Designing the real-time stealth and combat of Desperados III

Desperados III from developer Mimimi Games and publisher THQ Nordic is a Wild West real-time tactics game that serves as a prequel to 2001's Desperados: Wanted: Dead or Alive. It's drawn critical success thanks to its well-implemented storyline, stellar maps, and hardcore tactical gameplay.

We spoke with Dominik Abé, creative director and founder of Munich, Germany-based Mimimi Games to find out more about the game's development and its brilliant tactical stealth game design.

Who are you, and what is Desperados III?

Hi, I’m Dominik Abé. I’m the creative director and founder of Mimimi Games. Desperados III (Steam, GoG, Xbox Live, Playstation Store) is a prequel to the beloved classic Desperados: Wanted Dead or Alive (2001). It’s a story-driven, real-time tactics game with a focus on stealth, set in a ruthless Wild West scenario. You play up to five characters simultaneously and combine their unique skills to master large maps swarming with enemies. But the best way to understand the game is to check out the free demo we have on GoG.com and Steam as well as Xbox One (a PS4 demo will follow soon).

Desperados III identifies itself as being a “Real-Time Tactics” game. What would you say that means?

The game runs in real-time. While in safety, you have to carefully plan your next steps if you want to succeed. When you engage your enemies, you have to execute everything with the right timing. There are no artificial pauses where somebody is waiting for someone else to make his move or turn. Still, you can pause the game with the so-called “Showdown Mode” at any time to rethink your tactics or to synchronize your characters’ unique skills. For me personally, real-time creates a more believable game world where everything is simulated and reacting to you at any time.

Perhaps it might be useful to contrast it with a few games that have elements in common? So briefly, how is Desperados III similar to, and different from, say, Hitman?

Hitman is often called a social stealth game. Meaning the player is trying to blend into a scenario and mostly executing their plans while nobody is looking. In Desperados III you usually play in hostile environments, where you have to avoid to be seen all the time. However, similar to Hitman, we also have a spy-like character, who can disguise herself to blend in. Also, in Desperados III we added the so-called “Civil Zones.” This is a new feature in the real-time tactics genre that allows you to move freely in designated areas of a level, as long as you don´t act suspicious. In the “Civil Zones” you can also listen in on conversations of NPCs to get hints on new gameplay options. The major difference from Desperados III to a game like Hitman is that you can control up to 5 characters simultaneously and combine their skills. For example, the spy can distract an enemy and make him look the other way so another characters can sneak by–or you can take down multiple enemies, who are watching each other, at the same time if you combine the skills of the characters in a clever way.

Perhaps the game’s most distinctive feature is “Showdown Mode” where (except on the hardest difficulty) the game stops and you can give orders to everyone that they’ll carry out immediately upon leaving the mode, which I think is an interesting solution to the problem of doing a mostly real-time game where you play multiple characters. What does it add to a game like this?

It is the tool that allows you to coordinate simultaneous actions with your characters by synchronizing the unique skills in one move. Due to the pause feature it also allows you to react to unexpected circumstances. Making an elaborate plan with 5 characters at the same time and seeing it unfold can lead to the most satisfying moments of the game. We decided against chaining more than one main player action because we want the game to stay real-time as much as possible. Long chains of pre-planned actions would make the game feel very different and would nearly change the genre into something like Door Kickers. In many instances you only plan one action of a character and play the other character in real-time while the plan is executed.

Desperados III itself tells players that they’re intended to save often, and it even pops up a timer, after a short while, reminding players how long it has been since their last save. This is in stark contrast with another philosophy, that of games that don’t allow backup saves at all, those with “permadeath.” It is an interesting decision that it kind of leans into quicksaving. How would you say the design of the game differs as a consequence of this decision?

From a player’s perspective: Having the ability to quick save all the time gives the players the minimum punishment for failure. This encourages the players to experiment more and be more creative in their approach. It lets you iterate your plan until your execution and all the details are perfect. Succeeding with a plan after several tries can be very satisfying.

From a design perspective: The game needs to be very precise regarding the game design and programming because every single moment can be recreated any time and it is expected to behave the same. Randomness is kept to a minimum because you could always get the better odds by loading your save game, which would make it redundant. The core game rules can be harder and less forgiving because the players can understand them more easily as they will try the same situations multiple times. As designers we have little option to cheat the players because they will always be able to reconstruct what we did.

I’ve heard it said that Desperados III is basically a puzzle game, where each mission has certain ways that the player is intended to solve it. At the end of each mission, however, there’s a number of alternate goals given by the game as a way to encourage players to try different things. Would you like to point out one or two of these goals as a particularly interesting thing for you to design and/or players to pursue?

We would say the game is one part puzzle game and another part tactical sandbox. We always design the levels with a golden path and some sub paths but the real fun for us begins when the players choose their own paths and strategies. The core mechanics are way too open to create situations or puzzles where only one solution would work. The alternate goals (we call them badges) at the end of a level are supposed to encourage the players to retry missions with different approaches. Most of the badges came about because we played the game and came across interesting ways to approach a mission. I would suggest to play some of the earlier maps like Flagstone, which don’t take too long, in order to see how different a second play-through of a mission can be.

One interesting aspect of the game is the way the game makes apparent enemy vision cones, especially those parts of the cones that are safe to be in if the player is ducked (as striped regions). This is a nod to playability over realism, of course. Was it ever a question if this would make things too easy, and did it affect the design of the missions?

One of our design pillars for Desperados III was to give the player as much information as possible. We do this because we always want the players to know why their plan fails. Of course, this exact information is far from realism but it represents basic perception of vision and sound in a way that is easy to understand and fast to process. Since the player has this information at hand, we can design very complex and exact timed enemy interactions. Understanding and analyzing those patterns is part of the challenge. Finding the right counter measures to tinker with those clockwork-like routines is a major part of the fun.

Perhaps related to that, what would you say is the role of randomness in a game like this? Is there any? Is the game world like clockwork or chaos? What do you think about randomness in a game world like this?

The quick save/load system makes most of the randomness redundant for this game because the player could always reload for better odds. Not having this “tool” at your disposal introduces some design challenges. For example, randomness in games often leads to a lot of joyful moments like getting legendary loot or a perfect maximum damage shot. A shot in Desperados III will always hit the target, which is essential for planning but it removes the element of surprise. Nearly all rules are binary–something either works completely or not at all. Designing a game with only binary rules can be very hard because you cannot balance the game by increasing or decreasing some values; you would have to remove/add a rule or feature completely. The only randomness in the game rules are the way the AI searches for you during an alarm or investigation phase. In this part of the gameplay loop the players can react dynamically to what the enemies are doing. Sometimes they are lucky and they will not get detected or even get the chance to perform a free take-down.

Desperados III borrows a lot from your prior game Shadow Tactics: Blades of the Shogun. How did you refine its systems for the new game? What would you say the game takes from the past Desperados games, which were released over a decade ago?

Shadow Tactics’ two main inspirations were Commandos and the original Desperados. When we designed Shadow Tactics, we tried to get the best out of both series. Then we modernized them and added our own touch. Therefore, the core design system is quite the same in Desperados III. We tried to improve and enhanced those systems. A lot of work went into improving the players feedback and more precise information–for example: showing how reinforcements would spawn on a map, displaying off-screen markers for guards or aiming lines with more information, etc. Those are rather small improvements but overall, they make the game more accessible and less frustrating.

Other things like the “Civil Zones” (areas where the player can move around freely) add a new layer on top of the core systems. Player actions now are rated as hostile, suspicious or neutral. For example, luring a guard into a civil zone to a certain location does not work if the player is seen while executing the skill because it counts as a suspicious action. Aiming with a knife in those zones counts as hostile and will alert the guards after a short time. In a hostile zone this would not matter because the players are considered hostile at all times, except disguised characters (i.e. a skill of the spy character Kate) or mind-controlled enemies (i.e. a skill of the voodoo character Isabelle). A main addition that was directly inspired by the Desperados series was the enhanced focus on gun gameplay – but the design is very different as we wanted to remove the randomness from it.

In the original Desperados guns have a hit chance depending on distance and gun type. In Desperados III we designed gunplay in a way that is very precise. We also removed the “action” part since the game can be paused. Therefore miss-clicking is not part of the challenge anymore. The challenge in your gun fights focuses again more around tactical elements like positioning, visibility and cooldown management.

At the end of a mission there is an entertaining end-of-level recap, where the player’s actions are pointed out on a map. I like that as a way to compile the player's actions into a narrative. What do you think that adds, and where’d you get the idea?

Because we know that a playthrough of a single map can be quite long we wanted to reward the players for their time spend. We want them to celebrate and recap what they were doing the last few hours. For me personally, the timeline after an Age of Empires match was a bit of an inspiration for this. Also, the replay shows the “perfect” timeline / playtime of the playthrough without the reloads. That is why the real playtime is way longer than the time shown in the replay. We also believe that it is a nice thing to share and compare with other players as to how they approached a mission. A whole playthrough video would be quite long for that.