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
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?