The biggest shake-up in the history of Frontier’s “Elite Dangerous” was the introduction of engineering with the Horizons update. This system allowed players to take materials, which were collected via various game activities but were unavailable from markets, and spend them to roll for modifications on individual ship modules. These changes would dramatically alter the modules used, introducing personal variability to modules, giving users a sense of individuality for their ships, and introducing unpredictability into the game. Now, CCP seems committed to introducing the same elements with “mutaplasmids.”

In a FanFest presentation on “Ships and Modules” on April 13, CCP Rise explained some of the mechanics that had been mentioned in the Fanfest keynote. Unfortunately, it seems like these may have been the danger Aryth warned us about earlier this year.

The basic mechanic is a familiar one: players enter “abyssal deadspace” instances, shoot the things inside, and get special loot related to to Triglavian Collective. This loot includes “mutaplasmids,” which come in three different grades, and which can be combined with regular modules. When combined, the modules are permanently altered with random percentage benefits and detriments, ranging from 10% to 40%. Only certain modules can be modified in this way, excluding resist modules, damage modules, and turrets and missiles. Modules that explicitly can be so modified include propulsion modules, tackle, EWAR, and local repair and booster modules.

The Road to Hell

The intentions are good. CCP Rise explained that the goals they intended to achieve with these introductions were to introduce variations in rewards, to have a major impact on how the game is played, to have a lateral effect that means the changes apply to everyone, to and to give players a greater sense of individual responsibility. All of these are noble goals. Greater variation means more sophistication, consistent with the goal mentioned in the keynote of adding depth to the sandbox. Major impacts and lateral effects mean that the changes, and so CCP’s development efforts, are not catering to a specific, small group of players, but rather will affect all players, even if indirectly through the introduction of a new aspect of the economy. And in a game where the economy is real and players are involved in every element of play, there are few places where a player can feel like their own efforts alone resulted in something. “I built this” is a sentiment that does not apply easily to Eve, so the ability to create new, potentially very powerful modules on one’s own, without the need for vertical supply chains and robust economic manipulations, is very enticing.

But good intentions do not always result in good decisions. During the presentation, CCP Rise used a particular word that underlines all of the problems we saw in Elite Dangerous’ engineering: unpredictability.

The Problem of Unpredictability

In Rules of Play, Salen and Zimmerman define a game as “ a system in which players engage in an artificial conflict, defined by rules, that results in a quantifiable outcome.” The issue of introducing unpredictability is that it makes it difficult to determine what the rules are, and it makes it difficult to quantify the outcome.

Eve is a game with a sophisticated sandbox, where players can gain an advantage through having a deep understanding of the mechanics. Unpredictability undermines this. But we don’t have to rely on academic theory to make this case: we know where this exercise ends up. Elite Dangerous recently revamped the entire engineering system to ensure that players could always continue to upgrade modules until they had the top-tier results. Why? Because a purely random, continuous range of benefits means that when players all perform the exact same behavior, a few players – through no predictable activity or behavior – will simply have better things, the vast majority of players will have strictly average outcomes, and a few players, again through no fault or mistake of their own, will have objectively bad outcomes. In a competitive game, this means that the hand of fate, not the skill of the players, will determine the outcome.

Building a Skinner Pod

Game developers want people to play the game. The longer time people spend in the game, the more likely they are to spend actual money on playing it. We, as players, should also want more players to play Eve. In the opening ceremony, CCP Guard expressed a desire to keep Eve running for another 15 years at least. If we want that to happen then we have to accept that CCP must continue to make money. That means bringing in new players and getting existing players to spend money. All of this is simply a fact of game development and design, there’s no evil here.

But gambling and fostering addiction is a cheap way to go about it. All game design is, in a sense, behavior modification. It’s the common element in vogue, sure. Get players to perform a behavior, use a random outcome to introduce a variable reinforcement schedule. Fixed reinforcement schedules train players to perform a behavior the exact number of times it takes to get the reward. Variable schedules, weighted properly, instead get players to keep performing the behavior. It’s not hard to see why so many games are falling back on it. Perform the loop, in this case, running the Abyssal Deadspace instance. Get the mutaplasmid, apply it to the module. Sometimes get something good. Sometimes get something bad. If the good outweighs the bad more often than note, we’ll perform the behavior again, and again, and again.

There’s no evil in that, this is simply a reality of life. But there are better ways to go about it that don’t rely on unpredictability. Rather than using continuous random chances, so that each module is unique, a schedule of tiered results can be used to create discrete, achievable, equal outcomes. If the range is from between 10% and 40% improvement, make those groups that are pre-balanced with detriments – a 10% benefit to statistic A always comes with a balanced detriment to statistic B, for example – or known ranges: 10%, 20%, 30%, 40%, with nothing in between.

Why? There are a few reasons. The first is that predictability can drive an economy: different tiers of outcomes adds a meta-level that is predictable. This is going to happen organically anyhow as players start identifying the rarity of modifications to stats in ranges. If a “35% or better” improvement becomes the rarest improvement, it will happen naturally. But there is little benefit to introducing unique continuous ranges. We have to remember that if things are unpredictable, the chances of a unique and wonderful outcome are real – but the risk of a uniquely trash outcome are also real.

The real problem is that of valuing our time. Purely random, continuous ranges – unpredictable outcomes – encourage repeated play. The new abyssal deadspace instances introduce a nice closed gameplay loop that players can engage in for rewards, but by making the outcome unpredictable, we do not know that we will actually get rewarded more than punished, and that’s the key to getting players to repeat the behavior. If my goal is a 40% improvement, and I have absolutely no idea how likely it is I will get that, then I may just end up giving up. If on the other hand I know exactly what I’m working for, and I know it’s possible to get it, and in fact I know exactly how likely it is to get it, then (and only then) can I make a decision about whether it is worth my time. There is no way for me to make a decision regarding lost opportunity costs when an outcome is truly unpredictable, and that’s a bad look for a game that can hook addicted whales at the cost of players who don’t have the time to roll the dice on their opportunity costs. Respect for the player is important, and unpredictable outcomes do not respect the player’s time or effort.

A Game of Chance, or a Game of Skill?

Eve Online is absolutely a unique experience. The sandbox is deep and vast, and in many cases CCP itself acknowledges that it has created a virtual reality, a complicated virtual world, where the players drive the action. It is a struggle, then, because introducing gamification into this virtual world threatens its function. Players are able to engage with the sandbox and in some cases bend and break the rules exactly because they know the rules. CCP Hilmar has said many times how he is regularly amazed by Eve’s players’ ability to discover how things work and find ways to gain an advantage. This is not unlike the real world, where scientific discovery or military intelligence both serve as only two examples of the many ways in which understanding something better means having a competitive edge.

The introduction of unpredictability defeats this. When a system is unpredictable, players cannot use it to gain an edge tactically or strategically. They cannot use it to develop doctrines, they cannot rely on it for an advantage in a fight. In the real world, things may not always be known, but they are always knowable.

Pull Up!

Eve isn’t Elite. Perhaps the most noteworthy difference is that Eve’s markets allows for trade and an economy. The destroyable nature of things means that no module is going to persist forever to break the game. This is probably not a catastrophic end to the game. But it is certainly a risk. A little gambling isn’t a problem on its own, unpredictability has a place – we don’t know if every shot we take will hit. Some fights are won or loss based on a random missed shot. On its own, unpredictability is not terrible because it can be accounted for. But looking to introduce that unpredictability in a way that has a “major impact” and a “lateral effect” is a dangerous trend towards gamification for the sake of gamification.

I don’t think that mutaplasmid modifications are going to become The Cancer That Is Killing Eve™. It’s not likely simply because the unpredictability means they can’t be used to change the meta. The danger here isn’t from the modules themselves, but from the trajectory they represent. If the efforts that CCP undertakes are meant to have major impacts, to affect every player, then this is good – that is development time well spent. But CCP should look carefully at these things before introducing them. The likely result of introducing unpredictability isn’t unpredictable.