Behind the Scenes of Hearts of Iron 4
While many studios in the industry these days tend to shed artists once a project is done, the ongoing projects PDX works on allows the team to not only focus on mastering the genre they create so well, but to keep the same team in place.
“A lot of bigger studios need a lot of people to make a game,” 3D Artist Joakim Larsen explains. “But when the game is done they’re through with people for a half a year or a year. We don’t; we need the resources all the time so we don’t kick people out when a game is done like a lot of big studios do. We just keep the resources and do the next project.”
“We’ve expanded a lot in the last few years,” 2D Artist Mats Virtanen says. “So there’s a lot of new people, but the majority of the core Paradox team from the early days are still here.”
Although longevity at PDX may be different than a lot of other studios, when it comes to deadlines the PDX team runs into the same thing as just about every other studio. At PDX it truly is a team effort to help everyone push out the final product.
“There’s a lot of things we want to do,” Larsen explains. “We want to make the best thing possible, so stealing people from other parts of the company is crucial to do that.”
Larsen continues, “When a project is nearing completion, usually a lot of people come to help and we steal resources from other projects.”
With multiple games being developed at any given time, the structure for the teams at PDX can’t be the same as it is in many other studios. “We have dedicated core teams for each project,” Virtanen adds. “But as we get closer to the release dates and deadlines, we ramp up with people from other projects that jump in and help for a week or a few days either with bug hunting or with just helping implement new features and such.”
“Right now, Hearts of Iron 4 is the resource deathball of the company,” Larsen laughs.
The People Behind Hearts of Iron 4
Most studios have a similar all-hands-on-deck philosophy of grabbing resources when it comes close to deadlines, but the practice of jumping between games mid-production to meet them is somewhat unique to PDX.
“We artists are floating assets,” Larsen explains. “During the production time of Hearts of Iron, I also worked on other games. Sometimes there’s nothing for me to do on Hearts of Iron, so I go to work on another project for a while before coming back.”
“I’m a bit more tied to Hearts of Iron,” Virtanen adds. “But I’m assigned to Europa Universalis as well, and of course when needed I can jump in to help on other projects.”
“That’s actually quite nice to not be locked down to a project for ever and ever,” Larsen comments. “To be able to do different things.”
One of the reasons why artists float from game to game is because there aren’t nearly as many artists as there are artists. Perhaps this is because of the style of game PDX has mastered, but the end result is each artist at PDX needs to be alright with jumping between projects and hit the ground running when they do.
“We’re not a graphic assets heavy company,” Virtanen says. “So it’s quite interesting when you see the ratio between programmers and artists.”
Although they are a growing company, as of this writing Virtanen estimates there are around 40 developers to six in-house artists, not counting the in-house QA team of 17.
Virtanen continues, “As part of the core team on a project, sometimes a lead will come to me and say, ‘Okay, Mats, you have this programmer as your designated programmer. You can go to him to get your things done.'”
“And I would say it’s quite a unique position sometimes where I get to boss the programmers around,” Virtanen laughs.
“It’s a very unique and cool way to work,” Virtanen concludes. “Since we are a very close-knit group, you get to have a lot of say in the actual look.”
Because of the way the PDX team chips in to help each other out when a game is nearing completion, just about everyone at PDX has their hands in the creation of Hearts of Iron 4.
Of course, not everyone at PDX was working on Hearts of Iron 4 the whole time.
“It started out with 13 people,” Virtanen recalls. “Not too big. Plus we had three programmers and one scripter. That’s the core team on developing the project. We’ve had at least seven to eight people working on [Hearts of Iron 4] since it started.”
Throughout the production, though, sometimes people hopped on and off the project as other ones demanded attention.
“When we ended up doing the credits, basically everyone in Paradox Development Studios in one way or another contributed,” Virtanen says.
Making Hearts of Iron More Approachable
If you’ve played Hearts of Iron 3, you’ll know that one of the biggest challenges to the game isn’t really in its strategy. It’s interacting with its not so aesthetically-pleasing, but feature-heavy interface. Making the interface more approachable would be a main focus for Hearts of Iron 4.
“An issue that all of our games have more or less is usability,” Anna Jenelius, Senior QA Manager at PDX, explains. “They are so complex. We want them to be easier to access, but not lose the depth of it.”
“Hearts of Iron 3 had a very clunky interface,” Virtanen admits. “People felt that it was very, very intimidating and hard to learn. So we started to think about how we can make this game usable while keeping the depth.”
This is easier said than done.
“It’s not purely QA, but we still give input on usability and it’s something we work on closely with our UX designer,” Jenelius states. “There’s so much information, how do we present it in a way that people understand?”
There’s an inherent tension between functionality and aesthetics that many UX and UI designers often face. The question becomes how can you make something both interesting and clearly understandable. Adding too much flourish and ornateness can create confusion, while an interface that’s too bland can create boredom.
“As an artist, I want to make everything as pretty as I can,” Virtanen adds. “But making something too pretty might draw the eyes away from the really important stuff.”
The solution for creating a user-friendly interface while not losing the depth was found through iteration.
“We have a UX designer who’s putting a lot of thought into this,” Virtanen explains. “He makes a lot of mock-ups to figure out things like, how many clicks do we need to get to this point, which tabs are the most important and how do we lead people into a process where they do the right clicking in an efficient way?”
“Hearts of Iron 3 was huge,” Virtanen adds. “I see all these buttons and all this text and all these lists.”
“I mean, I think it’s intimidating. I never actually learned to play it, in a good way,” Virtanen laughs. “One thing we did was start moving away from these walls of text, replacing them with icons because they’re easier to understand. Pictures are prettier to look at.”
Of course, there’s a balance that needs to be determined.
“This being a Paradox game, you’ll never get away from some of those lists,” Virtanen admits. “I mean, if you have 50 divisions you have to be able to click on each and every one of them.”
One of the biggest challenges for Virtanen on Hearts of Iron 4 was in making the counters for the units, which you can see in the screenshot above.
“I can’t remember the exact size on it, but it’s roughly around 35 pixels by 70 pixels,” Virtanen recalls. “In that you have to fit the division type, the strength and the organization bar, which army group they’re in, the number of divisions, fit in if they’re seasoned, if they’re elite groups, if they’re using different variants and so on. That’s a lot of info to cram into those small counters.”
Virtanen continues, “So you have to remember it is a grand strategy game so we need that information to be conveyed in the best way and still not cover everything up.”
Helping to make Hearts of Iron more approachable isn’t done completely in the UI. There’s a new feature in the game that lets users avoid lists completely.
“We have a new feature called the Battle Planner,” Virtanen explains. “We’re iterating on it, and it’s still under development, but with the battle system you group divisions into one army, let’s say, assign a general as a leader and basically you draw actual battle plans across the map so you can execute the orders.”
Virtanen continues, “Hearts of Iron 3 was very micro heavy, where you had to actually move each one of your divisions manually. Of course you can still micro manage your divisions [in Hearts of Iron 4], for example if an enemy is trying to cut you off from supplies you can manually send a few divisions their way to stop them, but the micro way is very intimidating for new players.”
“We want to cater to new players, so we tried to make a system where this really huge and complex operation of making war during WWII is a bit easier,” Virtanen concludes.
Utilizing QA to make a better game
As any game dev knows, there’s a lot of work to go into creating a game. This is compounded when you’re creating a game with the scope of complexity such as Hearts of Iron.
Although Hearts of Iron 3 was initially launched with some bugs, the team at PDX really wanted to combat this for Hearts of Iron 4. To help, they rely on QA throughout the whole development process and not just at the end.
“We try to involve [QA] as quickly as possible,” Virtanen explains. “That’s something we really try to focus on a lot these last few years, because you can’t really release bug-infested games.”
“We don’t do that anymore,” Virtanen laughs. “We really listen to our customers and our fans, we owe so much to them to not release buggy games.”
Tackling the complexity of a grand strategy game isn’t easy, so the team at PDX harnesses the power of their fan-base to deliver the best product available.
Jenelius continues, “We usually have about an hour a day for all of the testers to go through the forums and take in bug reports, going through them and trying to reproduce them and put them in our internal database.”
Although their in-house QA team plays a huge role in giving feedback, the QA process involves more than in-house personnel. To help make sure everything is covered, the team at PDX leverages a robust user community.
“We know we can never cover all the bugs, because we are only so many people,” Jenelius explains. “We use the community, we have beta testers that test for us. And we continue to listen to people after release, for years after. That’s a whole other scale, because in one hour after we release we get more man hours of testing than we could ever get in-house.”
“Having a good QA team and a good QA process is essential to making a game into a good product,” Larsen adds. “It’s a very iterative process. Very seldom something doesn’t change, because a lot of people look at it and give feedback on how to make it smarter and better.”
“Our QA testers can come up with their own solutions and give feedback,” Virtanen comments. “If we’re closer to a deadline it’s tough to iterate, of course, but if I have the time I’ll try it out.”
“Many of our testers have a background in usability from university,” Jenelius explains. “They may not be experts in it, but they have experience with it and they can speak the language. And we also have a user research department in-house that helps.”
“So if we’re wondering whether or not a tutorial does the trick, we can go to them and say, ‘Can you run this through ten people and see what they think?'” Jenelius adds.
Feedback at PDX is a two-way street, and it’s not always QA who gives feedback. They also get feedback from the rest of the team.
“[Prioritizing] is always tricky,” Jenelius says. “Often it’s done with the developers to see how severe they think something is. We found this one, do we need to do it now, in this next patch or later?”
“Of course, if it’s a crash that’s always a high priority,” Jenelius laughs. “Because we have a close relationship with our developers we’re always trying to talk to them to see what they think.”
When you’re working towards a common goal, being open to getting feedback is crucial. One of the key ways you can achieve this is through building relationships with the people you’re working with, a fact that the team at PDX knows well.
“We have a great advantage at Paradox because we have a tight relationship with our developers,” Jenelius says about the QA team’s relationship with their programmers. “We’re not embedded with them, but almost. They’re literally next door, we have lunch with them and hang out with them.”
“It becomes natural,” Jenelius continues. “We report something and they say, ‘Okay we fixed it in this way. What do you think?’ They want our feedback because we are so close to each other, both physically and in our relationship. They genuinely care about what we think.”
Asset creation for Hearts of Iron 4
With the amount of depth and scale that goes into a Hearts of Iron game, that means there’s a large amount of assets that are going to be created. And by a large amount, I mean a crazy, insane amount.
“We have over 1,000 unique items in the game right now,” Virtanen says. That doesn’t even count the number of assets that need to be created for each of those items.
As is to be expected, the team at PDX really integrated their community’s feedback into building the experience for Hearts of Iron 4.
“We saw that players really liked the pace in Europa Universalis,” Virtanen explains. “So we thought about how we can use the guidelines from Europa Universalis. But every 2D asset was made from scratch for Hearts of Iron 4.”
Laresen explains further, “When we start a new project, we cherry pick everything we need from earlier projects so we can start prototyping the new game. As production moves along, Mats replaces all of the buttons, windows and so on to fit the new game. Europa Universalis is quite a different art style than [Hearts of Iron 4], so we can’t really reuse buttons or icons and stuff like that.”
To help crank out the assets, the in-house team at PDX leveraged others in the Paradox family to help churn through the assets and leveraged some gold standard assets to ensure the quality.
“When we started Hearts of Iron 4 we did three initial characters in-house,” Larsen explains. “How we wanted them to look, the different solutions we wanted. That set the baseline for how assets should look for the game. So when people start doing the bulk of the work, like the 60 characters, they can refer back to those initial three.”
On the 3D side, plenty of assets had to be made but fortunately they were able to use existing assets as a base mesh for new ones.
“The assets made for Hearts of Iron 3 were actually quite good,” Larsen comments. “But our engine has developed quite a bit, so now we have physically-based rendering, a really good material system where we can make things look like they’re supposed to. Cloth can look like cloth and metal can look like metal.”
“We have specular, gloss, metalness – we have all that fancy stuff,” Larsen laughs. “So, for example, for some tanks we could pull the Hearts of Iron 3 assets as a starting point and build the high poly model on top of it.”
“Some assets were really good,” Larsen continues. “We could just do some passes on it with NDO to add normal map details. So we could reuse some of the work, but it wasn’t just like we could take Hearts of Iron 3 assets and plug them into Hearts of Iron 4, because it’s a whole different rendering pipeline in the engine and a whole different quality of assets.”
For Hearts of Iron 4, all of the tanks were built as high poly models that were then baked down using NDO and texture maps to capture the high resolution detail to a lower poly model. Although some game engines may choke with the number of assets needed for such detail
“We didn’t really have a huge problem with polygons,” Larsen explains. “Say we have a lot, a lot, a lot of stuff on the screen. We could have a couple million polys on the screen, but that’s never really a problem for our engine as it is right now.”
While many games have level of details (LODs) to add an extra amount of complexity, the PDX team didn’t really need to use LODs.
Still, optimization is something Larsen wanted to ensure was on top of everyone’s mind for the 3D assets.
“If we can make an asset 1,000 triangles, then it shouldn’t be 3,000,” Larsen says. “A tank in Hearts of Iron 4 will range anywhere from 1,200 tris to almost 5,000 tris depending on the type of tank, since a lot of them look quite different. But we tried to bake down to keep the assets as optimized as possible.”
“I’ve worked at other studios before where we had to do a lot of LOD steps and stuff like that. Our game is not that poly heavy, so we don’t have LODs. The only thing we have LODs for are trees. But everything else, characters, tanks, we don’t LOD them. We just pile them up when they’re at a great distance, so that’s a nice option so you don’t have to do a lot of LODing.”
Larsen adds, “We also tried to do smart things like most tanks must share all the spec and gloss maps since they’re all metal. You should always have an optimized base to work from, since there’s going to be people with old computers who also want to play, and you’re going to want to add stuff onto your game later on. So you really want that base to be optimized, and we have that now with Hearts of Iron 4.”
Conveying the Harshness of War
Because Hearts of Iron 4 is set during WWII, most players have a pre-conceived notion of how things should look simply because there’s so much media out there around WWII. This played heavily into the artistic style of the gameplay.
“When it comes to beautiful concept art, there’s a lot of really good artists out there who are willing to share their paintings,” Virtanen explains. “But when it comes to designing and creating UIs, there aren’t as many.”
Virtanen continues, “So that’s something we’ve mainly done through looking at our own earlier games and other people’s games. And then basically you just start trying to iterate on the solutions. This is a good way to go, this is a beautiful interface, how did they do this? And so on and so on.”
“We’re making a specialized game,” Larsen adds. “So while we can be inspired by a first-person shooter, we can’t really do it the way they did it.”
Larsen continues, “I play a lot of games, and when I play I’m always evaluating how they solved different things and how they’ve done it. Always evaluating, would this fit into our game?”
Playing games isn’t the only research the team at PDX used for inspiration. In addition to exploring capabilities of a variety of game engines such as Unreal Engine and Unity, they also considered the historical accuracy of their game’s look.
“A lot of pictures were basis for a lot of things [in Hearts of Iron 4],” Larsen explains. “So when we were doing naval combat for boats, we were looking at aerial pictures of ships being bombed to see how does it look from a flying range?”
Continuing, Larsen says, “Also if you look at a real big battleship firing, you get big waves and so many things happening. We also want to have that effect in the game to show the power of the boat. This helps show it isn’t a small thing, this is a battleship firing these shells.”
“And of course, I can’t even count all of the WWII documentaries I’ve watched,” Virtanen laughs.
“We want to convey the feeling that there’s an actual war happening,” Virtanen continues. “So we have a lot of particles, we have a lot of bombers to drop their bombs, a lot of things should happen to give the players a certain feeling that there’s a war going on.”
To help give the game a more harsh and war-torn feel, Virtanen designed the UI using mostly grayish or black backgrounds with a lot of saturated icons on top of it.
“Hearts of Iron isn’t a game where you’ll look at it and say, ‘Oh that’s a really nice vacation island'”, Larsen states. “It’s quite harsh and contrasting. We don’t have really nice waves, shores and birds flying around. It’s supposed to be quite harsh because there’s a war going on.”
It’s a World at War
One of the biggest challenges for the PDX team was the sheer scale of the game. After all, it is the entire world at war.
“This game has a time cycle, night and day. It also has a year cycle for seasons,” Larsen explains. “So everything should look great from all distances, all angles, during all times of day and during all seasons of the year. Creating systems and solutions for having the game look good at all ranges and at all times is a really big challenge.”
Larsen continues, “Luckily, we had some really good help from coders doing post effect systems to build up volumes and place them all over the world. To control things like, here in the winter at night this area should be this bright and have this much blue, and so on. Also, since this game is the whole world, going through the whole world map and making all the rivers look good, all the shores look good is another big challenge.”
One of the solutions the PDX team employed was through exaggerating the units and tanks.
“Unites and tanks, all are quite exaggerated. They have realistic proportions but they need to be much stronger than if it were a first-person game,” Larsen comments. “So when you build the model in ZBrush to the correct scale you can make any sort of tweaks you need to make sure it doesn’t look too cartoony. All the models are done from the camera perspective to exaggerate the pose so the players can see on the models what kind of units they are.”
Larsen explains further, “It’s tough when just seeing the helmet, weapons and shoulders of a character to tell what kind of unit it is. Of course with tanks and planes it’s easier, but we really want the art to fit the game so we saturate small barrels on some tanks just to make them more visible. [We also gave] higher texture values for our stuff to show a better silhouette around the object.”
Planning for the future
Most of the games PDX have worked on in the past have included quite a few expansions and add-ons that get released after the core game is live. It stands to reason to assume that Hearts of Iron 4 will, one day, see its fair share of expansions.
And the PDX team has already started to think about that.
“I’m older and wiser now,” Virtanen laughs. “So whenever they want me to make an interface I always try to have some more padding or spacing; some empty areas, because I know someone will want to put a picture on that space.”
Even when you plan ahead, it’s impossible to predict the future.
“With future expansions when we add features, you will have to rearrange and work with what you’ve got. But as much thought as you can put into that now will save a lot of work later on,” Virtanen says.
Even with all of the research done, sometimes even the pros at PDX make mistakes that can hurt their future plans.
“Many of our fans are very interested in WWII,” Virtanen explains. “I’ll often get PMs with screenshots of minor corrections that might say, ‘Excuse me, sir, that Panzer 4 didn’t actually have that spare tracking on the side, it had it on the back side.”
“My point is when you start out with a new project you should try to schedule it. Nobody knows how many icons will go into the game when we start making it, but get a general sense of the direction we’re going in.”
A huge part of this can be found by determining what approach would be best for creating the assets in the game and whether or not a 2D or 3D approach would be better.
Continuing, Virtanen says, “So for now we’re trying to have huge tech trees with icons on them. That makes me as an artist say, ‘Okay, I can do these icons. I can model them in 3D, but they’re really, really small so that wouldn’t be time efficient.’ Since we have a lot of interfaces, we try to make an efficient workflow to make it as pretty as you can while still managing to get the job done.”
Advice from the Team at Paradox Development Studios
By now you should have a pretty thorough understand of some of the work that went into the creation of Hearts of Iron 4, as well as hopefully picking up some great tips and tricks to try out in your own games. Before finishing our chat with the team at PDX, we wanted to get some final advice for ways to both get into the games industry and be successful in it.
“I would say be humble and try to be open-minded about things,” Virtanen recommends. “It’s better to be as raw as you can in the beginning and then focus on a niche rather than the other way around. You’re going to have a hard time trying to get to that special animation position because the industry keeps expanding and getting new people.”
For many in the games industry, QA is seen as an entry-level position and a stepping-stone to another position. Then again, the way QA is handled at PDX isn’t necessarily the same as many other studios. And Jenelius is quick to point out that QA isn’t necessarily a means to an end at PDX.
“The last time we put up an ad for QA testers we had like 80 applicants,” Jenelius explains. “So you have to be quite lucky to get the job.”
“I suppose QA is an entry level job, but I don’t like that it is. It shouldn’t be.”
“Especially for us,” Jenelius continues. “Working on complex strategy games, we need people who have actually played our games. We can’t spend half a year teaching people how to play our games. So you have to have thousands of hours in our titles otherwise you’re not going to get it.”
“You shouldn’t apply for QA if you want to be a game designer,” Jenelius states honestly. “That [QA is] entry level or a stepping stone, I don’t like that at all. If you don’t actually want to work QA, you’re not going to enjoy it and we’re not going to get the best production out of you. Everyone loses if you don’t actually want the job.”
“And the other way around, if you’re a very good QA tester and you move onto game design then we’ve lost something. We want you to stay in QA if you’re good,” Jenelius concludes.
With a very open environment where feedback is encouraged throughout the entire production process, you can’t really be tied to your work too closely.
Larsen explains, “Be prepared to get critiqued. Everyone’s going to have feedback, so prepare for that. Take critique and make the best of it. Listen to what people say.”
“You have to be prepared to kill your darling sometimes,” Virtanen laughs. “You may have this thing you put a lot of time and effort into and it looks beautiful, but it doesn’t work for this game, not on this day. You have to be able to take that and realize it’s not time wasted. You still learned a lot, but sometimes things just don’t work out and you have to scrap it.”
If you are looking for a role in the games industry as a QA, Jenelius offers some great advice.
“Apply for betas,” Jenelius says. “Not just playing the games, but testing and giving input. Don’t see it as an early access thing, but we really want your feedback. Learn how to use the bug reports system and how to write a good report and so on.”
“There’s a big difference between testing a game and playing it,” Jenelius continues. “Get into the mindset of, ‘The game wants me to go this way, but what happens if I go this other way instead?’ Have the mindset of not going with the game, but against it and trying to actively break it. That’s where you should start if you want to be a tester.”
And many have found success using this technique. Some of the testers that are at PDX now started by participating in betas.
“The best advice is probably to network if you’re wanting to get into the games industry, whether it’s QA or anything else,” Jenelius concludes. “Also be good at showing what you’re good at. Make a portfolio or prototypes if you want to be a programmer. You can’t go into an interview and tell them you’re good at stuff anymore, you have to show them.”
While just about any creative is familiar with a demo reel, another approach of getting in as a tester can be to show a reel of how you’ve broken a game. This can be done by simply streaming the game as you break it.
“Actually [we’ve hired] some streamers,” Jenelius says. “We knew this guy was good at breaking, in this case Europa Universalis, because we’ve seen it in his streams. That’s also a good idea to actually proving what you can do in streams.”
From an artistic perspective, Paradox’s 3D Artist Joakim Larsen offers some great advice.
“If you want to become a 3D artist in the industry, spend time learning modeling and texturing to build up your skills if that’s what you want to do,” Larsen suggests. “When I wanted to be a 3D artist, I bought a year subscription to Digital-Tutors and went through the modeling and texturing series during the whole year. That got me to a level where I could apply to school to further develop my skills.”
“But also spend time learning how games are made. Understand the features of at least one [game] engine. For example, understand all the features Unreal Engine has, how it works and the technical solution behind it. That way you’ll understand when the coders say, ‘No we can’t do that’, you’ll know the cause. Because you’ll always be at the grace of the coders and what they can achieve.”
Virtanen adds, “I’m a 3D artist as well from the beginning, so of course everybody loves DT. We use a lot of your services.”
Flattering us with his words, Larsen continues, “You can do anything, but you can’t do everything. You can’t know everything. You’ll have to go to Digital-Tutors to learn how to do things. Be open to expand yourself to new ideas and new solutions. Off and on, I would be like ‘How do I do this layer type in After Effects?’ or something like that and I would go learn how to do it. If I go to YouTube to find it, I usually get some guy who starts with, ‘This is how you start to program’ and on and on. But with DT it’s like, ‘This is how we do this specific thing.’ No added crap, just this is how to do it.”