Is There A Simple(x) Solution To Optimizing Communication For “Hard Hides”?

Algorithms. What are they? Some might say they are the songs that made Einstein move (Al-go-rhythms). Others might say they are the nefarious micro minions in your phone, tablet or TV that serve you up viewing suggestions you never knew you wanted, but have to admit you do (hello every British baking show ever). Whatever algorithms are (boring answer: instructions for processing data to solve problems; cool answer: google ‘generative art’), they might be the drumbeat of the dog’s approach to problem-solving “hard” hides.

Good Dantzig Hunting

Paraphrasing from a blog post on George Dantzig (Aleteia.org):

The day that 23 year-old George Dantzig arrived to his statistics class 20 minutes late was the day that changed mathematics – and George’s life. George slid quietly into an empty seat in Professor Neyman’s class on this unremarkable day, and proceeded to write down in his notebook what he presumed to be two unremarkable homework problems displayed upon the chalkboard behind a lecturing Professor Neyman. Over the ensuing days, George worked on these two problems, finding them unusually difficult, but eventually prevailing with solutions. When George next saw Professor Neyman, he casually handed him the results of his work, and Professor Neyman stuffed what he presumed to be some unremarkable homework into his bag. Weeks later, an excited Professor Neyman found George to celebrate his work. Ignorantly, but not incompetently, George had solved two of the most famously unsolved problems in statistics! In arriving late to class, George had missed the professor’s discussion of the two problems being famously unresolvable. In George’s own words, “If I had known that the problems were not homework but were in fact two famous unsolved problems in statistics, I probably would not have thought positively, would have become discouraged, and would never have solved them.”

At the root of George Dantzig’s success is equanimity. Without knowing it, he gets to work on two of the hardest problems in his field, discovering the difficulty of the problems (for him) through the act of solving them – not by beginning with cultural preconceptions or myths of how impossibly hard the problems should be. Dogs search in this same way. They are “late to class” every time someone is describing what a problem should be. When they search, they perform an act of discovery.

The flower of Dantzig’s equanimous approach to hard problems, is elegance. Elegant solutions to complex problems are often authentic and intuitive (one of Dantzig’s solutions was thought to be so elegant it was published in a math journal). Had Dantzig been influenced (by others or by his own thoughts) to ignore his intuition and work in a way that was not authentic to him, he might never have solved the problems – and he definitely wouldn’t have solved them elegantly. Dogs do their best – most elegant – searching and problem-solving when we support their intuitive and authentic approach.

So, at the core of our exploration of hard hides are these guiding principles: meet all problems with equanimity, and, follow the intuitive & authentic path.

We’re Dantzig On The Ceiling

Imagine for a moment that you are in a wicked game of Marco Polo with Lionel Richie. You have only your sense of hearing, and spatial/touch awareness to guide you. Can you guide yourself closer and closer to the master of Motown? It could be hard, what with his ability to walk on ceilings and to mock you with the buttery angsty vocals of Hello (is it me you’re looking for). Even if it took you “All Night Long”, you’d probably catch Mr. Richie by making whatever move seemed best at the moment and repeating that process. In a sense, you’d be guided by the principles underlying the Simplex Method. (in another sense, you’d be guided by the soulful sound waves of one soft-rocking 80’s legend…)

Developed by George Dantzig in the late 1940s, the Simplex Method is considered to be one of the most important algorithms ever invented. It is still widely used to solve optimization problems. It led to the field of linear programming. It’s kind of a big deal. You can go down the mathematical rabbit hole (a simplicial cone, if you will) on your own, but a practical explanation of the method as relayed by Jim Collins will suit us right down to the ground: “You find optimal by never really knowing what optimal is. You do it by a series of iterative steps of the next best step.”

Collins goes on to explain that: for multivariate problems with endless permutations, under certain conditions, all you need to do is find the local optimum – what’s the next best step? (see the Tim Ferriss blog post with Jim Collins)

If we agree that scent work is full of multivariate problems with endless permutations, then it’s not hard to see that the optimal outcome to a scent work problem is unknown by both the dog AND the human. If we disagree, then you possibly believe a human can know the optimal outcome to most (all?) scent work problems, and knowing the optimal outcome, the human can teach the dog (and/or the handler) the steps to take to reach that optimal outcome.

To get back out on the dance floor for a minute, knowing the optimal outcome is like practicing a dance routine to a Lionel Richie song, crafting and rehearsing the specific moves for that routine with the intent of executing a flawless performance VERSUS learning many dance moves and routines, set to many different songs, with the intent of dancing to a randomly selected song – maybe even one you’ve never heard before – and possibly needing to merge your skill and intuition together to find the optimal dance expression for that song.

Lest you think we can’t talk about algorithms and dance, the University of California San Diego developed an algorithm called Dance Dance Convolution which takes songs and dance steps and creates choreographies. The robot can now create “the robot”. Domo Arigato!

Dogs are primed to create choreographies to scent songs – but they need to load up with dance steps. I’m not talking about specific patterns or protocols taught by a human to a dog. I’m talking about ensuring that the dog has a strong, energetic, reliable connection to scent from source before you toss him out on the dance floor to boogie with the vast catalogue of birch beats.

Liner notes: I’m agnostic on how, exactly, you train/cultivate your searching dog as long as you can answer affirmatively to these questions: can my dog search for the possibility of odor (search without already smelling scent)? Does my dog react energetically to the presence of scent from source? Can my dog persist at source (keep working to progress as close as possible to something his nose can’t touch)? Is my dog motivated to communicate source to me (indicate or clearly fixate when and where source can be identified)? Can my dog do these things in a variety of environments?

You’ll never answer 100% affirmative to those questions. You’re trying to get to 80%, then it’s time to dance.

The “First Best Step”

Let’s imagine a hard hide. It’s blind for the handler. Odor presentation and environmental factors (area size, topography, temps, wind, presence of other stimuli) are such that the dog will behave unexpectedly or in ways the handler has never seen before. This is a scenario where the song the dog is dancing to is one he’s never heard before – in other words, the odor picture is a doozy. The first best step for the dog is to gather some notes of odor without acting on preconceptions.

This requires complete trust from the human. It could take the form of slow, pensive, ambiguous behavior. It could take the form of wild, energetic, bombastic, sweeping behavior. The human has to go with it – follow the dog – even if these early moves don’t make sense.

Name That Genre

The first best step is also the next best step: gather more notes of odor. As more notes of odor are gathered, the odor genre may be revealed: soft rock, 90’s grunge, euro-pop, minor key, the singular stylings of Yanni? Some of these genres are much more simple, predictable and easy listening than others, like the music of a 90s boy band – or a picnic table seat hide drifting crisply along the underside of the bench seat, hooking a dog by his nostrils the way N SYNC’s “Bye, Bye, Bye” hooked everyone with ears in the early 2000s. Some genres can be quite complex, demanding, opaque, unconventional, love/hate, like the music of Joanna Newsome or Frank Zappa – or, like a hide tucked into a tiny sock left laying on the wood chips of a neighborhood playground, the odor creeping along the ground, spreading out like a thousand little funhouse mirrors distorting reality, crawling up the equipment, teasing the dog with a familiar melody of odor notes on the breeze, only to subvert the expected, twisting the dog in knots with a cyclonically cryptic trail to nowhere, until… the dog submits to the experience, letting the music of the hide take him where it must, take him to a surprisingly, yet inevitably, satisfying finish. (sometimes satisfying to the human because it’s over!)

A gathering of odor notes may need multiple “listenings”, to identify the genre. You might think you’re listening to music from the Funk genre, then change your mind and label it Urban Dance, then listen more carefully and realize it’s Acid Jazz. Your dog sometimes refines the genre of odor he’s working in, double or triple-checking the scent collecting in and around the bathroom trash can as he works out why it’s also collecting around the toilet and a little bit in the sink, behind the door, and on the tiled wall opposite the sink & toilet; finally realizing it’s an elevated, contained, diffuse, lofty source in a wall-mounted paper towel dispenser.

Just as you can never go wrong with an 80’s station on Amazon Music, some genres of odor are always welcomed by the dog… and some are universally rejected (please do not investigate Electro Swing. Just don’t). Most all genres of odor can weave their way into a dog’s consciousness (Macarena anyone?!) with a few exposures, wether he’s excited about it or not (you can test this by investigating Electro Swing. I dare you).

From Steps To Moves

As luck would have it, the iterative process of gathering more notes of odor is the dog’s intuitive simplex method. The dog checks each effort to gather notes of odor against the previous effort, considering the local optimum: do these notes of odor lead me to a more complete picture of the path to source? Step by step, the dog moves closer to a solution in a “more of this, less of that” fashion, which begins to take the form of an odor interpretive dance. The dog creates dance moves that can be combined, refined, and repeated, as he uncovers the notes, genre, and song of odor, and moves ever closer to an elegant expression of problem-solving.

Think of a song you’ve never heard before, but one you can place in the genre of country music. Before the song is over, there’s a good chance you’ll be able to intuit the remaining lyrics. There’s an even better chance you’ll be able to fit the song to the dance moves of a two-step. With some “hard” hides, if a dog is able to settle into the genre and get the melody of the odor, he can quite confidently shift to the perfect problem-solving moves and boot-scootin’ boogie his way to source. This is inarguably enjoyable to observe.

If the hide is of an unconventional, non-conforming nature, the resulting dance will probably be unconventional and nonconforming. If you prefer a night of conductor-led John Williams music to the wild excitement of early Ozzy Osbourne concerts, you may be a bit of a downer for your dog when he most needs you to let loose to the odor equivalent of Gangnam Style!

Simplex Remastered

A funny thing happens when a human commits to hitting the shuffle button on the infinite playlist of odor problems: she witnesses, in awe, the amazing abilities of her dog to not only run his own simplex method for problem solving, but to improve the method through exposure. An improved simplex method is like an ear finely tuned to as much music as possible, and a mind well-versed in the subtle details separating musical genres, and a soul always ready to receive a song and gift a dance.

A dog running his own simplex method meets all challenges with equanimity and follows an intuitive, authentic path to source. This dog is thoughtful, thorough, open to following connections that may not be relevant (this time), authentically matched to the notes of odor he gathers. This dog is a dynamic, creative, deep communicator, brave enough to dance to any song. From verse to chorus, chorus to bridge, he lets the iterative process of next best steps create the dance routine to fit the song. From odor edge, to pooling scent, from pooling scent to tendril to source, this dog brings you to the dance floor and guides you through the act of creating an elegant dance inspired by odor and environment. Like a doggy Derek Hough, your pup was born to dance to the melody of odor molecules, to dance and to grow, and to take you along on the journey. (check out Derek’s 50’s Paso Doble)

Something’s Not Twerking

There will be times when your dog samba’s with scent like he’s got 4 left feet – don’t fret! Many challenges can arise when searching the infinite playlist of possible odor problems, among them are:

• A rapidly changing environment. Think of turning the dial on an analog radio, scrambling the airwaves as you slide along FM from 88.1 to 107.9. You get snippets of classical music, the news, mariachi horns, an evangelical sermon, a Journey song. There’s no continuity or structure to this sound sandwich. A scent work search can present a similar problem. A dog might detect moments of odor clarity as he moves through the environment, not getting enough consistency from conditions to figure out what he’s smelling, and what he’s working towards. Environmental changes can be those that affect odor and/or those that affect the dog’s ability to focus on searching and problem-solving.

• A judgmental human. Ever watch a dance competition? Sometimes, the dancer is so concerned with what the judges are thinking that she can’t perform. The judges will often comment that she seemed like she was holding back, or like she was too conscious of technical details and not “feeling the dance”, or they’ll ask why a particular move was left out of the performance. The dancer will share that she is nervous, trying to please the judges, and afraid to take a risk or make a mistake. A searching dog can get judgmental vibes from a handler, too. Many dogs become aware that their handlers are not dancing with them, but judging their performance.

• A dance in the making. Not every performance is about presenting perfect problem-solving moves. Some performances are about creating those moves. The idea that there is ever a moment where a dog (and handler, for that matter) should “perform” can only be realized through controlling variables, controlling outcomes, and perfecting a path to a known outcome – even then, a highly polished search routine can go wrong. When we search with unknown outcomes and lots of variables, we have to observe the dog, and be prepared to end a searching experience without finding some – or any – of the hides; knowing that the dog may be piecing together useful problem-solving moves from multiple search experiences, planning to drop a new single – a new, successful search approach – in an upcoming search.

• A funk. If you and your dog have lost your groove for more than a few searches, you might need to step off the dance floor and assess. Are you mixing up your search exposure enough to include random variability of simple and complex search challenges? Are you maintaining the fundamentals of a confident, reliable, persistent dog, and a supportive and aware handler? Are you checking health and wellness of your dog and yourself? Do you need to take some advice from a guest choreographer (a coach or fellow handler)? Do you need to stop taking advice? Take the time to listen to your dog and yourself, the answers to a prolonged funk are usually out there with you during the search. When you’re ready, let go of expectations or goals – just get out on the dance floor – get back into searching – and let the notes of odor guide you.

May Handler Have This Dance?

You might enjoy your entire scent work experience learning to take your dog’s lead, learning to follow him through his process, honing your intuition of what move wants to happen next. But, what if your dog wants to give the lead over to you? Will you be ready to create the dance? Ready to lead your partner?

First, check with yourself to see if you’re open to walking into the club and just feeling the beat, or if you’d rather show up with your own song, so you can perform your rehearsed dance. If it’s the latter, no worries, much of scent work from classes to trials is predictable enough that you can play this way. Maybe you like the Gerry Rafferty song “Right Down The Line”. You might focus entirely on start lines, straight line movements, box to box checking, solid, reliable indication of the odor box. One pass perfection. Right Down The Line. Just consider, that when you bring your own song (or expect that the venue already has your song and intends to play it) and plan to perform your pre-rehearsed dance, it’s especially awkward when the music has changed, but you refuse to hear it.

If you want to join your dog in creating freestyle choreography for anise, birch and clove conundrums, it can be an exhilarating experience. Remind yourself to begin with lots of options as you read your dog’s dance and try to intuit the patterns in his movement. You are now running your own simplex method and you should already have a vast catalogue of behaviors that your dog expresses when dancing to the music of odor. Your dog will want you following his lead, understanding his expression – even if he appears to be sending the spotlight your way for an eight bar solo – your moves should flow from the scent story your dog has danced throughout the search. For example, if your dog ends up hanging at a found hide, tap-dancing on it, you might have to recall that earlier in the search, at the opposite end of the search area, his leash got caught on a post, yanking him off his feet while he was following a scent trail. Your dance move may be to subtly support your dog returning to the area of the accidental leash correction to let him tango with odor once more.

A dog & handler who both commit to an iterative process of next best steps, will find that they welcome the opening notes of even the hardest of hides with confidence in their combined abilities to stay calm and curious, to follow the intuitive path, and to dance their way elegantly to source.

Happy Sniffing!