Designer Notes #18: Offworld Trading Company

In this episode, Bruce Geryk interviews Soren Johnson about his new economic RTS, Offworld Trading Company. They discuss how exploring a black map is one of gaming’s greatest hits, why the hardest part of designing Offworld was ending the game, and why Early Access games shouldn’t have QA. Also, listen to hear Soren correctly pronounce timbre!

Games Discussed: Offworld Trading Company, StarCraft, Age of Empires 2, Railroad Tycoon, M.U.L.E., Belter

OTC Designer Notes #19: Special Thanks

The following is an excerpt from the Designer Notes for Offworld Trading Company. The game, an economic RTS set on Mars, released on April 28, 2016, and is available for purchase here. (A Game Almanac, which includes the full Designer Notes, is available as free DLC.)

Offworld Trading Company was not an easy game to make, perhaps most especially because people needed to believe that it would work in the first place. Thus, I need to thank Brad Wardell, Derek Paxton, Brian Clair, and everyone else at Stardock for believing that an economic RTS would actually be fun.

I also need to thank the team for taking the leap with me and working so hard to build the game. Dorian Newcomb, my business partner, was the game’s official Art Director and unofficial Producer, making sure that life on Mars looks amazing and that everyone was working in sync. Jason Winokur was the first programmer to join us, and he handled the graphics, supported modding, and generally made sure our project was in great shape. Dave Wagner implemented multiplayer, replays, saved games, leaderboards, and all the other details needed for a modern game. Jim Alley created all of our interface art as well as our great logo. Josh Hardy, after being lured away from the Star Control team, made our building, colony, and HQ models as well as our effects. Joel Bowers implemented the scenario system and then built the tutorials on top of them.

Zack Fowler implemented our terrain system and built the map editor on top of it. Tommy Truong helped us finish, animating our buildings and HQs. Mark Cromer made our sound effects and helped with audio design. Christopher Tin composed an amazing soundtrack that scales with the player’s progress. Kirby Runyon created 32 maps by hand, based on real locations on Mars. Andy Hull added some nice polish to our character popups (and invented the giant check for losing). Erik Ehoff made some key early concept art the helped to define the game’s look. Shentloc, our localization team, did an amazing job to help us ship in nine languages. (Ten if you count British aluminium!)

I’d like to thank my wife Leyla who, perhaps unexpectedly, fell in love with the game and sunk 1,000 hours into it, her very first RTS; her excitement and encouragement helped me keep going during the rough times. Finally, I want to thank the fans that provided us with amazing feedback and support during the pre-release phase; the game would have been much different (and certainly worse) without them. I specifically want to acknowledge the contributions of the following players: Zultar, Cubit, PBHead, Gameslayer, Blues, Blackmagic, Death Tacticus, Indczn, Kingmorgan, Jaiwera, YerAnd, GalacticWino, Roler, TheSpinCycle, Veivi, UltraPope, Dermas, Heisenberg, Showcasemike, and Sir Rogers.

Offworld would still just be an idea in my head without the amazing team we have at Mohawk, and I will forever be in their debt for believing in the game and committing to making it great. I think we will all remember that year when we were playing the game internally as one of the best of our careers — iterating on it constantly based on daily games while slowly becoming aware that we were making something truly unique. Actually finishing the game was very hard work, so I am very proud of what we created. I hope the game will mean as much to you.

OTC Designer Notes #18: Adaptive Gameplay

The following is an excerpt from the Designer Notes for Offworld Trading Company. The game, an economic RTS set on Mars, released on April 28, 2016, and is available for purchase here. (A Game Almanac, which includes the full Designer Notes, is available as free DLC.)

We are most proud of Offworld because it makes the player think about each game differently, adapting to the events and environment of that specific match instead of using the same build order or pet tactic over and over again. The game encourages adaptation because so many key parts of the game are randomized each time:

  • Random MapsEach map is randomly generated, with different quantities of each resource available. To encourage interesting randomness, we associated each resource with specific terrain types (Carbon with Craters, Water with Lakebeds, Silicon with Sand, and so on), which determines how likely it is that resources might appear on a tile of that terrain type. Then, we limited these types of terrain to different sections of the map; if Sand fields are in the northwest while Lakebeds are in the southeast, then the Silicon and Water will be in different parts of the map. By preventing different resources types from being too close together, each potential founding location is defined by what resources are close and what resources are not. We also create a dead zone of resources near the middle of the map surrounding the Colony to encourage players to found near the edges of the map, making longer shipping lanes more necessary.
  • Black MarketThe black market typically has six to eight items for sale, and they are selected at the game’s start from a set of eighteen (not unlike the ten card stacks chosen at the beginning of a game of Dominion). As mentioned previously, they are chosen not purely randomly but with certain guarantees (for example, either an EMP or a Power Surge will always be available). Advanced players watch the black market carefully before deciding which HQ type to use. For example, Underground Nukes penalize Scavengers (who are so dependent on maintaining their Carbon supply) but barely affect Scientists (who can still put secondary buildings on top of Trace levels of a resource). Pirates and Magnetic Storms are dangerous for Scientists as they often ship expensive resources across the map; on the other hand, EMPs and Power Surges favor Scientists as they have protection against both. Circuit Overload is dangerous for Robotic players as they like to maintain a positive rate of Power. Expansive HQs are a good choice if Bribe Claim is not available because they will be the only ones with extra claims. Furthermore, the items change how each other can be used; Holograms are much more powerful in games without Spies than in games with them.
  • Random Prices – With the Random Prices option turned on (which is highly recommended for veterans), the starting price of resources also changes from game to game. Although half of the resources stay the same, a quarter of them are reduced by 50%, and a quarter of them are increased by 100%. This option was added once veteran players developed general starting strategies for most of the HQ types. For example, two Steel Mills and a Metal Mine on Iron is a typical Robotic opening; however, what if Iron starts high at $40 and Steel starts low at $30? Then, a Steel Mill is going to lose money by converting $40 of Iron into $15 of Steel (although the actual conversion should be a little better because of adjacency bonuses). The Robotic player can stick to the familiar strategy but might be better off looking for something else. Perhaps in this scenario, Power started higher and Aluminum lower? In that case, a Geothermal Plant would be much cheaper than normal and immediately produce some serious money. Basically, the best players will reevaluate all of their opening moves depending on the set of prices revealed at the beginning.
  • Random EventsDuring the game itself, random events occur that shift prices of resources significantly. Some events (Oxygen Surplus, Food Shortage, etc) will affect just one resource, driving the price either up or down; these events are the ones which can be created artificially with a Hacker Array, so players should always view them with a bit of suspicion (even if a Hacker Array is not visible as it can be hid with a Hologram). Other events affect multiple resources at once; for example, a Pipeline Leak will drive up the price of Food, Oxygen, and Fuel. Finally, Solar Flares and Dust Storms affect how buildings function; while the former boosts Solar Panels, the latter penalizes them but boosts Wind Turbines. Taken together, all these events ensure that players need to adapt to random circumstance during the game as well. For instance, a Silicon Shortage might suddenly make Glass Kilns unprofitable; should the player scrap them for something else, or just hold on until the market balances out again? The answer probably depends on many factors, such as the player’s stockpile of Silicon, the Colony’s demand for Glass, whether the player still needs Glass to upgrade, the cost of switching to new buildings, and so on. The random event, however, forces the player to consider the situation carefully.
  • Other Players The decisions made by other players are, of course, not actually random, but they definitely require the player to adapt. For example, the desirability of each HQ changes depending on the other founds. If other players found Scavenger, then perhaps the player should pick Scientific, which has good protection against sabotage. If everyone goes Scientific, then perhaps founding Robotic and building Power is a good move as Scientific players consume plenty of Power with their secondary buildings. Seeing many Robotic players means, of course, that life support resources will not be in such high demand. Besides the founds, players need to pay attention to what everyone else is building. Is everyone skipping Power? (If so, put down an early Geothermal Plant.) Did most players build Farms and not Reactors? (If so, do the opposite and build Reactors..) Are multiple Patent Labs and Optimization Centers up early? (If so, build Refineries as the price of Chemicals is about to rise.) Basically, watch what everyone else is doing to anticipate where the market is heading.

We hope that we’ve made a game that lives up to our ideal for strategy games — that the player should always be planning, always be reacting, always be thinking.

OTC Designer Notes #17: Stock Market

The following is an excerpt from the Designer Notes for Offworld Trading Company. The game, an economic RTS set on Mars, released on April 28, 2016, and is available for purchase here. (A Game Almanac, which includes the full Designer Notes, is available as free DLC.)

The hardest part of the game design to get right – and one that changed significantly only a few months before release – was the stock market. The system had both a thematic purpose (a modern economic game needs a stock market to reflect success) and a gameplay purpose (the system serves as a desperately needed victory condition for a game in which wiping out the opponent militarily is not an option). Basically, we needed a way to end the game that was more interesting than simply counting who made the most money.

Our initial idea involved buying out other players and actually acquiring their HQ as well as all of their claims and buildings. After a player had no more shares available on the open market, she would be vulnerable to a buyout, which any player could trigger by paying double value for all of the shares owned by other players. Thus, a player could defend himself by buying up his own stock or attack other players by buying up their stock, in preparation for a later buyout.

This system worked reasonably well except for two problems. First, players often felt that the game spiralled out of control following a buyout because they inherited a lot of new buildings all over the map that they didn’t have time to manage. Second, buyouts were all-or-nothing affairs; when two players were racing to buyout a third player, the winner would usually snowball forward and easily dominate the rest of the game (or, if we tried to balance out this effect by making buyouts too expensive, an even worse situation occurred in which players saved up their money instead, in hopes being able to afford the final buyout).

We solved both these problems with the subsidiary system, under which players no longer acquired all of the eliminated player’s buildings and claims but instead now owned shares of a new subsidiary, which is the eliminated player’s company run by a modified AI that focuses solely on making money (and avoids advanced options like sabotage, patents, auctions, and stocks). The subsidiary simply exists to distribute its profits to shareholders, according to the ownership percentage. This system is much easier for players to handle as they don’t need to manage their subsidiary’s buildings and also easily handles split ownership, avoiding the problems of the original all-or-nothing system.

Because subsidiaries enabled partial ownership of companies, we were able to add one more wrinkle to the system — the majority buyout, which occurs if the other players own six shares (meaning more than half) of a player’s stock. In a majority buyout, the targeted player is instantly eliminated and turned into a subsidiary. This feature prevents players from staying in the game far after the point it becomes obvious that they don’t have a chance; if a player’s rivals own six or more of his shares, his chances of winning are slim at best. Most importantly, the player is not left around with an opportunity to unbalance the game by, perhaps, using sabotage maliciously to keep a specific opponent from winning. If a player has clearly lost, the company should be run by the neutral, subsidiary AI as soon as possible.

One final important change to the stock system came very late, less than five months before we shipped the game. Some players asked for more transparency in how close players are to buyouts. Thus, we experimented with a system in which players could buyout an opponent’s self-owned stock one-by-one (for double cost). Stocks owned in one’s own company became the equivalent of hit points, and when a player ran out of stock, her game was over. The system led to some tense games in which players could easily see how close the game was to ending as they lost control of their own stock, one share at a time. For the first time, three-player games were playable because the two leading players could each end up with half of the last-place one. However, free-for-alls with more than three players suffered because players could team up against one specific player and buy one or two shares each, forcing him out of the game.

Next, we tried a middle ground. Players could buyout shares one at a time until only five shares remained; at that point, the remaining five shares have to be bought all together. This mechanic struck the right balance between making buyouts more transparent while also protecting players from being knocked out by the group. One final small change gave us our final stock system – during a buyout, players have to pay 20% extra for each share owned by a third party, which gives a bonus to players who invest in buying some of the first five shares. Many players didn’t even notice this final tweak, but it felt better; if a player buys the first five shares in an opponent, she is now best positioned to finish the buyout.

We are often asked how Offworld compares strategically to more conventional RTS games. Some parts of the game are so different that it is hard to even make comparisons; however, the development history of the game’s stock market is interesting because it ended up in a place that mirrors the most famous dynamic in RTS games – rush vs turtle vs boom. As a short explanation, these three strategies have a high-level rock/paper/scissors relationship. A rusher (who sacrifices economy for early military) will beat the boomer (who sacrifices defenses for a strong economy). The rusher, however, loses to the turtler (who builds defenses in preparation for an attack). Finally, the boomer beats the turtler by outpacing him economically by the end game.

Strategically, our stock system parallels rush/turtle/boom. Going for an early majority buyout of another player is akin to the rush, which is done by investing one’s money not in economic growth but in buying another player’s stock. This rush, however, can be beaten by a turtle strategy, which means buying up shares of one’s own stock, especially that all-important fifth share that prevents a majority buyout. However, the turtler will lose to a player planning to boom by investing as much money as possible in upgrading her HQ, expanding production, and usually building the first Offworld Market. Completing the circle, the boomer risks losing to the rusher who only needs to buy those first six shares to knock her out. Most games of Offworld have specific key moments when this strategic tension is clear – when a player buys his own fifth share and blocks a majority buyout attempt or when a player spends all her money to upgrade to the final HQ level but forgot to buy enough shares to defend herself. The original stock system did not have these dynamics, so it’s interesting to consider how we came upon the rush/turtle/boom mechanics of a traditional RTS, basically on accident.