HoI II, with the player having to split available industrial capacity between consumer goods, supplies and new production as well as dedicating resources to reinforcements and upgrading existing units (or he can turn it over to the AI). The interface for those settings and the main queue where you prioritize new production has not changed much.

However, it too is affected by the ‘practical knowledge’ rating that affects research. New units have a base time to complete, but the better the related practical knowledge for that item is, the shorter the actual time becomes. Since this will apply to all further units, it encourages setting orders to produce multiples of the same thing in a chain, with each one starting when the previous one finishes, kind of like a… production line.

One small problem with the system is that once you commit an order to the queue, you can’t edit it, say to change the total number to produce. If you change your mind immediately, you can cancel it and start over, but otherwise you’re stuck keeping an eye on it.

Chain of Command

The basic land unit of construction and maneuver in the HoI series has always been the division, with brigade attachments to provide specialized abilities. HoI II expanded on this idea with a more extensive brigade system, that was also used for carriers and air units. In HoI III, the basic unit of maneuver is still the division, but the basic unit of construction is now the brigade.

Divisions are now made up of three or four brigades (or five with the right research), and can be constructed as a unit, or assembled out of component brigades (is possible to have individual brigades running around, but the system doesn’t really support that as a regular thing). Each country has a number of pre-done construction ‘templates’, which can be modified and saved, so a player can easily tailor his production to the structure he desires.

The truly new part is that higher levels of command are now directly represented in the game. Divisions can be grouped together into corps. This does not force the units to stay together (and, sadly, there is no easy way to select every unit in a particular corps at the same time), but does generate a headquarters unit which is an extra brigade. As with previous games, historical leaders are given to each country, and their abilities directly impact the performance of the division they command. A leader assigned to a corps increases the odds of extra divisions reinforcing and taking part in a combat. Corps can be similarly grouped into armies, which add to the organization of child units (which is the unit’s combat effectiveness). Armies can be grouped into army groups, whose leaders decrease supply consumption. And finally, army groups can be grouped into theaters, whose leaders can reduce the combat penalty for having too many divisions in too small a space.

All of these have their own headquarters units, which can take part in combat, but are not especially useful as such. Instead, the main purpose is that each headquarters has a range in which their bonuses apply. Corps have to stay relatively close to their divisions (say, 4-5 provinces), while a theater command in the middle of the Pacific can cover most of that ocean.

But more important than any of the above abilities is the fact that these can be turned over to AI control—at any level. You can let the AI control a particular division. Or corps. Or an entire theater. The AI is, as ever, not the best player, but neither is it truly incompetent. The point is to be able put entire sections of the game under AI control, so that the player can concentrate on those parts that are truly important. Invading Poland as Germany in 1939? Set the western theater to AI control, so it can handle anything Britain and France might do, and then concentrate on Poland.

This is aided by giving the AI directions as well. Once a unit is set to AI control, you can tell it to act offensively or defensively, and can flag particular places as goals. These could be places to defend, and places to attack. Once the UI is set, selecting the appropriate HQ will display a dashed line ‘front’ on the map which gives some idea of what the AI thinks of the situation.

All of this only partially applies to naval and air units. All the headquarters are land units, so while naval and air units can be grouped into smaller units (‘divisions’), they then must be integrated into the land force structure (possibly just at the theater level).

How Wide is my Front?

Combat is largely the same as it was in previous titles, including the ‘movement is combat’ model of HoI II. There are some tweaks and additions, the most important one being that battles are marked on the map with little bubbles that give the current winning/losing percentage. Battles that are going well are in green, ones that are going poorly in red, and the undecided middle ground yellow. Sadly, a big attack by weak units against a well prepared defender will often end in a defensive victory, but the combat rating (and bubble color) are based off the current gross numbers, not who is actually taking damage.

All units have a combat width, and generally each combat has a maximum width that units can fight in (with attacks from multiple provinces increasing the maximum width). Units in excess of the maximum width do not take part until other units are forced out of line (usually from loss of organization) to make room.

Along with the manpower pool needed to reinforce units and create new ones, there is also an officer pool, generated from leadership. When a unit is under heavy strain in combat, it can shatter instead of retreating, which removes it from the map, and forces it to be reformed, either with its parent unit, or somewhere in the home territory of the country. Each unit needs a number of officers, and the ratio of available officers to needed officers adjusts the odds of this happening.


Troy Goodfellow pronounced HoI III a ‘must have’ strategy game on Three Moves Ahead because it solved the ‘virtual viceroy’ problem. I’m less positive than that, as the AI runs into the same problems that SSG’s old Panzer Battles and Great Battles of the American Civil War series ran into: you can tell the AI where you want it to go, but you can’t tell it how you want it to get there.

That said, the hierarchical command system, with the ability to turn control over to the computer at almost any level is, largely, the key to solving the bigger problem of needing to have too many units to control in a real-time game. What is needed now is better communication.

I was surprised to see the armor research model go back to the ahistorical linear model of the first game, but every other place where HoI III goes back to the first game (and there are several), it is for the best. II showed a lot of lessons had been learned from the first game, and this one generally takes the best of the previous two.

The map gets really fine-grained in this title, and that is part of an over-detailed aesthetic that is partially alleviated by all the AI controls. For me, the HoI series remains Paradox’s least engaging effort, largely due to the expectations I bring to a WWII title. HoI III is the best of the series so far, and worth looking into for any grand strategy fan.