What do I do right after correcting my dog?

Finn

Finn is the case the concept is named from, and his sessions contain both the rule and the mechanism. He is a major flight risk whose obedience collapses in particular contexts, and the trainer's plan is built around the recovery rather than around the repetitions: an hour-plus walk chosen so that one pressured sit can be absorbed before anything else is asked. In the recall session the diagnosis is turned on the handler's own timing - he went back too fast, so the two attempts failed to stay distinct in the dog's head and became one prolonged stretch of negative reinforcement, and the lesson Finn would have taken is that coming back prolongs the stress instead of ending it. The recovery here is four or five minutes of free sniffing, and the movement is doing work of its own: the trainer says a jogging pace opens a dog up, which is the same claim he makes about people.

Louie

The consequence the trainer draws is fewer and better: four or five minutes, or the one down he does all session, because he wants that repetition to be worth something. His indoor sessions use the pool and jumping up for petting as the return to baseline rather than as a treat, and the trainer names it as functional. His file is also where a named handler appears - Christian, credited with having already run a couple of prediction errors on Louie's reactivity - and where the deliberate decision is made to let the reactivity surface rather than manage it away, so that there is something to learn from.

Sage

Sage is the anxious, freezing, hypervigilant type, and her file is what the rule costs when it is followed honestly. The gaps are ten to fifteen minutes indoors, where the tracked emotional-state data showed she was worse than outside; four place repetitions across an hour-long video; a go-home lesson where the gaps are snuggling on the couch until she is totally back to baseline. The trainer states the trade plainly - fewer repetitions in a good emotional state beat pushing through - and says out loud what it costs him: it takes more of his day, and sometimes that is the only option. His analogy for getting it wrong is the one an owner will remember: being handed a hard maths problem and told to solve it, and solve it, and solve it.

Lola

Lola is the nervous dog who freezes rather than flees, and her file supplies the handler's half of the concept. The correction is only half the repetition; the other half is the manner in the seconds afterwards - make it fun at once, with no residual I-hate-you energy, because pressure can shut her down fast. The named bad outcome at the end of getting that wrong is the velcro dog, which the trainer describes unprompted in the family's living room and defines the business against. Her file also makes the dog's own recovery the outcome metric: what you are looking at is the recovery, how long it takes and where she goes - flight, or to a person and then back to normal - and the second is what the program was for.

Stormie

Stormie is where the ten seconds after a correction are audible. She is soft, she is dramatic about the down, and she goes flat after refusing - and what the handler does is hold the demand and the reassurance in the same breath rather than in sequence: it's okay, but I need you to do the command; no, it's just the rules, I'm sorry; I'm not going to hurt you. When a run of repetitions fails he stops asking, cues a let's go, moves her and rubs her until the tail is wagging, and reads the tail to know when to go again. The fullest statement is made to the dog alone with no client watching, with a schoolchild-and-equation illustration attached - and in the same session he says, unprompted, that he hates hitting the button every time.

Zeus

Zeus's aggression work is the honest test of the concept, because the correction there is hard and deliberate. What his file adds is the distinction between a stopped behaviour and a settled dog: minutes afterwards the trainer says, while the footage runs, that Zeus is still a little ramped up and is sniffing compulsively everywhere the other dog has been. He does not present the calm that follows as the whole result. The structure of the repetition is also its own kind of return: the correction is followed immediately by release and praise and then by an ordinary walk down the field with both dogs loose on their lines, and what the handler reads next is what the dog chooses - his attention has moved from the other dog to him.

Roxy

The material in Roxy's file is mostly Ramsey's, and it separates the startle from the recovery. The jump at a sudden noise is involuntary and is not the problem; what matters is what the dog does thirty seconds later, so the instruction is to wait about forty-five seconds and watch rather than walking him away. Reassurance is ruled out entirely here - it's okay, buddy tells him it was a big deal, and the handler's unbothered calm is the information. For an object rather than a noise the intervention is the register: walk him over and talk to him like a peer, dude, look at this awesome box, and if it keeps mattering make the object part of a game. All of it starts from reasoning from the dog's side first: think dog thoughts.

Kodak

Kodak, a flight risk who did not trust his handler, is where the dosing gets a number. The separate walking excerpt watch concerns Rocko. It offers a comparison: the trainer says the street keeps creating requests and deliberately gives Rocko several minutes without more work. That comparison is not another event in Kodak’s history.

Bullet

Bullet arrived green and his heel work is where the reset is given as much narrative weight as the repetition. The description is generous and concrete - wander, follow me around, get some pets, listen to me yap at the camera, sniff everything on a walk - and only after that, with the dog completely back to baseline, is the criterion raised. The reason is stated as a rule rather than a feeling: do not keep pushing before letting them reset. His file also explains why the heel in particular needs it - sit and down are resting positions, whereas heel holds the dog's attention the whole time, so it is a duration exercise with a cost. And in a third session the handler grades his own dosing against himself and says the last repetition should not have happened.

Bo

The heel starts costing the command itself, the trainer says out loud that he has probably poisoned it a little and that his criteria were too high, and then, before any further work, he calls the dog in for affection and names what he is doing: reconciliation. The re-entry is at a lowered bar, and the behaviour appears. His crate-aggression session has the same shape with different vocabulary: the technique is called a prediction error, and between repetitions Bo is explicitly let back to normal rather than worked continuously - and the growl is marked with a disappointed tone rather than an authoritative correction.

Dolce

Dolce is the protection dog whose confidence ran behind her obedience, and her file supplies the theory and the practical reset in different sessions. The practical version is a handler asking, mid-session, for a three-minute break so she can come down rather than accumulate pressure; and a coaching instruction to take the pressure off temporarily, rebuild her mood with the tug until she is having a good time again, and only then re-apply.

Dutchess

Dutchess is curious rather than resistant, and her lesson gives the reset in its simplest owner-facing form: when the command is not coming, stop asking, walk around for a minute, and come back to it - do not push it over and over. The reason offered is mental fatigue, and the claim attached is that a fatigued dog simply taps out, in anything but the very high-drive breeds. Her owners' lesson is also where the underlying theory is taught directly: prediction error and classical conditioning as the two halves of learning, with the shake-off named as the visible sign that a dog has processed something.

King

King's file turns the concept into a diagnostic. A senior trainer coaching a junior one gives two questions to ask whenever something is not working - what is the dog's emotional state, and how clear was the criterion - and then the default that follows from them: never reach for more pressure, reach for play. He tests the logic on the junior handler directly, pointing out that hammering him at that moment would not have helped him learn. In the store session the register is different but the shape is the same: repeated trips back out into the open air between passes at the thing the dog is afraid of, and an honest closing line about what the session cost the dog and what it was for.

Luna

Luna's sessions show the break used as a tool rather than described as one. Free sniffing is inserted between blocks of hard work, twice, after the longest fights; in the night session the plan is to stop, let her sniff and open up, sit with her and do nothing for a long stretch, and only then reintroduce a small distraction deliberately. The same session carries the reason the collar number is not a property of the dog: with almost no distraction present, the level comes down a lot. Her material is thinner on narration than the others here - the reset is visible in what the handler does rather than stated.

Egg

On a long woodland walk the trainer insists that every release between heel reps be 'real': the dog goes back to baseline and gets lots of freedom, not 'okay, next rep, next rep'. He states the cost of skipping it: learning is stunted.

Samson

On a state-park trail the spacing is given in minutes. Release to freedom immediately after the rep, then two to six minutes before the next ask, apart from a casual 'let's go', so Samson can return to baseline and learn from what just happened. Later the trainer notes Samson is getting happier each time he comes off the 'okay'.

Ace (Chihuahua-terrier mix)

The negative case, stated outright. After a correction the trainer says was set a little high, Ace is released, succeeds straight away, and is out exploring. The trainer says that is why no harm is done: what produces a shut-down, suppressed dog is following a high correction with restriction, 'force you to heal for 10 minutes'.

Madam

A senior trainer coaches a junior handler through a down taught with pressure, and teaches the reset by asking questions: what does her body say after those reps (suppressed), how long should you wait (longer, until she's super happy again), and is she back to baseline yet, so is it safe to do another rep.

Shadow (German Shepherd puppy)

At an owner lesson Connor shows the puppy-sized version: three or four reps, then back off to decompress. When her tail drops during a sit he stops drilling and plays until the tail and ears come back, because once the dog's mental state drops, so does the quality of the training.

Callie

On a state-park hike with a sensitive dog, runs between work sets lift her mood, and short breaks let her come back to baseline.

Penny

For a dog who won't play yet, or who has gone into appeasement after pressure, running with the handler is both the reward and the repair.

Sofia

Taught and narrated in Spanish. After pressure on the down the trainer runs with her, explaining that when a dog is a bit low or suppressed from being taught, running makes her feel better, and that exercise and exploration relieve stress. Later he withdraws his attention while she sniffs so she can return 'para su baseline a neutral'.

Phoebe (the Rottweiler)

In downtown Angier the trainer gives a harder correction, then sends her off to run and be free, and warns against the alternative: going straight back into heel drills makes a sad-looking dog.

Wiley

Wiley goes after a runner on a path. The trainer says that 'cannot slide' and corrects it unambiguously, then does 'a makeup' straight away: sniffing, ball, praise, back to normal life.

Chunk

Pressure ends the behaviour and is followed at once by opening the dog back up ('I don't wanna repress'). There's a stated limit even when the pressure is low: after Chunk bolts from a barely-there correction, the trainer resets rather than escalating, because 'he could genuinely just traumatize himself out of the experience'.

Milo

The template comes from the dogs themselves. Milo corrected his housemate Josie for playing too rough: stop, a moment of stillness, decompress, then play again with the relationship intact. The trainer's human version: stop the dog, have him give up and freeze (not a pin), then be warm straight away, say okay, and make it clear he isn't still mad.

Stormy

At a private owner lesson: 'slow is fast', let her return to normal before asking for more. Freedom is 'a really underrated reward'. A dog pushed until she tries nothing is 'really hard to teach'.

Bandit

On a rainy proofing walk in a park the trainer works one turn in heel several times in a row: the first time Bandit lags hard, the second he tries going round the other side, the third he gets it but apprehensively. The success is paid with a big release of freedom and praise, and the trainer then says out loud that what comes next is downtime so Bandit can return to baseline - and gives the run of attempts as the reason.

Beo

On a park proofing walk the trainer counts the gaps as the pay. He tells the camera that the session has had plenty of positive reinforcement even though there have been no treats and hardly any toys: what there has been is freedom, praise, some petting, and 'lots of time in between reps to return to baseline' - a neutral place where Beo can reset, re-engage and feel good again. Here the return to baseline is part of the reward structure rather than a pause from it.

Kane

On a park walk late in his program the trainer states the rule as a property of the whole program rather than of one session: repetitions are sparse and organic, with neutral, free-wandering time between them, so the dog returns to baseline and can learn from the next one. He spells out the counterfactual in the dog's terms: if the pressure 'comes right back' in the next repetition, the dog learns that as soon as you ask for a command it is going to be bad - like a drill instructor shouting at a child who is being taught maths, the same picture Sage's trainer uses.

Reaper

Reaper's file shows the repair half of the rule after a hard correction in behaviour-modification work. On a walk set up as a choice point, the trainer says beforehand that if Reaper does anything 'dumb' he will punish it harshly, because it would cause major problems; Reaper fixates on a dog across a field and sprints at it, the correction lands, and within seconds the trainer is walking on with him, praising him and saying he wants him to feel happy again - 'we're buddies'.

Stormi

At Stormi's go-home lesson the rule is handed to the owner as an instruction. After a recall reinforcement leaves Stormi 'a little bit worked up', the trainer tells the owner to give her about thirty seconds and just be happy with her before asking again, and explains it in terms of nerves: a nervous dog who thinks she is in trouble and sees no resolution will lock up, which he reads from her tail. Indoors you wait for her to settle; outdoors it is easier, because running with a nervous dog, he says, releases 'feel good chemicals'.

Akira

Day two: the trainer leaves long gaps between repetitions so she can fully recover and feel happy again, and only then runs the second, third and fourth trials, which is where he says the learning happens.

Anna

After a big reaction to a small e-collar input, the handler turns the level down and lets her return to neutral before going on; he reads it as the collar being new to her.

Astro

In the kennel approaches the trainer explains that he is only partly letting Astro return to baseline within the session by not repeating the approach over and over; the real return to baseline is when he leaves and comes back for a second trial, which is where the learning will show.

Baxter

At a private lesson the owner is told that pressure is a little stressful for the dog just as for a person, so it cannot run for a whole walk; give time in between for him to do his own thing and reset to a neutral baseline.

Bun

Twice in the down sessions the trainer stops when Bun starts to look at him oddly under pressure, gives him a run and some ball to get goodwill back, and only then starts again. After a hard heel rep beside a barking dog he gives a sniffing break to reset to baseline before the next rep.

Drago

In a public heel Short the handler keeps the leash loose, lets Drago walk around and do whatever he likes, and runs the next heel rep only when he judges the dog to be in a good emotional place.

Fiona

After a string of down corrections while the e-collar was still new to her, the handler stops and lets her reset; in another session he lets her go back to neutral after strengthening the down, before moving on to heel proofing.

Irwin

In the park, after Irwin gives the trainer a look as if he wanted to bite and is corrected, the trainer gives him a breather to reset before asking again.

Jack

The fright version: when Jack starts panicking at cart and door noise in a store entrance, the trainer steps back outside to a less intense area, lets him sniff, and only then tries going back in, because pushing too hard too fast would cost a young, previously feral dog the trust he has built.

Kali

In September sessions the trainer says what baseline means for her: not frozen, not feeling pressured, back in a headspace with no pressure in it. After a stressful heel rep he lets her run and celebrate and then gives plenty of recovery time, and he waits until she is fully back at baseline before the last reps.

Max (Goldendoodle)

At the go-home lesson Kelton puts the return to baseline inside the learning model itself: a surprise, then a period where the dog is neutral and nothing is happening, then a second trial, where the learning shows.

Piglet

Piglet's trainer turns the rule on himself: Piglet is not happy when she hears one command, and he wonders whether he needs to leave more time for her baseline to return between reps, although he has been leaving a long gap.

Rocky

The clearest statement of why, on day one: several minutes of hanging out and watching TV between down reps, because if he is not allowed to return to baseline the pressure never ends and he learns that it never went away even though he did the right thing. At a later follow-up, after one serious correction, the trainer sends him off to sniff so he does not become a glue dog, and checks that he has fully recovered and is still doing regular dog stuff.

Whiskey

After a run of down reps the trainer lets her reset to baseline with sniffing before the next one, and reads what she shows him to decide that a little more e-collar pressure next time will not cause any real negative affect.

Reassurance after a fright: say nothing, or say it's okay?

How to read it

The two are answering different events, and the record makes the distinction available without drawing it: Ramsey's case is an involuntary startle at something in the environment, where sympathy would ratify the dog's reading of it; Stormie's is pressure the handler himself has applied, where the reassurance is about the relationship rather than about the object. Nobody on film states that distinction, and both instructions are given as general practice by the handler who gives them.

How long is long enough?

How to read it

These may be different kinds of repetition rather than a disagreement: what follows Zeus's correction is a walk in which nothing is asked of him and the handler only observes, which is arguably the reset rather than a second trial. But the same trainer also says, in the same session, that Zeus is still ramped up minutes later - so the observation is being taken from a dog who is on the record as not yet back to baseline.

Let the freeze resolve itself, or break it?

How to read it

Both handlers make the same cut between a dog who is working it out and a dog who is stuck, and both say the stuck dog needs to be moved. Where they differ is the threshold and the tool, and neither gives a way to tell in the moment which one you are looking at. Lola's second lesson offers the nearest thing to a test - grade it by the recovery and by where the dog goes - but that is a judgement taken afterwards rather than a decision rule for the thirty seconds you are standing in.

What do I actually do in the minute after I correct my dog?

In these lessons, the next interaction matters as well as the correction. Louie’s session describes free time and returning to baseline (watch, 25:23). Lola’s trainer asks the handler to make the interaction pleasant again and avoid carrying frustration forward (watch, 09:29). Watch what the handler does next and whether the dog resumes normal activity. These are examples of recovery within particular lessons; they do not establish a universal one-minute sequence or show that every correction is appropriate.

How long should I wait before asking again?

Longer than feels natural. The figures that get named are four or five minutes of free sniffing for a hard recall; four place repetitions across an hour for a nervous dog; ten or fifteen minutes indoors for the same dog when she was worse inside; and an hour-long walk to absorb one pressured sit. The one numeric rule anybody gives is a ratio rather than an interval: early on, ninety per cent of the session is walking and ten per cent is commands, sometimes ninety-five and five, tightening to eighty-twenty once the dog is comfortable, with seventy-thirty called intense. The shortest figure given to an owner is at Stormi's go-home lesson: about thirty seconds of just being happy with a worked-up, nervous dog before asking again indoors - a re-ask after a reinforcement rather than a second hard repetition.

Isn't all that sniffing just wasted training time?

The trainers argue the opposite, and one of them puts the sniffing and the pressure in the same sentence on purpose: the exploring is what rehabilitates a dog, mixed with the negative reinforcement, so that what he learns is both that he can cope with stress and that his life returns to normal afterwards. On that account the exploration is not a reward bolted on to the training, it is the second half of the lesson - the evidence that complying makes the pressure go away and the ordinary world come back. The bookkeeping is stated out loud in one session: the work is named, the fight is admitted, and the chill time follows regardless.

How do I know when he is actually back to normal?

The readouts the trainers use are behavioural and informal, and they are worth knowing because none of them is a clock. One handler reads the tail: he resets the dog with a let's go, moves her, rubs her, and goes again when the tail is wagging. Another names the shake-off as the visible sign that a dog has processed something. A third gives the negative version, which is the most useful one, because it tells you what not to be fooled by - the behaviour has stopped but the dog is still a little ramped up and is sniffing compulsively everywhere the other dog has been. A dog exploring the environment is back; a dog checking on you, or stuck on the last thing, is not.

My dog jumped at a loud noise. Should I comfort him?

On the instruction given to one family, no - and the reasoning is specific. The jump itself is involuntary and is not the problem; what matters is what the dog does thirty seconds later. So you wait about forty-five seconds and watch rather than walking him away, and you give him no input, because oh, it's okay, buddy tells him it was a big deal. If the thing that frightened him is an object rather than a noise, the intervention is the register you use: walk him over and talk to him like a peer - dude, look at this awesome box - rather than sympathetically, and if it keeps mattering, make the object part of a game.

Can too much of this go wrong? What is the worst case?

A nervous dog, or a dog who does not listen well, reinforced so heavily with the collar that it becomes frightened to leave the handler's side and spends the rest of its life following them around - and, in the trainer's words, repressed so hard that the dog has no soul any more. He says he is not a fan. That is the outcome the two-beat correction is designed to avoid, and it is why the same program corrects lightly on a dog like that and then releases the pressure completely. The standing goal is stated separately and used to cancel a planned exercise on the spot: try not to traumatise a dog, in any session.