Reproduced from the Sitka Gear Blog.
Intuition is terrible at solving problems. It’s arrogant and assumes it knows things it doesn’t. It jumps to conclusions without really accounting for nuances, and can really make a mess of things.
For example: imagine you're on a game show, and the host points to three doors. Behind one door there's a truck, and behind the others there are goats. So you pick door No. 1. The host plays it up, asks if you're sure, and then, surprise: he opens door No. 3, revealing… a goat. He asks, "What would you like to do now? Do you want to switch to door No. 2, or will you stay with door No. 1?"
What are you going to do? No, seriously, decide whether you’ll switch or stay before reading on.
If you're like most people, you gave it some thought, worked through the scenario in your head, and determined that it really doesn't make a difference, at least statistically, whether you switch or stay. There are two doors left, and one of them contains the truck, so each door has to have the same odds, 50/50.
If that was your thought process, your intuition may have just shot you in the foot. Because keeping your original pick offers only a 33% chance at the truck, while switching fully doubles your odds to 66%.
Is your intuition struggling to comprehend the nuances of this problem? That’s OK, ours was too. Here's the explanation: No matter where the truck is, your original choice had a 1 in 3 chance of being the truck and a 2 in 3 chance of being a goat. Switching doors is only bad if you initially chose the truck, which only happens 1/3 of the time. Switching doors is good if you initially chose a goat, which happens 2/3 of the time. So the probability of winning by switching is 2 in 3, double the odds of not switching.
Like we said, intuition tends to oversimplify even the simplest of problems. Not a big deal when the solution is mathematically testable and you can prove your initial conclusion wrong.
But what about more intricate and less calculable problems, like camouflage? The available tests aren’t mathematical. They’re not even objective. So each year we see a piece of cloth that slightly better resembles a particular species of plant, or maybe a slightly different arrangement of lines and blotches. Sure, some sciency-sounding language will get tacked on to describe why the new pattern is improved, but what we're looking at is just the artist's best rendition of what he or she intuitively figures will hide hunters from deer.
Of course, each year hunters kill deer while wearing the newest patterns, which proves they must work better. Or does it? The same guys did the same thing years ago in the old patterns, and our forefathers did the same thing back before there was printed fabric. There's anecdotal evidence that a certain camouflage works, but that's true of every pattern.
So when we were first trying to choose a deer pattern for Sitka Gear, we asked, "Which one works best?" We studied and probed for data, but there was nothing empirical to show that one pattern worked any better than any other. It was a crapshoot, 50/50, doesn't-matter-what-you-choose kind of scenario, all because the problem, How can I hide from deer? was being treated with a knee-jerk intuitive solution: look like your surroundings.
Intuition was shooting the entire hunting industry in the foot. And yet we needed intuition. We needed more of it. We just needed to use it differently than everybody else.
Sure, intuition is arrogant and surprisingly bad at solving problems, but if you want to find a useful problem, you need some measure of arrogance. You have to start with an intuitive assumption that the common understanding is wrong, or at least incomplete. And you need intuition to come up with new questions and new problems that didn’t exist before.
That’s how GORE™ OPTIFADE™ Concealment came to be. We worked with our associates at W.L. Gore who are, by and large, a bespectacled group of scientists and engineers. This whole bit about intuition and problems and solutions, this is just how we think, how we’re wired. And it’s almost completely backward to the way most hunting companies think.
So we set out to fix camouflage. We considered the problem, “How can I hide from deer?” and asked, “Why are we trying to hide from deer in the first place?” Sure, it sounded a little unorthodox, but it allowed our intuition some room to roam: Do we really care if the deer see us? What if they do see us? Wouldn’t it be cool if they just went back to eating acorns? What if we could get away with a little more movement? How would we have to look to make that possible? What would the deer have to see, or not see?
And that led us to a better problem, one that’s quite obvious when you hear it. But as far as we can tell, no camouflage designer had ever asked it before — at least not seriously. The problem is this: How do deer see the world, anyway?
To solve it, we had a lot to learn, which meant we had a lot of scientists to call. We talked with ophthalmologists and veterinary biologists, and a few of them seemed to think we were nuts. But one gal mentioned a handful of names, and one of those guys mentioned a handful of names, and as we expanded our search, this one name kept coming up: Dr. Jay Neitz.
So we Googled him. Dr. Neitz is an animal vision expert at the University of Washington Medical School, and he developed with his wife a gene therapy that cured colorblindness in a pair of squirrel monkeys, turning them from dichromats – animals whose eyes contain only two types of color receptors – into trichromats – animals with three color receptor types, like most humans. If he could do that, he could definitely devise a simple test to understand what deer see. So we called him. And we were in for a surprise.
“Deer vision is very different from human vision,” he said. He talked a million miles a minute using terms we’d never heard, but we caught just enough to get that he’d already studied deer vision and could easily answer our questions. We were furiously scribbling notes.
“And it’s not just deer we’re talking about,” he said, since ungulates — hoofed animals like deer, elk, goats, sheep, and pigs — all evolved with very similar eyesight. “They’re all dichromats with cones that perceive yellow and blue, meaning they’re colorblind to the red spectrum.”
It was kind of surreal hearing that for the first time. For years we’d been obsessed with deer, and now were we beginning to understand how they see the world.
A deer’s sight picture is more than twice as wide as ours, he said — 280 degrees, compared to our measly 120 — so it’s like they’re seeing in panorama. With that enormous swath of vision, deer lose a little bit of clarity, so if we see 20/20, deer see a slightly blurrier 20/40. Also, deer eyes are equipped with many more rods than ours — highly sensitive photoreceptors that can’t detect color, but can be triggered by a single photon. So deer can see much greater detail in much darker conditions than we can, albeit in black and white.
Dr. Jay Neitz at the University of Washington Medical School.
When Dr. Neitz agreed to help us develop a new pattern based on his findings, we were ecstatic. We had on board the one scientist in the world who could help us most. Our minds felt soupy from all the new information, but we were primed to take on the next logical problem: How do we build camouflage around deer vision? And what should we be trying to get deer to see?
The answers were more nuanced and complex than we ever could have imagined.
Whose help would you ask for if you were building camouflage around a different species’ vision? Our intuition said, "Soldiers," and we went with it. Militaries around the world have a vested interest in keeping their soldiers effective and alive in the field, and they probably know a thing or two about building and testing camouflage. But Who helps them with that?
We did a few web searches, and Lt. Col. Tim O'Neill, Ph.D. (U.S. Army Ret.) popped up. The guy's profile was pretty impressive. An expert in visual biophysics and human visual performance, he created the first digital camouflage pattern in 1975. He was deeply involved in the design of the most current patterns for the Canadian Department of Defense, as well as the U.S. Marines and Army.
A guy like that would probably love to talk to a few gear junkie hunters, right? Right?
Well, we called him anyway, explained what we'd learned from Dr. Neitz, and asked if he could help.
He proceeded to blow our minds with the efficiency and thoroughness. These are what he called "the basics."
"It all has to do with how the eye and the brain work together," he said. "So when humans detect a target, we are simply spotting something that doesn't belong. It may be a valid target – a true detection – or it may not be anything of interest – a false alarm."
There are two different visual systems running at the same time that make such target detection possible. These are known as the focal and the ambient systems, he said.
The ambient system, sometimes called the "where is it system," is an ancient way of seeing and evolved long before the focal system. What's amazing is that it uses a different part of the eye than the focal system, and it has a distinct anatomical pathway to the brain, called the tectopulvinar. All the images that feed in from this pathway are registered by the brain in the unconscious.
If the brain picks up something of interest from the ambient system, it will move the eye so that the object sits squarely in the receptors used by the focal system, or the “what is it?” system. This second neurological pathway then brings the image into the brain's consciousness so we can start gathering information about it.
Understanding this process means camouflage has to work in two ways: First, it has to avoid stimulating the ambient system. Essentially, that means breaking up the hunter's 3-dimensional shape, not just his silhouette, which has to be achieved with an effective macropattern. Second, if the hunter is detected, the camouflage has to prevent, or at least delay, recognition by making the hunter appear to have a completely different texture, which is where the micropattern comes into play.
Lt. Col. Timothy O'Neill, Ph.D. (U.S. Army, Retired), the grandfather of digital camouflage.
Lt. Col. O’Neill’s “basics” were fascinating, but this understanding of visual systems had only ever been applied to camouflaging items from human vision, and as we'd recently learned, deer don't see the way we do. Could it be possible that the eyes and brains of deer operate in the same way?
We got back in touch with Dr. Neitz and asked if he thought these dual visual systems might exist in deer. He said, "Absolutely,” and we breathed a sigh of relief. “These dual processes take place in very [evolutionarily] old parts of the brain, and are common to all vertebrates. In fact, a similar phenomenon occurrs with all the senses, not just vision. Sounds and smells stimulate your senses all the time, though you’re not always conscious of them. When your brain deems something to be of interest, you will either react to it unconsciously, in the case of a loud noise that causes you to jump, or your brain will shift consciousness to begin assessing the stimulus.”
So we were on. We called Lt. Col. O'Neill back and he agreed to help guide us in the development of a scientific pattern around Dr. Neitz's findings, though he said that for the actual design, we should work with Guy Cramer.
At that time, Cramer's algorithmic digital patterns concealed more than 1 million soldiers in about a dozen countries worldwide, and those numbers have since grown substantially.
We listened as he explained the process of creating an effective pattern.
“To do this well requires a deep understanding of optical systems, organic shapes, and some rather complex mathematics,” he said, explaining that he'd coded much of that knowledge into a set of progressive computer algorithms.
The process begins with the Macropattern Algorithm, which determines the most disruptive layout of light and dark colors for a particular three-dimensional shape – in this case, the human shape.
From there he applies his Symmetry Disruption and Symmetry Axis Algorithms, which directionally position the areas of light and dark to conceal the structural repetitions of arms and legs that can stimulate the ambient system.
The Movement Concealment Algorithm refines the macropattern to obscure the pivot points of the limbs and torso, which, even with slight movement, can reveal our shape and make us easier to identify.
And then he applies his Fractal Algorithm, which is pretty intricate. Fractals are geometric shapes that tend to repeat at larger or smaller scales within the same object, and nature is full of them. For example, a twig is a small version of a branch, which is a small version of a parent branch, which is a small version of a tree trunk.
When the brain analyzes a given setting, it quickly takes note of the common fractals and then ignores them, treating them as background noise. So if the pattern is not a fractal – like an artist's random array of blotches or branches, for example – or it's a fractal that doesn’t normally occur in this particular setting, it will stimulate the ambient system and alert the focal system to investigate.
Then comes the micropattern, the small digital pixels that add background noise and texture to confuse the eye's focal system. The idea is that if you're detected, you should be difficult to identify so that a deer's delayed reaction could offer you an ethical shot.
The micropattern is initially laid in with the Fractal Algorithm, but it isn't quite right until the application of two further algorithms.
The Boundary Luminance Gradient Algorithm pulls the darkest and lightest colors into a thin line to form a border between the two or three dominant color layers, which creates a stronger disruption between similar hues.
The Three Dimensional Depth Algorithm then arranges the lights and darks in a way that simulates depth. An alert deer will be looking for an even surface, like the hide of a natural predator or the contour of a jacket. But if, when they look at you they see holes and raised portions, their brain will interpret it as texture, and therefore not a threat.
When all the shaping algorithms have been applied, there's one final matter to settle: the colors. The Camouflage Color Algorithm is used to reflect the primary colors within a particular environment, or an average of two or three environments, as well as the shades required for proper simulation of shadows and reflections, all in the correct percentages. For dichromatic deer vision, this algorithm would require some major tweaking.
Guy Cramer, owner of Hyperstealth Biotechnology Corp.
It was pretty clear we had our dream team. The stage was set to develop a measurably better camouflage pattern – which is an incredibly long and intricate process. The deeper we got involved, the more we realized how little we actually knew, how little we had to contribute. It was pretty obvious that this was now the experts’ problem to solve. So we waited.
We talked and dreamed and thought about the new pattern a lot. We knew the results wouldn't look anything like what hunters were used to, and we worried that people might not use the pattern because it was too different. Of course, we also hoped it would catch on and completely disrupt the hunting industry.
Sure we were expecting it, but it still felt like a surprise when we got the call saying Dr. Jay Neitz, Lt. Col. Tim O'Neill, and Guy Cramer had produced the first-ever scientific camouflage pattern for ungulate vision. Per our wishes, they used algorithm inputs to avoid stimulating the ambient system while on ground, and optimized the micropattern for delaying recognition at engagement distances beyond 35 yards.
Shortly afterward, they developed a pattern for whitetail hunters with algorithm inputs for preventing visual stimulation in elevated engagements, for hunting from a treestand. They optimized the micropattern to delay recognition at engagement distances of 10 to 25 yards.
Sure, the patterns looked cool, but seeing all that abstract scientific thought come to life was what really got to us. Every color and pixel was placed with the sole purpose of preventing the stimulation of an ungulate’s visual processes. Minimal stimulation meant minimal detection, and as we considered the implications, we realized this was an entirely different way of thinking about camouflage. We weren’t trying to look like our surroundings, which is what intuition had always said to do. Instead, we were becoming nothing in the eyes of our prey.
Traditional hunting camouflage and its paintbrush-wielding leap of intuition just seemed obsolete, and this new thing was so radically different that we felt weird calling it by the same old name. So we decided to call it something else, something that seemed more accurate: concealment. The first pattern became known as GORE™ OPTIFADE™ Concealment Open Country, and the second as GORE™ OPTIFADE™ Concealment Elevated Forest.
Half of our goal was accomplished. We had a scientific solution to the problem of concealment. But the other half was just as important. We needed to prove the patterns’ effectiveness, and that led to an interesting problem: how can humans empirically test a pattern that's not designed for human vision?
Again, it required expert help. Dr. Neitz gave us exceedingly precise measurements of deer vision, which Guy Cramer then built into filters for photos and videos, allowing us to see the way deer see. In this deer-vision world, we could fully test our patterns, compare them against others, optimize stand placements and correct concealment weak points.