Training in Practice

You now know how dogs think. You understand conditioning, markers, and reinforcement. You have the tools. But there is a gap between knowing all of that and being able to use it with a real, unpredictable dog in an unpredictable environment.

This chapter fills that gap. It starts with the single most important variable in dog training — and that variable is not your dog.

Your Half

Every concept in this chapter — threshold, reading your dog, management, progression, corrections — has something in common. They are all things you do. Not your dog. You.

The science works. Conditioning is mechanical. A marker delivered inside the 1-to-3-second window produces an association. A command followed by consistent follow-through — meaning you never give a command you cannot ensure happens — produces compliance. A dog kept below threshold can learn. A dog pushed past it cannot. None of this varies. The method does not fail on its own.

The variable is you. Whether the marker lands at the right moment or two seconds late. Whether you enforce the command you just gave or let it slide because you're tired, or in public, or embarrassed. Whether you put your dog in an environment they're ready for — or one they're not, with no leash, no control, and no way to address what happens next. Whether the rules you set yesterday are the same rules you hold today.

When training works, it's because the handler did the right thing, at the right time, consistently enough for the dog's half to follow. When it doesn't work, somewhere in that chain — timing, consistency, environment, enforcement, follow-through — the handler's half broke down. The dog responded exactly the way conditioning predicts. The inputs just weren't what you thought they were.

This is not blame. This is the most empowering idea in this entire system. If every training problem traces back to something you did or didn't do, then every fix is within your reach. You don't need a different dog. You need to change what you're doing — and do it consistently enough, for long enough, that the dog's half catches up.

And consistency means more than repeating the same commands. It means never putting the dog in a position you can't address — an off-leash dog park when you have no recall, a busy sidewalk when your dog is still reactive, letting an excited stranger approach a dog that isn't ready. It means the lines you draw are the lines you hold. It means the rules apply on tired Tuesday the same as they did on motivated Monday.

This works from two directions at once. Enforcement ensures the dog can never successfully ignore a command — a leash means "here" always results in the dog coming to you, whether voluntarily or with guidance. And engagement ensures the dog doesn't want to ignore you — because your side is where the best things happen. One without the other is incomplete. Enforcement without engagement builds a dog that complies but doesn't care. Engagement without enforcement builds a dog that loves you but knows the rules are optional. Together, they build a dog that understands commands aren't requests — and chooses to follow them anyway, because following them has always led somewhere good.

Your dog is a mirror of your handling. A dog who comes every time had a handler who made every return rewarding and never let the dog practice ignoring the command. A dog who walks calmly on leash had a handler who was more interesting than the environment, consistently, for weeks. A dog who falls apart at the park had a handler who moved too fast, managed too little, or stopped being the most rewarding thing in the picture.

There is one honest caveat. Some things live outside this loop. A dog with deep imprinting from its earliest weeks — trauma, deprivation, or lack of socialization during the critical developmental period — carries patterns that no amount of handler consistency can fully erase. Neurological or psychological conditions can create ceilings that training works around, not through. These cases are real, and they require adjusted expectations and often professional help. But they are the exception. For the vast majority of dogs and the vast majority of problems, the principle holds: your half comes first, and the dog's half follows.

Everything in the rest of this chapter is a specific instance of this idea. Threshold is about not pushing the dog past the point where learning stops — a handler decision. Reading your dog is about paying attention to what the dog is telling you — a handler skill. Management is about controlling the situation when the dog can't — a handler responsibility. Progression is about not moving faster than the dog can keep up — a handler discipline. Corrections are about reading whether the dog is confused or choosing — a handler judgment call.

The tools work. The science is settled. The only question is whether you hold up your half.

Threshold

Every dog has a point where they stop being able to learn. Below that point, the dog can hear you, respond to commands, and make choices. Above it, the dog is reacting — not thinking.

A dog over threshold cannot be trained. Their brain is flooded with adrenaline, fear, excitement, or drive, and no marker, no treat, no correction will reach them. You are not dealing with a stubborn dog. You are dealing with a brain that has gone offline.

Your job is to find the line.

What this looks like in practice

If your dog reacts to other dogs, there is a distance at which he notices the other dog but can still look at you, still sit, still respond to his name. That distance is your starting point. Train there. Build reliability there. Then close the gap — a few feet at a time, over days or weeks — until the dog can function closer and closer to the thing that used to send him over the edge.

This applies to everything, not just other dogs:

The test is simple: Can your dog respond to you right now? Can they take a treat? Can they sit when asked? If yes, you are below threshold — this is where training happens. If no — if the dog is staring, lunging, barking, frozen, or completely ignoring you — you are past threshold. Back up. Increase distance. Reduce intensity. Find where the dog can think.

If the dog cannot respond to you, you are too close, the environment is too intense, or the situation is too much. That is not a training failure. It is information about where your dog's limit is right now.

Reading Your Dog

The behavior is always visible before it happens. Always.

Dogs do not go from calm to lunging in a single frame. There is a sequence, and if you learn to see it, you can intervene before the explosion instead of reacting after it.

The early warning signs

The specific signals vary by dog, but common ones include:

These signs appear seconds before the bark, the lunge, the bolt, the bite. Seconds. That is your window.

Why this changes everything

If you catch the stiffening, you can redirect with a "leave it" or a "sit" before the dog crosses threshold. The dog is still thinking. The dog can still hear you. You can mark and reward the alternate behavior, and over time, that new response starts to replace the old one.

If you wait for the lunge, you are too late. The brain is offline. Now you are managing — holding a leash, creating distance, waiting for the dog to come down. You are not training. You are surviving the moment.

How to build this skill

Watch your dog. Constantly. In different environments. When another dog appears across the street, what happens first? When a stranger approaches, what does your dog's body do before they react? Every dog has a personal sequence — a chain of signals that always plays out in the same order before the behavior you are trying to address.

Your dog is broadcasting everything. You just need to learn the channel.

This skill takes practice. You will miss signals at first. But the more you watch, the earlier in the chain you will start to see things. Eventually, you will catch the first ear flick, the first weight shift — and you will have enough time to act.

Management vs Training

These are two different things, and confusing them is one of the most common mistakes people make.

Training is teaching a dog to make a different choice. It requires the dog to be below threshold — able to think, able to hear you, able to respond. Training changes future behavior.

Management is controlling the situation so the unwanted behavior cannot happen or is interrupted. It does not teach the dog anything. It prevents damage in the moment.

How to tell which one you are doing

The mistake

The mistake people make is trying to train during a management moment. Shouting commands at a dog in full meltdown is not training. Correcting a dog who is over threshold is not training. It is noise. It may actually make things worse by adding stress to an already flooded brain.

When the dog is over threshold: manage. Get through it. Create distance. Hold the leash. Wait for the dog to come down.

Then set up a controlled situation below threshold: train. This is where the learning happens — not in the crisis, but in the setup you create afterward.

Both are necessary. Management keeps everyone safe while you are building the skills. Training is what eventually makes the management unnecessary.

Progression

All training follows the same pattern: start easy, get reliable, then make it one step harder.

This principle is already built into everything the manual teaches — recall starts in a low-distraction environment and gradually increases distraction. But it applies to literally every behavior, every situation, every dog. It is the universal rule. And it is how you achieve generalization — a behavior that works everywhere, not just where it was first learned.

The four variables

You can make training harder by increasing any one of these:

The one rule

Change only one variable at a time.

If your dog can hold a sit for 30 seconds in your living room, do not next try a 30-second sit at a busy park. That changes distraction and intensity simultaneously. Instead, try a 30-second sit in your backyard. Then a 10-second sit at a quiet park. Build each variable independently. Each new environment where the behavior works is a step toward generalization — the point where the dog responds regardless of the setting.

What failure means

If the dog fails, you moved too fast. That is not a guess — it is almost always the explanation. Go back to the last point where the dog was succeeding and stay there longer.

This is not a setback. The dog just told you where their current limit is. That is useful information. A dog who fails at 20 feet but succeeds at 30 feet has a training plan: work at 30 feet until it is automatic, then try 25. The failure gave you the data. The frustration is optional.

Identifying the Source

This is where it gets subtle — and where getting it wrong can make things worse instead of better.

The same visible behavior can come from completely different places. A dog barking at another dog on a walk looks like one problem. But the cause behind the barking determines the fix, and applying the wrong fix makes the problem harder to solve.

Three dogs barking — three different causes

Fear-based barking The dog is scared and trying to create distance. The bark means "go away."

Excitement-based barking The dog is overstimulated and wants to get to the other dog. The bark means "let me go play."

Frustration-based barking The dog has learned it cannot get to the other dog and is redirecting that frustration into barking. This is especially common in dogs who are restrained on leash.

When you are not sure

Default to creating distance and observing. What does the dog do when the trigger moves away?

The reaction to the trigger leaving tells you more than the reaction to the trigger appearing.

When and How to Correct

A correction is information: "that was not what I wanted." But the same correction applied to different dogs in different states produces completely different results. This is why you needed everything above before this section makes sense.

Correction works when three things are true

  1. The dog understands the command. You have trained it to reliability in easier settings. The dog has done this before, successfully, many times.
  2. The dog is below threshold. The dog is thinking, not reacting. The dog is capable of making a choice.
  3. The dog is choosing a competing reward over your command. Not confused. Not afraid. Just making a different choice — the squirrel is more interesting than your treat, and the dog knows what you want but is opting for the squirrel anyway.

When all three are true, a calm correction followed by an opportunity to do the right thing and earn a reward creates clarity. The dog learns: that choice has a consequence, but this choice gets me what I want.

Correction makes things worse when any of these are true

The question to ask before every correction

Does this dog know what I want, and is this dog choosing not to do it?

If both answers are not clearly yes, do not correct. Redirect, lure, or manage instead. A correction without understanding is just confusion wearing the mask of discipline.

Key Takeaways

← Previous: The Tools | Next: Putting It All Together →