Pattern Recognition Bias

If i gave u a sequence of numbers: 2,4,6,8,10,12,.... what can you predict about the rest of the sequence?

The most obvious answer coming from our long education history would be that the given sequence is that of even numbers. This is an example of pattern recognition bias. Our left brain which is responsible for logic (nothing but pattern recognition) makes us connect the dots in such a way. Sounds familiar? Occam's Razor, the principle of parsimony, crudely states that the simplest explanation to a given problem is the most likely of all possible solutions. That is how our brain sees this sequence. We store rules that can be used to generate specific instances. A very powerful data compression principle employed by the brain. In a largely random world, we breakdown data into pockets of order and define a function to impose order on the limited sample space. Are there consequences? Yes. Since we assume a linear and logical world, some of the events appear to come out of the blue, unexpected or rare.

Let me take the earlier sequence itself: 2,4,6,8,10,12.......

Lets say we know more about the sequence: 2,4,6,8,10,12,14,16,19,20,23,28...

Until 12 what we had was the past and the present. Beyond 12 is the future. When at 16 we see 19 emerge, we are taken by surprise because the expected trend of 'even numbers' is no longer valid. Instead, the sequence turned out to be an ascending sequence. Our expectation of the unexpected is tempered by our expectation of a linear, orderly expected world.

It is possible to superimpose more than one type of pattern on an unfolding event while giving a very coherent and consistent explanation of the past and present. But there is an inherent bias based on our tendency to assume known patterns. If you noticed, I re-interpreted the sequence to be an ascending sequence from an earlier assumed even number sequence. I have made another likely error.

What if the sequence were: 2,4,6,8,10,12,14,16,19,20,23,28, 27,18....

Most of the times we try to confirm and assert a post-hoc explanation. History appears to be logical, business cycles appear to be logical, wars appear to be logical, rise and fall of civilizations appear to be logical. But if we were to ask people of those times, if they expected events of such scales, it is very unlikely that they would have seen it coming. In effect, the logic is to suit our needs. This logic however cannot be expected to hold reliably to forecast the future.

Coming back to the sequence 2,4,6,8,10,12,

The most obvious and probably the only truth that is confirmed is that the given sequence is not a descending sequence. From the start and middle, what we can reliably say about the 'full picture' is not what it is, rather what it is definitely not. We can rule out a pattern based on what we know. We can never confirm a pattern using the same premise.

“We now know a thousand ways not to build a light bulb” - Thomas Alva Edison, on his failures before finally inventing the Light Bulb