Flow, Challenge, and the Dog in Front of Me

When I was doing web development, I'd sit down to fix one bug and look up to find the room had gone dark. My coffee would be cold and untouched, and I'd have skipped lunch without noticing. One problem led to another and hours just disappeared. The work wasn't easy. The best days were when I knew enough to make progress but didn't know the answer yet, so I'd read something, try an idea, break something, figure out why it broke, and try again. It took all of my attention, but it always felt like something I could solve.

Music did the same thing to me before that. I'd sit down to work on something for a little while and lose the whole afternoon messing with details nobody else would notice, or chasing a sound I could hear in my head but couldn't figure out how to make yet. I didn't have a name for it. I just knew it happened sometimes, and when it did I stopped thinking about the clock or whether I was doing a good job.

About ten years ago I read about the concept of flow, and it described that experience pretty much exactly. The part that stuck with me wasn't losing track of time, though. It was the relationship between skill and challenge. If something is way below what you can do, you get bored. If it's way beyond what you can do, you get stressed out or overwhelmed. In between there's a zone where you know enough to keep going but you still have to work for it.

Seeing it from the outside

Not long after that I was between jobs and working at an Amazon fulfillment center. A lot of the work was the same motion hundreds or thousands of times a shift. To get through the day I started making little games out of it. Could I find a better rhythm, cut out a wasted step, or get faster without making mistakes?

After a while I started paying attention to the people around me too. Some of them had completely checked out. They knew the job, it wasn't asking anything else of them, and every hour dragged. Others always looked overwhelmed, like they were behind and worried about their numbers. And some had found a rhythm. I had no way of knowing what any of them were actually feeling. All I could see was how they moved and worked. But it made the idea a lot more concrete for me. It was the same job for everybody. What was different was how each person matched up with it.

Ten years later

I didn't expect to be thinking about any of this again ten years later on the other end of a leash.

When I started taking training seriously with Gunner, most of my attention went to mechanics like timing, markers, reinforcement, luring, shaping, criteria, and proofing. There's a ton to learn and I'm still learning it. But the more I train, the more I keep coming back to a question that feels familiar: what is this asking of him compared to what he can actually do right now?

Some sessions he's completely plugged in. He's not just waiting for food, he's offering behavior, moving with me, and coming back for more. Other sessions feel flat, and I get the behavior but I'm dragging him through it. And sometimes I ask for something I thought he knew and it all falls apart. My first instinct is to troubleshoot the behavior. Does he understand the cue? Is the reward good enough? Those are fair questions, but I've started asking another one next to them: how hard is the problem I'm actually giving him? A sit in the kitchen and a sit at the park while another dog walks by are technically the same behavior, but they're not the same problem.

What I can actually see

When I was coding I knew I was in the zone because I could feel it. Gunner can't tell me any of that. He can't tell me he's bored, confident, worried, or locked in, and I don't think I can honestly say whether a dog experiences anything like flow at all.

What I can see is his behavior. How fast he responds, how hard he drives into a rep, whether he's scanning around or sniffing the ground, whether he pushes through a mistake or quits, whether he takes food and tug the way he normally does, and whether he comes back for another rep on his own. None of those mean one thing by themselves. A missed rep doesn't automatically mean it's too hard, and sniffing doesn't automatically mean he's bored. But patterns over a session tell me a lot. I've started thinking of his behavior as feedback on my training, not just the thing I'm training. So I'm not using flow to guess what he's feeling. I'm using it as a reminder to watch what he's doing.

Too easy

If I drill something he already knows, like a sit, over and over, it can look great on paper. He's getting it right almost every time. But the quality slides. He gets slower, looks around between reps, takes the food like he's doing me a favor, and eventually the grass is more interesting than I am.

My first thought is usually that I need a better reward, and sometimes that's true. But the better question is whether the problem is still interesting to him. A high success rate isn't the same thing as engagement. The fix isn't always to make it harder, either. Sometimes it's a shorter session, mixing up when and how the reward shows up, or adding one small new piece so there's something to figure out again.

Too hard

The morning I took Gunner to the park to work on heeling and recall is a good example of the other side. I went early to avoid people, but after the rain the grass was full of crows digging for worms, plus bunnies and squirrels. For me it was a quiet park. For him it was the most distracting place in town, and nothing I had in my pocket could compete with it. We ended up working on him disengaging and turning back to me on the walk home, which was a lot harder than I expected.

I could look at that and say he failed. What really happened is I changed the exercise way faster than I'd prepared him for. When a session is too hard the signs look different from boredom. He hesitates or gets frantic, makes the same mistake over and over, can't stop scanning, won't take food like normal, or just tries to leave.

With a Malinois there's another wrinkle. Frantic doesn't always mean too hard. Sometimes it's arousal. If I bring out the tug and he's so wound up he can't think, the exercise might be simple and he still can't do it. That needs a different fix, like bringing him down, changing the reward, or building the behavior at a lower level of excitement first.

The middle

What I'm actually looking for isn't a perfect session with no mistakes. I want him to understand the game, have success be doable but not automatic, and have a mistake lead to another try instead of shutting down. He stays on the problem, he can take information from me, the reward still matters, and when a rep ends he wants the next one. I think the goal is keeping the struggle solvable, not getting rid of it. That's what the best days behind the computer felt like. They were hard in a way I could work with.

Working dogs

Working dogs make this easier to see than a puppy in the backyard. A herding dog is solving a problem that changes every second as the stock moves and the pressure shifts. A detection dog knows the game but not the answer, and has to keep searching when nothing has paid off yet. Even two dogs playing well are constantly adjusting to each other as they chase, back off, and switch roles. I can't say any of those dogs are in flow, but I recognize the same things that made it possible for me: enough skill, a real challenge, quick feedback, and a problem worth working on.

That doesn't mean I should make everything harder. Some challenges just matter more to certain dogs. A Border Collie might find moving livestock compelling on its own, and a retriever might love carrying things. For a Malinois it's chasing, gripping, searching, and moving with the handler. So my job isn't just to raise the difficulty. It's to find challenges that actually mean something to Gunner.

The dog in front of me

There's a part of the flow idea I missed the first time around. The sweet spot isn't fixed. As your skill grows, the challenge has to grow with it or you end up bored again. Training works the same way, just faster and less predictably. What was the right challenge yesterday might be too easy today, or too hard somewhere new, on a hot day, or after he's already had a big session. My plan tells me where I think we're headed, but Gunner shows me where we actually are, and the skill I'm trying to build is noticing that fast enough to adjust.

When I started, my main question was how to teach the behavior. That's still the question, but now I'm also asking what the session is like for him while I teach it. Is it clear? Is it too hard? How confident is he, how amped up is he, and does he want another try? That's not me throwing out learning theory. It's me using it while paying more attention to the dog.

I first found this in music and code, then watched it from the outside at a warehouse. Now I'm the one setting up the challenge. I can't know if Gunner will ever experience anything like what I used to feel behind a computer, but I can keep watching and keep adjusting until the work is hard enough to be interesting and easy enough that he wants to go again. I think that's a big part of training the dog in front of me.