Gregz
Arcane
An anonymous mobile games developer spills his guts about player manipulation, and it's much worse than I imagined:
As titles like Diablo Immortal raised discussions around predatory monetization earlier this year, I feel it’s a good opportunity to share my take on it. I believe predatory practices are an increasing issue in the industry, because developers become better at hiding them and because predatory monetization is mainly used in GaaS (games as a service) which tend to become the norm, even outside of the mobile sphere.
The mobile gaming industry is not black or white, I've worked with incredibly talented and genuine people who always had the player's interest at heart, as well as on games that reached the top charts while always respecting their player base.
But I've also seen unethical decisions being made. And while ethics related to the workplace have been in the spotlight for the past few years, ethics related to monetization are unfortunately still rarely discussed within companies or even in the industry at large.
Predatory practices tend to appear when the top management puts pressure to grow the revenues, when the few people in charge of the roadmap or monetization don’t care or don’t question the morality of these practices and their impact on players, and when the team members who disagree with these practices gave up on fighting against them.
Truth is, when a company starts abusing predatory practices, it’s often symptomatic of a bigger change of mindset & priorities, and the experienced, creative talents who care the most about players leave to be replaced by people who are aligned with the new mindset. This often leads to less opposition, more of these practices over time, and more talents leaving… which gives me little hope that companies can make a positive comeback.
Many videos already do a great job of deconstructing games to expose the psychology and design behind FOMO, loot boxes, battle pass, or premium currency packs but this is based on information visible to the player…and is unfortunately the tip of the iceberg. The most pernicious predatory practices are systemic, buried deep in the code, and fully understood by the few people who directly worked on them, and a lot of effort is spent to obfuscate these practices from players, precisely to avoid community backlash (or legal issues).
The concept of personalization and how it’s combined with fake randomness is in my opinion the root of many predatory practices.
Like any sport or physical game, most players assume that the rules of the video game they’re playing are the same for everyone. In mobile games or GaaS, it’s often not the case, 'personalization' is a widely used practice that dynamically changes the rule of the game under the hood to optimize spending opportunities.
- Personalization
Similarly to personalized ads or news feed when you browse social media, when you play, data is collected and processed through all sorts of algorithms that analyze your play and spending behaviors. It can then dynamically changes variables like the chances to get a reward, the difficulty of a level, or even the $ price of an offer, to put you in a situation where you’re more likely to spend. Any large mobile studio has dozens of market analysts, data analysts, and even psychologists who analyze players and their data to refine these algorithms. Personalization is present in pretty much any GaaS, what differs is the degree to which it’s exploited to squeeze players for money, which is often hypocritically presented as ‘Delivering a personalized experience’.
Mobile game monetization is straightforward, as a player, you engage with various progression systems and once in a while, a spike in difficulty slows down your progression. To overcome it, you can either pay or grind. The pinch is here to frustrate you just enough so you feel like it's worth spending 5$, but not frustrated to a point where that’d make you quit the game.
Finding and maintaining the sweet spot is challenging for the developers but it makes the difference between a failed project and a hit game. But ‘problem number 1’ is that this sweet spot differs for every player, depending on factors like their playstyle, motivations, experience, or how much money they spent in the game. The concept of ‘spending depth’ or ‘problem number 2’, is that if a player is willing to spend 20k, they should still feel the need to spend money after spending 19k, hence they have to be kept in this sweet spot of frustration, regardless how of much money is already spent.
I say ‘problem’ because those are only problems for greedy companies that pursue infinite spending depth, often at the cost of employees' and players' satisfaction. There are plenty of ethical mobile games that generate more than enough revenue to sustain hundreds of people for years without prioritizing these problems. But for companies that do, personalization is the ideal solution.
To illustrate with concrete usage, imagine that Mary and Sam are both playing level 97 in Candy Crosh. Mary is an experienced Match-3 player who already spent some money to get in-game advantages, while Sam is new to the genre and a non-spender. If the level’s difficulty was the same for both, Mary might win it easily and not feel the need to pay, while Sam might get too frustrated and quit the game.
If a model predicts that players are more likely to spend if they lose 4 times in a row with a few moves left, the difficulty variables can adjust themselves to ensure this situation will happen for both Mary and Sam, maybe by reducing the chances of candies to spawn for Mary, while increasing them Sam. Both will beat the level on their 5th try, probably thinking that they outsmarted the game this time, without realizing that it’s all decided in advance.
Or if another model predicts that Sam is about to quit the game because he has fewer and shorter sessions, the game might suddenly become easier and give away more rewards, the same rewards that Mary would have to grind for or pay for.
Now, adaptive difficulty or personalization is not new and can be a good thing if used with the right intentions, after all, most console or PC games let players choose their own difficulty, and games like Bioshock Infinite dynamically change the loots based on the player’s inventory, and smart matchmaking is a thing in competitive games for many years. But it is often insidious in mobile games for a combination of reasons:
In some cases, personalization can reach a point where 2 players from the same country have very different $ price points for the same in-game offer, solely because the game knows that player B is willing to spend more. And legally, nothing forbid companies from doing it.
- Making more money is the main intention (unlike Buy 2 Play games where it’s undebatably here to create a better experience).
- It's not transparent: Players are not aware that many variables in their game are adjusted on the fly and different from other players.
- It’s imposed: Players can’t opt out (The only data that require the company to ask for the players' permission are personal data, like country or name).
- It's combined with multiplayer features. Many mobile games have social features like tournaments. It’s as if in a shooter, regular players were unknowingly competing and investing money to play against players with an aimbot.
- Paying might actually backfire on the player. Imagine that a player spends $5 to guarantee a 6 on a dice roll in Monopoly, but after doing so, the game would subtly increase the price of the properties to ensure it doesn’t become too easy. They might feel instant gratification, but in the long run, the game is adjusting to keep them in the frustration zone.
Now if Mary, who is a loyal customer and already spent 2000$, realized that Actovision is ‘personalizing her experience’ to squeeze her for money, she wouldn’t be very happy and might even send a legal threat to customer care.
In general, simply because the vast majority of mobile players do not communicate with other players, the risk of community backlash or players realizing that they are playing a completely different game than their neighbor is low.
But even if they do, luckily for the developers, there is an easy way to exploit personalization without players noticing it.
2) Fake randomness
It’s not a coincidence if gacha, match-3 puzzle, and casino games dominate the revenue top charts, and if Blizzard’s first mobile game is a hack & slash.
What do all of these games have in common?
Progression is dictated by randomness, and randomness is very easy to manipulate.
Randomness in itself is already a powerful monetization tool, when selling loot boxes, developers don’t sell direct value to the player, they sell them a chance to get value, which reduces how quickly the content is consumed and how much needs to be produced.
But randomness combined with personalization can also be used in meta systems to hide soft or hard gates and regulate the players' progression.
Imagine a game where beating monsters rewards players with experience points needed to level up, and as developers, we want players to gain 1 level per week, regardless of if they beat 30 monsters or 10 monsters per day. If it was a fixed amount of points, it would be too obvious that player B is being punished for engaging too much with the game. Instead if it rewards them with a ‘random’ number, we can easily adjust the odds so the more monsters you kill per day, the more ‘unlucky’ you are with the drop.
Another more pernicious example with a gacha system. Let’s say 100 new heroes are released and players can unlock them through loot boxes, as a player, it’s presented in a way that any of these heroes can be unlocked from day 1. Some players might want to unlock them all and decide to spend thousands of $ on loot boxes. But what players don’t know is that half of the heroes are locked behind a time gate and will only have a chance to drop 1 week from now. It’s not that the odds are unfair, they might not even exist. It’s not any different than the rigged carnival games.
Legally, companies are obligated to disclose the odds of a loot box, but they don’t have to be too specific, so it’s extremely easy to get around it. They can just say ‘10% chances to drop a rare hero’, and so the loot box will drop a rare hero…simply not the one the game doesn’t want you to get.
I see games as a pact between players and developers, players trust the developers to be transparent about the rules of the game, and in exchange, the developers trust the players to play by the rules of the game. Personalization and fake randomness break this pact as the rules are not transparent in the first place and change on the fly without the player’s consent/awareness.
My advice for players would be to stay away from mobile games that abuse randomness in their systems or monetization, and I salute Supercell’s recent decision to remove loot boxes from their store.
Finding the line between what it’s ethical and what’s not is tricky, but this blurry line is often exploited by unscrupulous people to push their agenda, which always hurt the company in the long run. As developers, if you are in a situation where you need to hide a feature or some balancing behind randomness to avoid community backlash, it’s probably not a good idea to have it in the first place.
