For Science!

How Your Mind Screws with You in Games Like Diablo

Diablo Cognitive Bias 640

Is your mind tricking you into believing something that simply isn’t true about the game you play? It’s definitely trying to.

Our tendency to think in a way that leads us to draw illogical conclusions is called a cognitive bias. Here are 14 cognitive biases that may be affecting the way you play your favorite game. We’ll frame many of the following examples in the context of Diablo 3, but these biases apply to most games – especially games that include an element of chance, including dice-based games like Dungeons & Dragons.

1. Confirmation Bias

This is our tendency to focus on information that confirms our preconceptions, and we often recognize when others fall prey to this fallacy on forums.

At some point in Diablo 3, a new character named Kadala was introduced. Basically, she’s a glorified slot machine. A superstition began to emerge in the community: you could improve your odds of winning by “spam-clicking” Kadala – that is to say, gambling as fast as you possibly can. When a new forum thread would bring up this theory, you would inevitably see resulting reports from people who tested it and – gasp – appeared to have positive results!

Here’s what was really happening: by gambling faster, you’re not winning more often, you’re just getting the same number of wins in a shorter period of time. We discount the gambling cost and only consider the time cost, which grants the illusion that we’re getting more wins overall. Our mind latches onto the “evidence” that confirms our initial belief: all these wins must mean this myth is true!

This ties into…

2. Expectation Bias

This bias leads us to only report data that agrees with our expectations – and to ignore data that doesn’t.

One reason superstitions propagate is because people are far more likely to report “positive” results than negative results. This gives the semblance that a far greater percentage of people are confirming a given myth than actually exists.

Let’s say 50 people test a claim. 10 of those people excitedly post the “positive” result that fast-gambling worked for them. Five people post negative results – that they saw no difference in win rates. And 35 other people get negative results, but can’t be bothered to follow-up and reply to the thread. The status quo is something that people often don’t feel the need to report.

The situation becomes worse when the negative reports then become marginalized and ignored – surely, these few negative reports are the outliers who failed to rigorously test the hypothesis, right?

3. The Availability Heuristic

This is our tendency to overestimate the likelihood of events that have greater “availability” in memory – in other words, how recent the memories are, how often they appeared, or how emotionally charged they are.

For example, if you are a forum-goer or Redditor, you will tend to see a lot of excited reports of people encountering something that has a really rare chance of occurring, such as a rare item drop. Even if you know how rare the item is – even if you have the raw data on the odds of finding the item – the flood of reports will overwhelm your mind’s ability to keep in mind how rare it is, and you’ll think you’re really unlucky for not finding it. In fact, you’re probably within the statistical norm.

This ties into…

4. The Von Restorff Effect

Simply put, something that is unusual is more likely to be remembered. Another popular Diablo myth involves “double drops” – getting the exact same item drop twice in rapid succession. This myth often also encompasses when someone in your party gets the same drop you received shortly after you do – because we need to stretch the definition in order to find enough examples of this happening to perpetuate the myth.

Theories about server lag causing some form of rollback glitch abound, and this myth survives on the hope that someone can find a way to control this effect – such as with the Diablo 2 duping exploits. But we’ll discuss the illusion of control later…

Here’s the thing: how often do double drops happen? We believe they happen more often than they do, because we are discounting the hundreds of times we don’t get the same item back-to-back. The odds of the same item dropping twice in a row are exactly the same as any two other specific items dropping one after the other. On two coin flips, you’re just as likely to get heads twice in a row as you are to get a head followed by a tail. But getting the same item twice in a row appears unusual to us, so we remember it better, granting it more “availability.”

5. The Availability Cascade

Feedback loop

This self-reinforcing cycle leads a community to adopt a certain collective belief. As a new idea enters a community and rises in popularity, a feedback loop is caused: because it is rising in popularity, more people seek to adopt it, and so it continues to rise in popularity.

We often see this in the form of insight or advice gleaned from top-end or professional players across any game. People watch the top-end players to learn how to become better and pass along what they believe to be advice to other average and new players. The problem is that these observations are being made in a vacuum, without regard for the numerous other factors that come into play.

The assumption being made is that if you take tactic X and apply it to an average player, it will have the same affect as on a top-end player. That advice may be beneficial to a pro player, but detrimental to an average player that lacks certain skill sets, knowledge, or gear. On the surface, though, the advice seems solid, especially when it is being demonstrated by an expert, and so the advice propagates throughout a community.

During my years in a StarCraft clan, I would see average players pass along advice to new players all the time. Often, this advice would only benefit advanced players. New players – who should have been working on mastering the fundamentals of the game – were instead learning to execute advanced tactics, and couldn’t understand why they weren’t winning more games.

This leads to…

6. The Illusion of Truth Effect

A person is more likely to believe a statement that seems familiar to them than one that does not. Another way of putting this is: say something often enough, and it will become the truth.

If you keep reading on forums that X is a good tactic, eventually, you start to believe it and repeat it as well. Maybe it actually is a good tactic – but the point is you haven’t critically thought about it yourself. You just assume it is true because you keep hearing it.

Of course, this ties into…

7. The Bandwagon Effect

Bandwagon effect

The tendency to believe something because many other people believe it. This is similar to the illusion of truth effect, but is contingent on you seeing a lot of other people believe or adopt the idea.

Again, we’ll call upon a Diablo example, but I’m sure similar situations arise in many other games. When the Diablo 3 ladder released a couple of months ago, a certain build emerged as the “best” for climbing the leaderboards. All the top leaderboard players were using this build – until a few weeks later, when the top spots shifted to a new build.

What happened?

The “new” build wasn’t new at all – it emerged at the same time as the other build. However, it didn’t gain traction in the community. The first leaderboard-topping build gained traction because it was used by one or more top-players to place on the leaderboard. Then, everyone sought to mimic these top players to also place on the leaderboard, so they adopted that build, assuming it was the best.

In truth, it never was the “best” build – the “new” build was always superior. But be it due to lack of luck, time, or skill, those who first conceived of the “new” build couldn’t place on the leaderboard, so the build remained in obscurity until someone used it to hit a new record on the leaderboard. Now, the bandwagon effect has switched to this build…

8. The Gambler’s Fallacy

The tendency for humans to believe that the likelihood of a future event happening is somehow based on the past. This is the kind of thinking that leads people to believe that, upon rolling a 20 on a 20-sided die, it is less likely for the next roll to also be a 20.

Probability doesn’t care what happened in the past. If I flip a coin, the odds of it coming up heads are 50/50. If I tell you I will flip two heads in a row, what are the odds of at least one of those two turning out to be heads?

There are only four possible outcomes:

■ Heads heads
■ Heads tails
■ Tails heads
■ Tails tails

In three out of those four possibilities, we get a head, so there is a 75 percent chance that in one of the next two coin flips, I’ll get a head. But if I flip the first coin, and it turns up tails, does that mean that the next flip will have a greater than 50 percent chance to be a head? No – it’s 50 percent.

Our misunderstanding of probability leads to believe silly things, such as…

9. The Hot Hand Fallacy

The belief that you’re on a “winning streak.”

In Diablo history, we saw this in the form of the “loot server” myth. As per this theory, some servers (or game instances) allegedly have higher-than-average odds at dropping loot – and you can even tell which servers will be “loot servers” based on IP addresses.

What started this myth? When people who would get a few loot drops in a row and believe they were on a hot streak. They would feel that this server must be “lucky” and would seek to stay in it as long as possible to take advantage of the apparent increased drop rates.

The reason these hot streaks exist is due to…

10. The Clustering Illusion


This is our tendency to overestimate the importance of small streaks in large samples of random data. The human mind loves to look for patterns in chaos. Given a large sample of random data, it’s natural for some of that data to randomly arrange itself into something that may look like a pattern, especially if you only look at a portion of the data set. But when you look at the entire data set as a whole, you realize that there is no pattern there.

This is also because of…

11. Insensitivity to Sample Size

Our tendency to underestimate how much variation to expect in small samples, especially when dealing with events that have a tiny chance to occur. Too often on forums do we see people reporting results from statistically insignificant data sets.

As an example of how much variation we can see in data, let’s take everyone’s favorite die: the d20. When you roll a 20-sided die, every number has a five percent chance of turning up. In other words, the “drop chance” for each of the 20 different “items” is five percent.

In a run of 100 dice throws, you would expect each number to come up five times. I rolled a virtual d20 100 times, and in my data, two numbers only came up twice, and one came up nine times – roughly half and double the expected “drop rate,” respectively.

And that’s in a sample size of 100 data points (far more than people tend to post on forums), with only 20 different possible items. Diablo has dozens – maybe even hundreds – of items, meaning we need thousands of data points to get any meaningful results.

All of these biases tie into…

12. The Illusion of Control

Our tendency to overestimate our degree of influence over external events. This is the bias that leads us to develop these superstitions. We want to feel in control of our universe, like we can exert some degree of influence over random events. We seek out (false) evidence and search for reasons to convince ourselves that we are, in fact, exerting control.

13. The Contrast Effect

Contrast Effect

Simply put: when it’s placed next to black, grey looks lighter than it is. Next to white, it looks darker than it is.

The example I’ll make here is the original Diablo 3 – the one with the infamous Auction House. Because of the contrast effect, the items that dropped for us in-game felt even worse than they actually were. I’m not saying that the drops were great, but that they looked like complete junk next to the items available in the Auction House. The top-end Auction House items were still incredibly rare, but when you have hundreds of thousands of players, these rare drops add up, and they make their way to the Auction House.

I have friends who played Diablo 3 without ever visiting the Auction House, and they never complained about bad drops. They didn’t have a yardstick against which to measure the quality of their items, so they didn’t know that their gear was laughable compared to Auction House gear. The result? They didn’t think their gear sucked.

14. Rosy Retrospection

Everyone is familiar with this one: our tendency to remember the past as being better than it really was. Time has a way of smoothing over memories, making us forget the mediocre and glorifying the good. Diablo 2 was an amazing game, but we’re remembering it better than it really was.

What cognitive biases do you often see people fall prey to? Which do you manage to avoid?


About the author