How do I know if my dog is actually reliable?

Lola

Lola bolts when her nerves take over. At the second owner lesson Kelton tells the owners what eight good repetitions can hide: the two times out of ten she bolts. He acts on it, calling her in at a narrow spot rather than risk ending a good lesson chasing her. In a night session another trainer finds no bolting and refuses to credit it yet: it was dark and quiet, so the retest is two or three in the afternoon. At go-home the owners are told plainly that if she works out nobody is holding the long line, they will not catch her.

Bo

Two habits. On a trail, Bo's trainer marks a successful down as 'the first rep' while saying Bo does not know it well yet. In the next session he names the help in the first down and schedules it to shrink. Later, because the owner wants to board Bo, he stages the real thing: a mock boarding kennel, an unfamiliar crate, Connor (who has barely worked Bo) and Claudia (who never has), gloves, a door that will not latch, other dogs barking. The next round is to be handlers of different skill levels, not all professional trainers, then nails clipped on camera and a vet visit.

Zeus

Late in Zeus's program, after a session in which everything went right, the trainer says he thinks Zeus would have listened off leash. He declines to test it: Zeus is a very big dog, and there is liability. Removing the backup is not something you try out. 'You want to know not hope,' said twice.

Finn

At Finn's first owner lesson the trainer explains why the recall is proofed where Finn could genuinely leave: dogs know whether they can get away, and in the training yard he knows he can't. He also keeps a slip lead on as a backup, because Finn has slipped his collar once before.

Louie

Louie was a rescue with a bite history, placed with a family on rural land. The proof of the placement is not an obedience test. It is an ordinary family meal, filmed, with Louie loose among grandparents, parents, children, other dogs and food. Nothing much is asked of him and nothing is prevented. Earlier on placement day he approaches to be touched, and a staff member reads him as relatively chilled and not messed up by it.

Reaper

Reaper is a very large dog in a serious behaviour-modification program. At pick-up his lead trainer, in a bite suit, explains the staged handoff. Most people never give a dog the chance to try, so nobody knows what the dog will actually do: 'we won't know until we see.' Two months later, on his first public trail walk, the trainer notes that Reaper has never been able to go anywhere like it. At the first owner lesson, after going through how to handle visitors and tight spots on the trails, the trainer asks the family to have the people Reaper has had problems with at the next lesson. With him there and 'safety equipment just in case', the encounter can be set up safely and the outcome seen.

Fiona

Fiona shows some reactivity on walks, but her trainer suspects she has never actually been near another dog without a leash or a barrier. So the reactivity may be leash and barrier reactivity only. He sets up a test with two staff dogs walking through, and tells viewers not to test a small dog with a big dog themselves. Later the staff plan cat and small-animal tests once her foundation is good.

Irwin

His trainer tries a food-scented ball to build sniffing and play, and Irwin guards it: the first time anyone has seen guarding, including at feeding.

Hank

He is walked and tested on the leash first to find out if there is a specific issue. Then he is put in a play yard next to barking dogs so any fixation shows, because his trainer cannot correct what he has not seen.

Pepper

On a store trip Pepper's trainer hands the leash to a helper, Alec, to see whether her skills hold with someone else. She pulls a little toward the trainer. 'That's why we're testing it out.' She then holds a sit and takes the release from Alec. Later in the program the same trainer runs it as a rule on a busy outing: he deliberately bends and touches the leash while she holds a position, she takes it as a release, he corrects it and tries again - 'she would have never learned that if we wouldn't have tried' - so 'we make sure that we try all kinds of stuff like that'.

Nikki

A rescue dog being prepared for re-adoption, with a reported bite (at the vet, the handler says, and 'we don't know' exactly what happened). The report is not taken on trust either way. Two handlers have tried to get her to bite in gloves, and the test is built around what a child might do - pulling and tugging at her - because 'we wanna make sure that she won't bite any kids'. A week later a growl at the kennel door, seen once and not filmed, gets its own session: 'this is a video to see if she growls at me again so that we can correct it'. The handler predicts 'all barking no bite' and protects the hand anyway. When she goes back in cleanly, that is not the end of it: 'we'll keep testing to see if we can get a growl again like over the next couple of days'.

Chaparro

A young, large dog at his first in-home evaluation, with a Spanish-speaking family who had told the staff he bites. The staff let him meet them loose to find out - 'we wanna see if he's he's gonna bite him' - and came prepared for a bite. He does not bite, and the staff member interpreting explains the test to the family in Spanish: 'que queremos ver la diferencia de que si él va a morder en verdad o sólo es que está viendo o cosas así' - what we want to see is the difference between whether he will really bite or is only watching; since he does not bite, he is not aggressive, he has reactivity toward other dogs. The fights and the bite the family describes were their account, not tested on the visit; the reading is the evaluator's at a first meeting.

Apollo (the Sharkie)

Late in a place-duration session the trainer rates Apollo's recall but will not call it proven: it is 'good enough' that he is not worried about him running off, yet it still has to be tested around higher distractions - and there was nothing very distracting there anyway.

Astro

At an update evaluation the trainer explains why Astro's recall on a six-foot leash proved little: he never got close enough to anything to want it, so he never had to leave something to come back. The longer flexi lets him get interested first. At go-home the same caution applies to greetings: he now seems interested in people, 'he might be fine for greetings i don't know for a certainty'.

Callie

On Callie's first day the trainer does not assume she is or is not dog-reactive: seeing another German Shepherd, she walks her toward it to find out - 'maybe she has', 'maybe she doesn't'.

Kodak

Kodak and Rocko had never been loose together in the same area. At the second owner lesson Kelton says he has not yet had them loose together in the house and plans to give it a shot that day, in steps - leash let out first, 'probably not loose at first' - so the family sees what the dogs do rather than guessing.

Lucy

At Lucy's in-person evaluation the trainer stages recall tests to find a failure and gets none: she whips round every time. He is 'disappointed but it's not her fault', and tries 'one last thing' - taking her off the leash - to see whether anything shows.

Milo

At the off-leash recall session the trainer says a recall from a dog with nothing better to do proves little - untrained dogs do that. What counts is Milo and Josi coming back while the children are running and they want to go with them.

Oscar

At go-home the trainer refuses the easy demonstration: calling Oscar from close by when he 'wasn't doing anything' is 'not like a big test', so he recalls him away from the clients instead, then uses the release to check whether Oscar came because he was called or because he wanted to (uncertain audio).

Peque (PK)

At a refresher the owners report that Peque has been reactive, especially to bigger dogs. The trainer sees her hackles go up but no 'straight reactivity', and rather than settle it either way plans to have that family meet him somewhere while he is out training, to see how she does with them.

Piglet

Kelton takes Piglet to chickens as a deliberate test of her predatory interest, wanting her to try something so it can be penalised while he is there: 'I'd like her to do something so I have a chance to penalize her for it'. She mostly ignores them, and the next chances are planned - one more here, and another with a cat at her go-home lesson.

Rocky

After a few reactions in the same two or three days, Rocky's owner stopped letting him near dogs - and says in the lesson that this means nobody knows whether he has already grown out of it. The trainer's answer is to let him go and check, and respond only if he does something inappropriate.

Thumper

Thumper's first meeting with a dog on a walk is treated as a first: the trainer says she did not know what to expect and assumes the worst with any dog she is meeting for the first time. He does better than expected.

Test it, or don't take the chance?

How to read it

Both halves are the same principle - know rather than assume - applied to different stakes. Where failure is recoverable and informative, the trainers stage it; where it is dangerous or irreversible, they refuse to find out. The 30 Sep footage adds the trainers' way of running a costly test: staged by a professional with protection - gloves for a reported biter (Nikki), the trainer present with safety equipment for the visitors Reaper has had trouble with - rather than left untried or tried by the family alone.

My dog has been perfect for weeks. Is he trained?

Maybe, or maybe nothing has happened yet. Kelton's example is a dog who bolts two times in ten: if you only ever see the eight good ones, you think he is fine. Test him in the situations that matter, where he could fail.

When can I take the leash off?

When you know, not when you hope. Zeus's trainer believed Zeus would listen off leash after a perfect session and still did not try it. Finn's trainer says a recall only counts where the dog could really leave.

How do you test a dog properly?

Make the test the real thing. For Bo, who was going to be boarded, that meant a mock kennel, an unfamiliar crate and people who barely knew him, then handlers who are not trainers. For Pepper it meant handing her leash to someone else in a store.

My dog is reactive on walks. Does that mean he is aggressive with dogs?

Not necessarily. Fiona's trainer suspected her reactivity was only leash and barrier reactivity, because she had never been near a dog without one, and set up a controlled test to find out. He warns owners not to run that test with a small and a big dog themselves.

The rescue says she bites. How do you find out?

On film the report is neither accepted nor dismissed. Nikki is handled in gloves the way a child might handle her, by more than one person, and a growl reported at the kennel door is deliberately set up again on camera 'so that we can correct it'. One clean result is not the end: 'we'll keep testing ... over the next couple of days'. At Chaparro's evaluation the staff let him meet them loose, prepared for a bite, to tell real biting from watching and reacting. These are tests run by trainers with protection, not something the clips suggest an owner try alone.

My dog always comes when I call him at home. Is his recall reliable?

Not yet proven. Several trainers say a recall only counts when the dog leaves something he wants. Calling a dog who 'wasn't doing anything' is 'not like a big test' (Oscar's go-home, uncertain audio). Milo's trainer waits until the children are running. Astro's trainer says a six-foot leash never proved it, because Astro never got close enough to want anything.