A visual illusion

If it is this easy to be mistaken about simple things it must be much easier to be mistaken about difficult things. From Geekolinks.


There is a bright band on the top of the lower box, so they are not really the same color.

Yeah, that's what I was thinking too.

Is this a subliminal message to the mistaken Paul Krugman to read David R Henderson?


Try cupping your hands so that both the background colours are excluded. Then the two boxes still look different colours. As Thomas says, it's probably the effect of the bright band. So ... FAIL!!

All that means is that the background is not really the issue. Without your finger, you brain is performing a 'shape from shading' operation and assigning 'actual' colors based on the assumed orientation and lighting of the two objects. When you put your finger across, you obscure the shading cues and then perceive the upper and lower objects as 2D (or at least as trapezoidal rather than slightly angled) and then you perceive the colors as the same.

But as for the moral of the story? The 'different colors' interpretation is actually the right call. When you're perceiving the world, what you really want to know is what colors the *objects* you're seeing are, not what are the actual RGB values of the various patches of visual field (unless, perhaps, you're an artist and are tying to mix paint).

Tried taking away the background and shape cues:


The illusion persists but not so strong (at least to my eyes).

Maybe banding, shape and background all contribute to it.

Further to test if background contributes to the illusion at all:


To my eyes it does; I'm still fooled. Like so many other explanations, it seems a little bit of everything.

Yep. If you "see" the RGB values you're at an evolutionary disadvantage to things that interpolate the true color based on the cues.

Probably something people working on visual processing systems like self-driving cars have to address.

I don't think the distinction between simple and difficult things is apt. It's about particular types of perception and reasoning, particular perceptual and cognitive situations where the brain's methods fall down.

This is, by the way, one of the stronger nails in the coffin of rational expectations theory. Humans make systematically irrational mistakes. As a result markets don't perform as if they consisted of rational individuals, which might be the case if all those human errors were random.

"This is, by the way, one of the stronger nails in the coffin of rational expectations theory."

What is the alternative? The government figuring out our biases and only using them for good?

If my bias is full employment, and it is articulated across all periods, then if the government figures out that bias and prevents me from going austere when I shouldn't, I would buy that.

Maybe England should have.

But you are irrational in all your conceptions, so the government should just ignore you.

During a recession, people do what is rational for them individually, but presume that it is irrational for their fiscal agent, the government, to do something opposite from them.

Who is irrational if you know this?

What makes you think you know this?

"I don’t think the distinction between simple and difficult things is apt"

Actually, it's the comparison that's not apt.
I'm not sure it's even possible to make this kind of mistake about complex things, because understanding them requires a different set of mental tools.

I agree that his connection to rational expectations is inapt, but the propensity of perceptive errors increases with complexity.

Arguments against rational economic man included cognitive and emotional biases. I've never seen someone attempt to include visual biases as a basis for rejecting REM, unless he was being snarky.

Then you want to read Schiller's Irrational Exuberance and the critique of RET.

I've read IX. Was there a chapter on optical illusions that I missed?

I believe in short term irrational behavior, or at least bounded rationality. I just don't see any connection of this illusion to REM. I'm familiar with emotional and cognitive biases, rational ignorance, and fat finger mistakes. Where does this one fall?

Willi, I'll look for it, but meanwhile here is a paper he wrote critiquing RET from a behaviouralist perspective:

Abstract: This paper surveys critically the literature on rational expectations and the dynamic structure of macroeconomic models. The theoretical framework common to this literature is set forth for the reader unfamiliar with it. As this is done, problems are brought out which are usually ignored in existing literature. The topics discussed are: (1) rational expectations and the natural rate hypothesis. (2) optimal linear forecasts and their applications, (3) the general linear rational expectations model and its use in econometric policy evaluation, and (4) information and convergence problems in rational expectations models.

Link here: http://www.sciencedirect.com/science/article/pii/0304393278900326

Willi, I have not found my copy to check my memory, perhaps confusing it with Kahnemann.

Or, Perhaps the Animal Spirits in me caused a Money Illusion

I must be defective. They both look like the same color to me (sans white band) without doing anything. I have seen other illusions that worked, but to my eyes this one is lame.

How do the boxes appear to other people anyway? To my eyes they both just look like grey boxes. Is there a difference between the shades of grey or something that I am missing?

Yes, I think the perception of different shades of grey is the point. If you focus intensely, you don't have to manually block out the background or shading.

You must have a very acute focus of mind if you are not immediately tricked.

........or insensitive visual receptors?

Inability to be fooled by certain visual illusions is a sign of mental illness. You should get yourself checked.
Try the hollow mask illusion, if you can see through that then you have scitzophrenia.

That is funny. I did try the hollow mask illusion at this link... http://www.wired.com/wiredscience/2009/04/schizoillusion/ and at first I could not see the illusion. I was watching the video and trying to figure out what the narrator was saying. Towards the end of the video I suddenly "saw it" and now I can't stop seeing it even though I have re-watched the u-tube clips. Kinda of aggravating really. I try by force of will to make myself not see the illusion like I could at first but no luck.

The strange thing is that when I first came back for seeing the wired link I could see the bottom square as "white" and the top square as dark grey. But that illusion did not stick around for long. I now see two dark grey squares again.

They actually look quite different to me. When I cover the band in the middle the bottom box gets progressively darker for a second or two until they match. Remove the finger and it seems to instantly get lighter though. I'm somewhat near sighted, monitor a bit above my head.

They look the same to me as well.

The problem I have with so many "optical illusions" is that they are contrived. In the sense that they are derived from actual neuro-sensory processes, I accept them as examples. It is real illusions or false perceptions occurring naturally that interest (or concern) me most.

The brain anticipates motion from visual information. This is extremely important for things like driving, where you must respond to hundreds of moving objects, judging distance, speed, bearing and heading.

It becomes concerning in eyewitness testimony where a person will swear on a stack of Bibles that they "saw" something, and video cameras show that they saw no such thing. And I'm not talking about split second or centimeter differences that a baseball umpire would be expected to judge, but large, slow, coarse chains of events. I've seen enough of these cases to know the witness actually believes what they are saying, and their disbelief when seeing the video is unsettling. Eyewitness testimony is very unreliable.

All still life, portrait, and landscape paintings would be considered "optical illusions" by the standard set above. The artist tricks the human mind into seeing depth, color, shading and light that do not exist.

If you want to see (and hear) a real interesting neuro-sensory illusion, check out the McGurk Effect. It tricks you even when you know about it. My question is whether there are more examples than the one provided. Maybe that's why there are so many misheard song lyrics. :)

Maybe you want to think of this as a stimulus problem, and not a cones and receptors problem, when discussing limiting this simply to visual stimuli. Stimuli of all sorts cause you to react in ways you didn't anticipate.

Here's an example from the grocery store: List price $2, and a 30 cent coupon; you buy because it is on sale; next week, list price $2, and you buy with a 20 cent coupon; next week list price $2.10 and you buy with a 30 cent coupon.

Did you notice that the actual price, after coupons, has been increasing every week?

I understand and accept what your saying. I'm a wealth management advisor and see first hand people's cognitive and emotional biases.

I understand the psychological impact of 99 pricing. It happened to me today when pricing exercise equipment.

I accept that someone might be caught off guard with your coupon example, but coupons are a more interesting case of price discrimination rather than deception. We know that low cost goods are price inelastic, and we observe price points where elasticity is different for a price rise than a price drop.

I'm still eagerly awaiting an example of how an optical illusion eroded faith in rationality. A modified version of rationality is making an optimal decision with all available information and how it is perceived by you. It's only irrational as viewed through the objective function of another person. Irrationality is suboptimal behavior based on your OWN objective function.

Loss aversion is extremely rational to me. Myopic consumption and lack of self control is perfectly rational when we believe the person has a hyperbolic discount rate caused by lack of faith in the control of their future. Eat, drink and be merry for tomorrow we shall die.

You're sort of making the argument that God doesn't exist because rabbits eat their own shit, and God would never do something so stupidly inefficient.

REM is a good place to start under the right conditions. In an irrational world, where do you start? How do you come to any decision? If it's not rational, it cannot be modeled - by definition. We have some proof that models work reasonably well under many circumstances, and when they don't, we could hardly do better.

Re: "I’m still eagerly awaiting an example of how an optical illusion eroded faith in rationality." What we are talking about is not cones and eye receptors, but short cuts the brain or gut uses to leap to conclusions. Optical stimuli in that respect are no different than the pricing examples I gave, or the "Money Illusion" Schiller refers to.

As to "In an irrational world, where do you start? How do you come to any decision? If it’s not rational, it cannot be modeled ..."--I guess is where it comes to is Odysseus asking that the other sailors tie him to the masthead. As a financial advisor, I am sure you tell clients to tie themselves to the masthead and weather out the storm, but some don't while others do. Sometimes to make money all you have to do is avoid making mistakes that others make. You might like reading: Why Smart People Make Big Money Mistakes, a book written by Belsky and Gilovich. They tried to model some behaviour--the mistakes people make and why--from a psych perspective and behavioural economics perspective.

There is one optical illusion that is not contrived. Ever been disappointed with the photos that you took when the lights poor?

Here is a more powerful example of the same illusion:


My color picker displays different RGB values for various spots in both squares. values between 125-128. So its not really just one color.

Good trick.

Most people would ve be surprised what you actually "see" without all the visual processing between your eye and your cognitive centers. At some point Nature made a trade-off between a few factors -- accurate perception (probably expensive to evolve), visual processing time (cheap and flexible), and the ability to "predict" what is coming next and show to your brain instead of what your eye actually sees (a function of the visual processing). So we process images into something a bit different than what our eyes "see."

This color illusion is a lot more impressive:

I remember when I was in Amos Tversky's class in grad school in 1990 he made a similar point.

He put up a set of standard optical illusions. Then he said something like, "Psychologists who study visual perception understand pretty precisely why the visual processing in our brains produce these illusions. Nevertheless, the illusions still work on them. I understand pretty precisely why the cognitive processing in my brain produces systematic biases. Nevertheless, I am still subject to these biases."

This conclusion always stuck with me. I think the right response to learning about cognitive biases is not to conclude that you can now avoid them. It is to be skeptical of your own conclusions. Which is why you pull out your spreadsheet and calculate when it matters. Just like you pull out your color meter.

The reverse actually happens though. When people learn about biases they become more sure of their own conclusions and begin to doubt everyone else more.

There was a study that showed that learning about biases and about logical fallacies made students more sure of their own beliefs and less willing to accept anyone else's beliefs. It also seems to blind people to their own biases when they know more about biases.

I really am not impressed by this recent trend, associated with the glorious name of Daniel Kahneman, to revive hoary optical illusions popular 50 years ago in Ripley's Believe or Not and other conjuror's and confidence man's tricks as proof that humans are not 100% accurate and rational.

Well, duh. Did anybody other than economists ever believe such a thing?

Everyone knows that other people are not rational.

I remember that stupid had drawn picture of a woman who could be perceived as young and beautiful or old and ugly, depending on how you viewed the picture, and it "helped" your choice if it was framed by the observer.

I have no doubt that framing has an effect and is a bona fide cognitive bias, bit I object to illustrations of this based on contrived circumstances. In other words, the picture was designed to trick you and the motive of the "seller" wasn't immediately apparent. There are few natural events that elicit a framing bias, and when politicians and salesmen frame things, it is blatantly obvious.

"Did anybody other than economists ever believe such a thing?"

For one, people who have been falling for these confidence man's tricks since as far back as we can document.

Moreover, the point is that visual and cognitive illusions cannot be defeated simply by knowledge of the fact that these illusions exist. Some people who fall for scams or psychological techniques like good-cop-bad-cop sometimes know the techniques exist and yet are still fooled by them.

If it is this easy to be mistaken about simple things it must be much easier to be mistaken about difficult things.

A non sequitur. Optical illusions work because the brain uses shortcuts to process visual information. They're not fundamentally different from a moire effect when wearing stripes on TV.

Not just short cuts. They are fundamental to our ability to understand the world. People who are immune to those illusions usually have horrible mental illnesses.

There are many examples of higher order illusions which are not merely optical. For example the classic experiment where the subject watches a group of people pass a basketball around, and fail to see the person in the gorilla suit walking through them.

Or more harmfully, the witnesses who saw a crime and "know" that the perpetrator was black.

Does this explain why tax cuts seem to kill jobs while tax hikes create jobs?

Were more jobs created after the many Bush and Obama tax cuts than after the Clinton tax hikes in 1993, but the illusion of Clinton being president in the 20th century and the tax cuts were in the 21st century?

Reagan was an illusion, too?

The illusion was that Clinton kept spending under control in the 1990s, in that he accomplished it while striving to do the opposite.

They are exactly the same color...tested with an on-screen color tester and both squares have an RGB color value of126, 126, 126.

Comments for this post are closed