Baxter

A dog who plays, which the trainer treats as the best news available at the start: he chases the ball, brings it back and growls back in play.

Moderately sensitive to the e-collar: nothing visible at 5, a clear but small response at 10. The trainer's own advance guess was that the third household dog, Ellery, would feel it lower than Baxter.

After his first feel of the collar he went flatter in the game, and the trainer worked to bring him back up. He may also have wanted to guard the ball.

Physically strong and able to brace: the down was hard for him because he can lock his legs against downward pressure, not because he knows sit better.

A large head for his neck, so the flat collar has to be set by feel.

Calm on the road, with a recall the trainer thinks is probably already better than Tyson's; the owner agrees. One unexplained remark about protection instincts during the walk.

Jealous of attention. Tyson keeps out of his way and, in the owner's words, was afraid of him. She adds that a dog was 'kept in a cage' and so shares poorly; the sentence runs on from Baxter's jealousy and probably refers to him, but it does not name him. 'Mr. Jealous', said while the dogs are being greeted, is probably Baxter for the same reason.

Manageable on walks, walked off-leash in the neighbourhood, and prone to stopping and falling behind. The owner walks him loose while she walks Tyson so that she only has one dog pulling.

Knows 'come' and a second call, 'hurry hurry', but the owner finds his recall slow and sloppy. Off-leash he ranges further and looks for cats.

Cautious at the car since hitting his head getting out of it, which the trainer calls learning from natural consequences.

In a hot-weather ball game at home he ended the game himself by trying to bury the ball.

A groomer who also trains told the owner that Baxter is not reactive and is easy to train. This is the owner reporting a third party.

The e-collar was brought in to reduce how hard the owner would have to pull, not because Baxter needed more pressure.

The trainer's word is friction: a method the owner finds physically hard to repeat, such as bending and hauling on a leash, is a method she will stop doing.

After one hard down, no second rep; the owner was asked to make the call.

He was tired, breathing heavily, and had been confused. A stressful prediction error needs time before the next trial, and back-to-back failures teach nothing.

When the first taps flattened him in play, the trainer changed the game rather than pressing on.

Motivation and emotional state govern learning; the trainer had told the owner indoors that he will sometimes do the technically wrong thing to keep the dog's state up.

Taps called out aloud on the second visit, for the owner to hear on the day and on the rewatch.

Because she had told him she could not see his presses in the first lesson's video.

The trainer became the dog.

A dog cannot tell the owner her timing was off and can be confused by practice on it; a person can report exactly what he felt and when. He says he now teaches new trainers the same way.

For Baxter specifically: the collar became the backup, and the task became finding him a reason to fail.

His recall already works in easy conditions, so there is nothing to reinforce until something tempts him away.

Tyson was worked before Baxter at the second visit, and by the trainer's hand before the owner's.

The first visit had named Tyson's pulling a safety issue, and he had just pulled her down on the retractable.

Operating the e-collar

Finding and choosing the level

Pressure, release and the flat dog

Prediction error, and spacing the reps

Everyday standards make emergency obedience

Play and verbal first

Homework after visit 1: down and sit, with a little pressure

The walk: a longer line instead of a six-foot leash

The recall sequence, and what each word must mean

Release back to freedom

Dose, one dog at a time, and practice until automatic

Baxter learned 'out' within the session.

The trainer, 27 May

The second down trial was much easier than the first, without the e-collar and with barely any pressure.

The trainer, 27 May

Baxter's recall is probably better than Tyson's. The owner agrees.

The trainer, 24 June

The owner's timing became correct - on a person.

The trainer, 24 June

Things are 'a hundred times better than the first time'. What is better - the walk, her handling, or the dogs - is not said.

Unattributed, end of 24 June

Baxter is not reactive and is easy to train.

A groomer-trainer, as reported by the owner, 24 June

On 27 May Baxter comes on a line to 'come' with 'uh-uh' and collar taps, and once to 'come' without the collar.

The first pressured down is long and hard. The audio has minutes of repeated 'down', the trainer's running account of head-shaking, sitting and evading, a reset, and a finish he calls close enough with the dog partly on his side. The easier second trial rests on the trainer's narration; the audio cannot show the dog's position.

The owner does not handle Baxter on camera at the first visit. The trainer demonstrates; she is told she will try it next time.

On 24 June the owner calls Baxter several times on a short walk and he comes. Her first full correction sequence on him reached no collar at all, because the transmitter had been left behind, and he came on the marker and the repeated 'come'. Later recalls come without the button. No filmed moment establishes that the owner stimulated Baxter at all.

One recall came toward the owner and went past her.

The owner's first correction on a real dog (Tyson) and her first on the trainer both put the tap straight after 'come' with no 'uh-uh'. The same error appears twice in one afternoon, which is the concrete case for the drill.