Field Notes · An evolving learning library from Angier K9
ANGIERK9FIELD NOTES
Explore Angier K9

The session: dose, venue and proofing · He has to be allowed to get it wrong

Should I stop my dog before he misbehaves?

A dog cannot learn that a choice is wrong if he is never allowed to make it, so in training you stop blocking the mistake and let it happen where you can answer it - carefully, not too often, and with the dangerous ones staged by a professional rather than tried alone.

The answer

Mostly, no - not in training. Kelton, Reaper's trainer, says he can't punish the bad choice if the dog doesn't make it, and calls that a mistake a lot of people make. Kelton tells Lola's owners he needs dogs to blow him off, or he never gets the chance to make the command reliable. At Roxy's go-home lesson the family is told to stop managing the jump with the leash or their body: let it happen, and either she doesn't do it and you praise her, or she does and you have the chance to answer it. For a recall, Finn's trainer names two conditions - the dog has to want to go somewhere else, and has to actually leave you - or there is nothing to practise. The September 2026 sessions say it again in each kind of case. Kelton again, with Reaper: 'I'm searching for it i'm like i need to see it you need to actually do the behavior for me to be able to correct it'. Pepper's: 'I let her make the mistake and I just correct it every time she makes the mistake so that it's real life'. To Sofia's family Kelton says it in Spanish: 'voy a dejarla escoger lo que va a hacer si es algo bueno o algo malo y va a tener las consecuencias del error' (I'm going to let her choose what she does, good or bad, and she'll have the consequences of the error). For Hattie and the household cat the trainer hopes 'she goes after the cat', because 'you can't correct her for looking'.

So the trainers go looking for the failure rather than waiting for it. They let a dog wander until something interests him before calling him, open the truck door beside a dog in a down, keep a lead long and loose past other dogs, let strangers greet a mouthy dog on condition he is corrected if he mouths, and, before a go-home lesson, one trainer hopes the dog will do one dumb thing in front of him because otherwise the call comes the next day. When it happens, it is welcomed: the most a dog has ever tested you is a good day, a blown recall is a learning rep, and a session with a lot of argument in it can still be a good one - the trainers' own term is healthy conflict. With some dogs, progress arrives in a handful of these moments rather than rep by rep. In September 2026 the same welcome is heard on Beo's down ('i need that to happen sometimes that's part of the proofing process'), on Kane's uninvited jump, which is allowed, answered with the conditioned punisher and then put on cue, and in Finn's Shorts, where the handler hopes Finn won't come straight back so there is a recall to prove.

The wider set of session notes sharpens what the 'choice' is. It is the act, not the feeling or the look: Bandit may pass another dog with his hackles up, because 'he hasn't actually made a bad choice yet he just feels a certain way', and Peque's trainers point out that they did not restrict her choices 'artificially before she did anything'. It is not the thought either: on Kalypso's place cot the trainer will not correct her for considering stepping off - 'don't preempt it let her make the choice first'. And the shape around the choice is the same in Astro's, Piglet's and Penny's sessions: let it play out; a good choice passes or is praised; a bad one gets a consequence, then 'let's go' and move on. Preventing the chance instead leaves the handler 'stuck doing that forever' (Penny) and a dog who 'never learns' (Piglet).

The same files set limits. Do not run the error setup too many times in a row. When the failure is dangerous it is staged, not awaited: the first meeting of two dogs who have never got on is run where professionals can contain it rather than at home; a dog's introduction to the two dogs she has had problems with is moved to a day with time to make it foolproof. The September 2026 sessions add two more staged failures: Hattie's trainer offers to run the first cat set-up himself, because what to punish and when 'would involve reading the dog' and 'that first one a lot of times like darn near takes care of' it; Reaper's second owner lesson is planned as 'a scenario where we can punish the bad reaction' to visitors, so the family need not micromanage it later; and Nikki's kennel return is re-staged on camera because 'we want to see the behavior so we can correct it'. And some handlers in the Roxy household do the opposite of letting it happen - they set the dog up so the behaviour cannot be completed, a down placed back from the door so the door can close on the attempt. The record holds both and does not reconcile them. What an owner should take from it: in training, with your trainer, let the mistake be possible; do not stage bites, bolts or fights at home on your own.

Angier K9 is a family-owned, play-based Board & Train serving Angier, Raleigh and the Triangle, with three full owner lessons.

What to do

In practice sessions, stop blocking the mistake with the leash or your body: give your dog the chance to jump, break or leave, and be ready to praise the good choice or answer the bad one. Find temptations strong enough that he might fail. Don't repeat the same setup over and over, and don't create risky situations - with other dogs, with people he might bite - on your own; those are staged with your trainer.

What the trainers say

um a major correction because he needs to make the choice the bad choice in order for me to punish it i can't punish it if he doesn't make it that's a mistake that a lot of people make

Reaper · Watch at 47:41 ↗

then I then I have to figure out some way for them to I need dogs to blow me off when I'm training they have to if they don't blow me off ever I can't make make a reliable command because I can't ever have an opportunity to obligate it

Kelton · Lola · Watch at 1:20:33 ↗

all right, like let me like manage her with the leash or with other things or with my body and like make sure it doesn't happen, let it happen because either number one will happen, which is she won't do it, and you go, "Oh, good job." Right? um or they will do it, right? And then if she does it, then you have an opportunity to give a consequence for the choice she made and then she'll make a better choice next time, right?

Kelton - the lead trainer, at Roxy and Ramsey's go-home lesson · Roxy · Watch at 13:08 ↗

in order for me to proof it, I need him to be motivated to go somewhere else other than me and I need him to blow me off.

the trainer · Finn · Watch at 7:29 ↗

Um which is good. We need that. We want it. Yeah. Um because there's no learning without error, right? Um errorless learning is fake in my opinion.

Kelton · Lola · Watch at 5:12 ↗

So learning, both in dogs and in humans, is not errorless, right? So they make errors, and you need to give them those choice points for them to actually make the error, like not coming back or something like that, in order for them to start learning.

the trainer · Roxy · Watch at 26:50 ↗

you don't need to do it too many times in a row otherwise you end up with problems

the trainer · Roxy · Watch at 27:47 ↗

And that's what proofing is all about. You have to put them in situations where they're going to fail like that.

the trainer · Bullet · Watch at 10:08 ↗

This is actually the most she's ever tested me, which is kind of a good Thing, because that means I can like reinforce it while she's really giving Me some bite.

Connor - the trainer · Luna · Watch at 3:47 ↗

I hope that he at least a few times doesn't listen to you right now I'm hoping for it because that's the most instructive thing

Kelton - the trainer, at Zeus's second owner lesson · Zeus · Watch at 59:57 ↗

That's actually, I was telling him on the way here, I was like, man, if I could get only one thing I prayed for today, you know, like if I can only get one thing out of this go home lesson, it'll be that Sputnik does one thing that's dumb.

the trainer (probably Kelton) · Sputnik · Watch at 28:05 ↗

I'm always happy like I want that to happen with us not at home yes you have to do it you have to try it because if it happens at home could be a big disaster if it happens here we can mitigate the disaster

Kelton - the trainer, at Rocko and Kodak's first pack walk · Kodak · Watch at 12:08 ↗

glad he blew me off a little bit there because it just gives me an opportunity to clarify things but then we have an easy outlet for him to go have fun again that way he can see like we're still buds

the trainer · Teddy · Watch at 17:34 ↗

If the door opens and they you can't like put them back. They were already gone. But if you put them in a down a little bit away from the door, you can close on them, right?

Kelton - the trainer, at Roxy and Ramsey's second owner lesson · Roxy · Watch at 2:17 ↗

moments with him a lot of times with this kind of thing it's like you need like I'm searching for it i'm like i need to see it you need to actually do the behavior for me to be able to correct it

Reaper · Watch at 1:25 ↗

And again, I'm going to let him make his choice first, and if he makes a bad choice, then I will cause a problem.

Reaper · Watch at 5:31 ↗

just practice that where like I let her make the mistake and I just correct it every time she makes the mistake so that it's real life

the trainer, in-program owner lesson · Pepper · Watch at 17:41 ↗

con sus dueños y vamos a ver lo que ella va a hacer voy a dejarla escoger lo que va a hacer si es algo bueno o algo malo y va a tener las consecuencias del error

Kelton, in Spanish, narrating; meaning: '...with their owners, and we'll see what she does. I'm going to let her choose what she does, whether it's something good or something bad, and she'll have the consequences of the error.' · Sofia · Watch at 17:22 ↗

yeah and so like what i would be hoping for is that she goes after the cat i want her to go after the cat right

the trainer, follow-up lesson with the adopters · Hattie · Watch at 50:49 ↗

anticipated that i might ask him to uh when he recalled so i let him do it again this is where allowing them to make choice is important right so i allowed him to do it and then i penalized it

the trainer, narrating to camera · Kane · Watch at 15:08 ↗

down fail all right so that was perfect right i need that to happen sometimes

the trainer, narrating to camera · Beo · Watch at 30:30 ↗

that's what we want we want to see the behavior so we can correct it

the handler, narrating to camera · Nikki · Watch at 1:57 ↗

You need them to make a mistake. You need them to make a mistake. Because they need to try something and get a prediction error.

the trainer · Thor · Watch at 2:29 ↗

because uh kind of like what we were talking about um a little bit ago is that in order for them to learn they have to have the opportunity to make the choice um otherwise you're just avoiding it you're just managing it forever

the trainer · Piglet · Watch at 8:59 ↗

yeah and also then you're stuck doing that forever right um and so what's better is you just let them make that choice if they look at it and they're like okay and then they keep moving

Penny · Watch at 33:10 ↗

because if they have the option between making the mistake and not making a mistake and they choose to do it correctly that gives you an opportunity to praise um and and reward if they make the choice to make the mistake that's also very useful

the trainer · Kalypso · Watch at 40:14 ↗

We let her go look. She even walked towards the dog and we said nothing.

the trainer · Peque (PK) · Watch at 1:42 ↗

learn, afraid to try, and afraid to let your dog make mistakes, afraid to let your dog experiment, afraid to let them try. That's a certain failure.

the trainer, narrating to camera · Jack · Watch at 26:21 ↗

How it looks, dog by dog

Roxy

The Roxy and Ramsey household holds every side of the idea. At go-home the family is told to stop managing the jump and let it happen so the choice can be answered, and a long-line session says learning is not errorless - while warning not to run the error too many times in a row. The same trainer says nervous dogs are allowed to choose the wild option until nothing happens.

On film

all right, like let me like manage her with the leash or with other things or with my body and like make sure it doesn't happen, let it happen

Watch at 13:08 ↗

feel the urge to flee. Um that's actually what I do with like dogs like this that are really nervous. I just let them do stuff. I let them choose.

Watch at 15:10 ↗

If the door opens and they you can't like put them back. They were already gone.

Watch at 2:17 ↗

Lola

Lola is too compliant, and that is the problem: she gives the trainer nothing to enforce. He goes looking for the blow-off - long line, places she will want to leave - and teaches a second trainer to spot the moment she is considering it, visible in her face before she moves. Kelton tells the owners errorless learning is fake, in his opinion.

On film

you just look for opportunities for her. You look for, you try to create, like, times where she blows you off. Like, you need her to blow you off. If she doesn't blow you off, she won't.

Watch at 17:26 ↗

yeah you can see when they're thinking maybe I might bolt right maybe I might run away and you're looking for that scenario to happen so that you can make it a bad thing

Watch at 2:10 ↗

Finn

Finn's recall work states the two conditions for proofing - he must want to go elsewhere and must actually get distance - and the trainer builds them on purpose: the truck as bait, noises, the store where it went badly before. He also checks whether a correction makes Finn want to bolt. In a September 2026 Short the handler, keeping Finn cautious because of his flight-risk history, says it would be convenient if Finn didn't come back right away, so there would be one more chance to prove the recall with the collar.

On film

in order for me to proof it, I need him to be motivated to go somewhere else other than me and I need him to blow me off.

Watch at 7:29 ↗

and I want to see how he handles that little bit of pressure there from the e-collar, see if that makes him take off

Watch at 3:00 ↗

Did give me one more opportunity to flip the recall with the e-collar and like didn't come back right away

Watch at 0:13 ↗

Deuce

Connor tells Deuce's owners he will spend the next few days setting up scenarios in which Deuce tries to jump on him, because some behaviours have to be set up to be fixed.

On film

but probably the next couple days I'll start working like I'll be trying to set up scenarios where he's going to like try to jump on me and stuff.

Watch at 2:35 ↗

There's just like weird behaviors that sometimes you just got to like set them up and just fix them.

Watch at 2:41 ↗

Bo

Bo is put on a long line and allowed to wander until he finds something worth sniffing, so the recall is tested against something real; the first call is refused and answered.

On film

I think I'm going to let him wander around a little bit now and get distracted so that I can do a recall with him.

Watch at 7:50 ↗

Bo, come. Come, come. No. B, come. Nope. Come. Yes. Okay. Good boy.

Watch at 8:15 ↗

Louie

Louie, a dog who has bitten, is the case where the failure is sought out and handled with care. A trainer keeps the room low-pressure so the reactivity will show and can be learned from; a clean statement from another session says it matters to have a few moments of authority over the dog, and that the owner will need a couple of those too. The same household is told they do not want anybody bitten.

On film

I still want to see it and um try to allow it to happen because if we can get it, then we can we can allow the prediction error to happen and he can learn from it.

Watch at 0:41 ↗

And it's important to have those little moments where you're able to have some kind of authority over the dog.

Watch at 26:16 ↗

Luna

Mid-fight on an empty street, Luna's trainer says this is the most she has ever tested him - which is good, because it is when reinforcement means something. He adds that testing is a stage that ends, and hands the same expectation to the owners.

On film

This is actually the most she's ever tested me, which is kind of a good Thing

Watch at 3:47 ↗

But dogs do their thing. They'll test you. You just gotta make sure that you reinforce it. Same thing as when the dogs go home, right?

Watch at 2:20 ↗

Zeus

At Zeus's owner lesson the trainer says before the owner handles him that he hopes Zeus disobeys her a few times while he is there, because it will happen later when he is not. When it does, he calls it the most instructive thing that could have happened, and says he would have been disappointed by a smooth demonstration.

On film

I hope that he at least a few times doesn't listen to you right now I'm hoping for it

Watch at 59:57 ↗

that was the most instructive thing that could have happened

Watch at 1:25:20 ↗

Dolce

Dolce's trainer closes a hard session by calling it good but not an everything-went-well one - there was a lot of healthy conflict.

On film

well that was a really good session with dulce i well maybe not the most everything went well type one we did have a lot of healthy conflict

Watch at 16:43 ↗

Kodak

Kodak and Rocko, who had only ever met badly through a gate, are introduced on a pack walk because the trainer would rather it go wrong with professionals present than at home, where it could be a disaster.

On film

just getting them used to each other because they've never interacted before ever other than through the gate and those never went good

Watch at 11:47 ↗

I'm always happy like I want that to happen with us not at home

Watch at 12:08 ↗

Dutchess

Dutchess's introduction to the two dogs she has had problems with is deliberately moved off a booked hour to drop-off day, so it can be made foolproof; the risk is an emergency, and scheduling is the safety measure.

On film

to say that is i don't want to bring the two in that she has a problem with quite yet i want to handle that on like the drop off day

Watch at 9:11 ↗

sometimes when there's bigger stuff like that i like to make sure everything is really nice and tight that way like it's very foolproof

Watch at 9:40 ↗

Bullet

Bullet wanted to go outside, so he was asked to down by the door; when he failed, the trainer said that is what proofing is - putting the dog where he will fail so he can see the consequences of both choices.

On film

so he really wanted to go outside and get some more freedom so when I was asking him to down

Watch at 9:40 ↗

So that you can show them the negative consequences of not doing what you ask and the positive consequences of doing what you ask.

Watch at 10:14 ↗

King

King is left in a down beside the truck with the door open, chosen because it is exactly where he will want to go.

On film

he's really going to want to go get in the truck when i open the door so that'll be a test for him see if he holds it down

Watch at 24:19 ↗

Teddy

Teddy's blown recall is welcomed as a chance to clarify, followed at once by an easy outlet to play so he can see they are still buds; a broken stay is predicted in advance and called a learning rep.

On film

glad he blew me off a little bit there because it just gives me an opportunity to clarify things

Watch at 17:34 ↗

learning rep he learned that if he keeps going away if he decides to do that thing where he runs off from me things go badly but if he comes back to me things are happy

Watch at 30:52 ↗

Reaper

On the trail Kelton (Reaper's trainer) forecasts progress by how many chances for a major correction he gets, because the dog has to make the bad choice to be punished for it - and he warns that owners make a big deal of a look instead. The September 2026 sessions make it Reaper's method. On a trail (16 Sep) the trainer announces the rule: let him choose first, cause a problem only for a bad choice, leave it alone otherwise. In the owner session (19 Sep) the next lesson is planned as a set-up with visitors so the bad reaction can be punished. On a walk (25 Sep) he hopes for 'at least one chance to punish' and lets the encounter develop; at the in-program lesson (26 Sep) he says he was searching for it, because the dog has to do the behaviour to be corrected; on 29 Sep he regrets missing a dog in the distance.

On film

um a major correction because he needs to make the choice the bad choice in order for me to punish it

Watch at 47:41 ↗

And again, I'm going to let him make his choice first, and if he makes a bad choice, then I will cause a problem.

Watch at 5:31 ↗

second lesson if you guys can have whoever you guys want you there um we we will like set up a scenario where we can punish the bad reaction

Watch at 59:49 ↗

really big one for reaper and so i'm hoping that today i get at least one chance to punish something

Watch at 3:39 ↗

moments with him a lot of times with this kind of thing it's like you need like I'm searching for it i'm like i need to see it you need to actually do the behavior for me to be able to correct it

Watch at 1:25 ↗

wish i'd been a little closer to that probably gonna miss that opportunity

Watch at 14:25 ↗

Sputnik

Before Sputnik's go-home lesson the trainer says the one thing he prayed for was that the dog would do one dumb thing, because a dog that behaves at handover produces a phone call the next day. In the field he recalls the dog once and then decides to see what he does natively.

On film

Please. Just do a dumb thing.

Watch at 28:21 ↗

Just got to recall the dog briefly there. Next time I'm just going to see what he does natively.

Watch at 5:36 ↗

Baxter

Baxter's recall already works in easy conditions, so the owner is told to find something he is more interested in than her; the trainer wants dogs to fail the first time.

On film

You want them to fail the first time.

Watch at 14:24 ↗

The main challenge is gonna be you're gonna want to find something that he's more interested in than you

Watch at 9:44 ↗

Bernie

From day one Bernie's trainer wants him to break so there is something to reinforce, and later lets him put his face on the table so it can be answered.

On film

So if he breaks it, that's going to be an opportunity for me to negatively reinforce

Watch at 17:47 ↗

if he goes to put his face on the table like that's the thing then like let him

Watch at 13:18 ↗

Major

Major's owner corrected before the passer-by test had happened; next time, she is told, let him choose whether to bark or be nice. The handlers then act as if they are just hanging out, on a fifty-foot line, so the decision is his.

On film

next time what happens is you want him to choose whether or not he wants to bark or be nice

Watch at 14:01 ↗

i'll put him on my long line i think it's 50 feet and we'll just be acting like we're just hanging out together and letting him make the decision

Watch at 10:27 ↗

Milo

Milo's reaction to other dogs only happens elsewhere, so the trainer sets up a meeting with a little pressure and a helper dog chosen to be unbothered, so the reaction happens in front of him and can be stopped.

On film

how it goes cuz I kind of want to like have a little bit of pressure I want an

Watch at 1:31 ↗

that's not the right dog i would use my personal dog for that that's my tester she's an anatolean

Watch at 9:00 ↗

Clark Kent

Clark Kent mouths people, so strangers become the classroom: anyone who asks may greet him, on the stated condition that he is corrected if he mouths them.

On film

Yeah, it gives me a chance to correct a few mouths.

Watch at 7:04 ↗

I'll only correct him if he mouths you.

Watch at 14:31 ↗

Sofia

On a walk (25 Sep 2026) Kelton narrates the principle in Spanish before two small dogs pass: he will let her choose, good or bad, and she will have the consequences of the error; she makes the bad choice and is corrected ('Eso no se hace', that's not done).

On film

en eso es que necesitas dejarla hacer malas decisiones y también las buenas y corregir

Watch at 8:46 ↗

estamos distraendo eso no funciona no funciona

Watch at 10:03 ↗

con sus dueños y vamos a ver lo que ella va a hacer voy a dejarla escoger lo que va a hacer si es algo bueno o algo malo y va a tener las consecuencias del error

Watch at 17:22 ↗

error. Sofía, vámonos. Eso no se hace. Eso no se hace.

Watch at 17:37 ↗

Whiskey

In a new place Whiskey is expected to fail and learn the rules are the same; a passing dog is set up as a choice point rather than managed away, and she makes a good choice.

On film

so we need to practice leader so that she learns all over again that the rules are the same she's gonna fail it's okay she'll learn through the failure

Watch at 8:26 ↗

right so i gave her a choice point she made a good choice

Watch at 10:51 ↗

Nash

Nash is freed near another dog rather than held, so that charging is his choice and meets the same consequence; he makes a much better choice.

On film

and that's that's that's his choice and then he's gonna get the same problem

Watch at 13:12 ↗

But now he's making a significantly better choice

Watch at 13:17 ↗

Hank

Hank is let into a play yard to fixate so the trainer can see it happen, correct it and redirect it to play.

On film

So, what we're doing is just using one of the play yard areas here and just trying to let him make a mistake, fixate on something.

Watch at 7:46 ↗

Nikki

Nikki's trainer adds a little social pressure while handling her to see how she responds, and she tries to bite - the response the session was built to find. After Nikki growled at the handler at her kennel off camera (28 Sep 2026), the next session is built to bring the growl back so it can be corrected, and later the kennel return is re-staged while she is more excited, to see whether it is still there.

On film

adding a little bit of pressure a little bit of like social pressure with that and stepping in to see how she responds to it

Watch at 6:08 ↗

that's what we want we want to see the behavior so we can correct it

Watch at 1:57 ↗

Now, I'm curious, because there's something else going on, and she is more excited about it,

Watch at 22:45 ↗

Bumblebee

Bumblebee is being allowed to show new behaviours so they can be corrected.

On film

wanting to play when he wants to play and things like that um so we're kind of just letting him show new behaviors and then correcting those

Watch at 2:49 ↗

Lucy

Lucy's lag in a prey chase cannot be fixed mid-chase, so it is set up at moderate distraction and answered at the level the real chase would need.

On film

right so the way that you get rid of the lag for the prey chase is you do it for the you punish the lag when they're not in the prey chase

Watch at 13:22 ↗

Rocky (the Rottweiler)

For Rocky the Rottweiler a helper wears gloves, a hat and glasses on purpose - things that normally trigger a dog - so the reaction can show up in the session.

On film

glasses right so things that normally would trigger a dog all right which is

Watch at 15:00 ↗

Beo

During a public down (23 Sep 2026) Beo thinks about getting up, and the trainer calls the near-failure 'perfect': it needs to happen sometimes, as part of the proofing process.

On film

down fail all right so that was perfect right i need that to happen sometimes

Watch at 30:30 ↗

that's part of the proofing process so he's thought about getting up there

Watch at 30:54 ↗

Hattie

At the adopters' follow-up (22 Sep 2026) they ask whether to punish Hattie for looking at the household cat. The trainer says the answer involves reading the dog, offers to run the first set-up himself, and says what he will be hoping for: that she goes after the cat, because there has to be a behaviour to correct, and you can't correct her for looking. That first correction by the trainer, he says, often nearly takes care of it.

On film

it right and then it's like i could tell you the answer but it would involve reading the dog

Watch at 42:23 ↗

things and the benefit of doing that is that if you've already laid that groundwork and then i do that first one for you that first one a lot of times like darn near takes care of

Watch at 50:20 ↗

yeah and so like what i would be hoping for is that she goes after the cat i want her to go after the cat right

Watch at 50:49 ↗

to correct yes yeah exactly yeah because like yeah you can't correct her for looking

Watch at 51:04 ↗

Kane

On a park session (21 Sep 2026) Kane jumps on the trainer at the end of a recall, anticipating being asked. The trainer lets him do it again on purpose - 'this is where allowing them to make choice is important' - answers it with the conditioned punisher (not the collar), and then asks for the jump, to put it on cue.

On film

anticipated that i might ask him to uh when he recalled so i let him do it again this is where allowing them to make choice is important right so i allowed him to do it and then i penalized it

Watch at 15:08 ↗

Pepper

Not a map member before this update. At the second in-program lesson (20 Sep 2026) the trainer gives the owner the rule twice. For a copied behaviour from other dogs: don't manage it ahead of time, let her try, and correct it at once, before it can happen a couple of times. For a boundary line: let her make the mistake and correct it every time, 'so that it's real life'.

On film

that right um and like the biggest thing is just like okay as soon as they try it like i don't like manage it ahead of time i just let them see what they're gonna do and i let them try it

Watch at 6:11 ↗

just practice that where like I let her make the mistake and I just correct it every time she makes the mistake so that it's real life

Watch at 17:41 ↗

Apollo (the Bully)

Not a map member; added from the session notes. In a Lowe's session Christian works Apollo through tighter spots on purpose, so that he has the chance to drift out of position and be corrected there rather than never being tested.

On film

So he has the opportunity to make mistakes and kind of wander off the box area

Watch at 9:37 ↗

Astro

Not a map member; added from the session notes. Asked at an update evaluation whether to give Astro a replacement behaviour when he fixates, the trainer says he normally does not: he lets it play out, a good choice passes, a bad one is punished, and they move on with 'let's go'. At a later evaluation he separates a real bad choice, which is punished, from something merely unwanted at that moment, which only gets the redirection.

On film

Interestingly, what I normally do is I just let it play out. I just see what happens. If he makes good choices, nothing happens. And then if he makes a bad choice, then I punish him.

Watch at 31:11 ↗

um and so a bad choice would have to be an actually like thing you don't want them to do again to punish it

Watch at 17:08 ↗

Bandit

Not a map member; added from the session notes. Kelton lets Bandit pass another dog with his hackles up: the feeling is allowed, and only blowing off a command or lunging is penalised. In a heel session he steps away from the people and the dog to see whether Bandit follows; he does not, which gives an easy chance to correct it, and he jumps back into heel.

On film

ramped his hair can go up right like that's all fine like i'm not gonna do anything about that because he hasn't actually made a bad choice yet he just feels a certain way

Watch at 13:06 ↗

big step to the right and see if he would kind of like clock that and follow me and stay in the heel um and he did not which i did not think he would and so that gave me an easy opportunity to correct it

Watch at 14:38 ↗

Callie

Not a map member; added from the session notes. At both go-home lessons the family is told to let Callie decide: if she chooses to jump on them, then 'nope' and get her off; around the garden plants, let her roam and correct each approach, so that with time she ignores them.

On film

her to jump on on you just like let her make the decision right and if she does choose to jump on you then you then you then you correct her by saying nope and

Watch at 15:46 ↗

let her free range and just watch her and then whatever she does uh whatever she does that you don't like her like licking the rust

Watch at 9:41 ↗

Doja

Not a map member; added from the session notes. At Doja's drop-off Kelton calls these moments choice points: constant managing means the dog never gets to decide, and is then let loose untested. On the retractable he shows it - a runner passes, he does not steer her away, and she does not even look.

On film

but if you're constantly managing then they never get to make the decision right and then when they're let loose they're

Watch at 4:00 ↗

instinct is to go right and make sure she doesn't go over to the guy on the leash right but instead i just waited to see what what is she gonna do

Watch at 5:19 ↗

Fiona

Not a map member; added from the session notes. In home-context training the trainer sets out to show the family how to answer barking, biting and jumping, so first he gives Fiona the chance to be hyper with a ball and to wander, to see what she does. This time she just settles.

On film

And I also want to observe the behavior myself to see what she does. So I'm going to try to let her have an opportunity to be a little bit hyper.

Watch at 16:38 ↗

So I'm just going to let her wander and see if she gets herself into any trouble or if she just chills.

Watch at 17:07 ↗

Jack

Not a map member; added from the session notes. In an exposure session with a nervous dog the trainer speaks to camera about owners in general: being afraid to let the dog make mistakes and experiment is a sure way to fail.

On film

have it is to be afraid to make mistakes, to be afraid to experiment, be afraid to learn, afraid to try, and afraid to let your dog make mistakes, afraid to let your dog experiment, afraid to let them try. That's a certain failure.

Watch at 26:16 ↗

Josi

Not a map member; added from the session notes. After a recall in which the e-collar's contacts were not touching, the trainer tightens it and hopes Josi gets distracted again, so that he can proof the recall; he walks away and lets her sniff.

On film

and the dog feels nothing if it's not touching them all right so hopefully she does that again so i'll have an opportunity to proof it

Watch at 5:56 ↗

Kali

Not a map member; added from the session notes. At the park Kali fixes on a dog across the way, and the trainer uses it to proof the recall: he lets her choose, and she comes. At the first in-program lesson for Kali and Kane the full argument is given - you cannot punish something that did not happen - but that passage is on uncertain audio (watch 44:56) and is not quoted.

On film

prove recall though I'm gonna let her make a choice see what she chooses

Watch at 3:26 ↗

Kelly come good girl good job good girl you did good you made the right choices

Watch at 3:39 ↗

Kalypso

Not a map member; added from the session notes. Two sides of the same rule. On a heel walk the owner is told to let the leash hang and look for chances, because a right choice can be praised and a mistake is just as useful. On the place cot the trainer will not correct Kalypso for thinking about stepping off: he waits for the choice, then 'no, place' and social pressure guide her back, with no e-collar.

On film

thinking like a trainer you want her to make the mistake you want her to choose

Watch at 40:07 ↗

not preemptively correcting i'm letting her make the choice first right i could get on her case now because i think she's thinking about stepping off but don't preempt it let her make the choice first

Watch at 9:44 ↗

Louie

Not a map member; added from the session notes. At the in-person evaluation the trainer explains proofing to the family as failing somewhere new: a dog that gets away with a failure at one store fails at the next, while one that is corrected and shown the right thing succeeds faster at the next place.

On film

But if I fail and somebody corrects me for it and then shows me what to do to get the right thing, now when I go to the next place, I succeed faster.

Watch at 23:34 ↗

Lunna

Not a map member; added from the session notes. On a training walk the trainer welcomes Lunna's distraction, because it gives the recall something to beat; she comes and is sent off to play.

On film

Ooh, we just want her to get distracted so practice a recall.

Watch at 6:13 ↗

Max

Not a map member; added from the session notes. In Max's first session the trainer tells the owner that dogs learn fastest by doing the commands among distractions, so that when they fail, the failure can be corrected and then the success shown.

On film

Because the biggest thing with dogs is they do learn quickly, but the fastest way they learn is having to do the actions with a lot of distraction and with a lot of random things around so that they can learn that the boy, when they fail, then we can correct that failure.

Watch at 8:32 ↗

Max (Goldendoodle)

Not a map member; added from the session notes. In home-context training Kelton leaves Max on place and goes indoors as a test: if he stays, good; if he breaks, that is the chance to reinforce it.

On film

And that's gonna be a test for him because we're gonna see if he stays. and if he doesn't then I'll have a chance to reinforce it.

Watch at 26:52 ↗

Oscar

Not a map member; added from the session notes. In a Short, Oscar is put in a sit as dogs approach, and the trainer leaves the retractable unlocked: if he goes, he will be stopped.

On film

And I'm not locking the Flexi. If he chooses to go, he'll go, but I'll stop him.

Watch at 0:16 ↗

Penny

Not a map member; added from the session notes. Kelton passes ducks without locking the retractable or saying anything, because otherwise the owner is stuck doing that forever; if Penny goes for them, there is a consequence. He says this proofing phase is different from early on. On a trail walk he lets some sniffs go so that there will be chances to recall her later.

On film

So that's why I didn't block the flexi or tell her anything because there's an impulse to go block that flexi and tell her Penny, let's go because you see the ducks.

Watch at 32:58 ↗

doing anything about that currently, because I want her to continue sniffing. If I just recall her every time she even thinks about sniffing, then she's going to stop doing it. So I need to let some sniffs go if I want to have an opportunity to recall in the future.

Watch at 28:40 ↗

Peque (PK)

Not a map member; added from the session notes. In public training the trainers point out that they did not hold Peque (PK) back from an approaching dog before she did anything - she walked toward it and they said nothing - so the choice was hers. At her refresher program the trainer says the only way to get a dog not to act on a hang-up is to allow a choice.

On film

So, importantly, something I want you to notice here is that did we restrict PK's choices artificially before she did anything?

Watch at 1:21 ↗

freak out they need to not do it yeah and the only way to let that happen is you have to allow a choice

Watch at 21:07 ↗

Piglet

Not a map member; added from the session notes. Piglet's cat and chicken work is built on the rule. At the first in-program lesson the trainer says that managing the dog so she never has the chance means she never learns and can never be trusted; the handler becomes the one who decides the consequence. In later sessions the trainer wants her to try something so he can penalise it.

On film

better choice right but if i immediately manage and prevent her from even having the opportunity to make the choice then she never learns

Watch at 47:28 ↗

I'd like her to do something so I have a chance to penalize her for it like I did

Watch at 1:05 ↗

Rocky

Not a map member; added from the session notes. At an in-program lesson the family is told to make practice hard enough that Rocky breaks, take him back, try again, and reward success at once. At a follow-up the owner keeps locking the retractable in advance and is told to let him go and call him right at the edge of the temptation instead.

On film

the idea is you try to make it hard enough that he breaks and does the wrong thing and then you go and you make him go back and then you try again and then if he succeeds you immediately reward

Watch at 45:41 ↗

Instead, just let him go and see what he's going to do, and then it gives an opportunity to reinforce the recall

Watch at 37:12 ↗

Sage

Like Lola, Sage stays so close that she gives the trainer no chance to correct her for leaving the radius (on uncertain audio, he adds that he worries how she will be with her owners). In a later session she offers a down twice on a sit cue, and he lets her get it wrong both times so that the two positions become distinct.

On film

whenever, I don't know, and then I'll correct her for that, but she hasn't given me a chance

Watch at 9:04 ↗

She tried down twice, and I let her do it. She got it wrong both times.

Watch at 19:09 ↗

Thor

Not a map member; added from the session notes. Proofing a stay, the trainer says the distraction has to mean something, because the dog needs to make a mistake and get a prediction error. When Thor stays, fun follows. At his refresher pickup the owner is told to let Thor go to do it and only then tell him.

On film

Like, you throw a toy by them. You run by them. You need them to make a mistake. You need them to make a mistake. Because they need to try something and get a prediction error.

Watch at 2:27 ↗

make the choice you don't tell him anything yeah you let him go to do it and then you tell him

Watch at 12:00 ↗

Titan

Not a map member; added from the session notes. At Titan's first in-program lesson Connor gives him a moment to choose when he fixates, and tells the owner a dog cannot learn unless he is allowed to. At the second lesson he baits Titan into jumping so it can be answered - it sounds mean, he says, but it is part of training.

On film

choice and he chose yeah yeah well that's the other thing is i do make sure like you cannot have a good learning experience if you don't allow somebody to learn

Watch at 20:47 ↗

get a little high energy with him and try to like kind of get it it sounds really mean to like bait him into doing something but it is part of with the training you do have to do it

Watch at 14:06 ↗

Wiley

Not a map member; added from the session notes. In a down-duration drill the trainer raises the criteria hoping Wiley makes a mistake, and corrects it when he does. On a heel walk he wants another choice point - a runner from behind - but will wait for it to happen naturally rather than stage it.

On film

Hopefully he makes a mistake.

Watch at 6:42 ↗

and give him another choice point i'm not gonna like set him up or anything like that in a in a non-organic way

Watch at 7:14 ↗

Why it works

As the trainers state it: a consequence can only follow a choice the dog actually made (Reaper, Roxy's go-home), and a command only becomes reliable when there has been an opportunity to obligate it (Kelton with Lola). The moment of testing is when reinforcement carries information (Luna), and with some dogs a few such confrontations carried through move the dog a long way at once (Louie). If the error never happens in training it happens later, without the trainer (Zeus's owner lesson, Sputnik's go-home). Limits stated in the same files: running the error setup too often creates problems (Roxy), and failures that could hurt someone are staged where they can be contained (Kodak, Dutchess). Kelton frames 'errorless learning is fake' as his opinion. A wider reading of the session record adds the reverse case, stated at Piglet's first in-program lesson: if the handler always prevents the opportunity, the dog never learns and so can never be trusted. It also shows that choice points need not be staged when ordinary life supplies them - Wiley's trainer waits for a runner to pass rather than setting him up.

On film

punish it i can't punish it if he doesn't make it that's a mistake that a lot of people make

Watch at 47:48 ↗

off when I'm training they have to if they don't blow me off ever I can't make

Watch at 1:20:38 ↗

This is actually the most she's ever tested me, which is kind of a good Thing, because that means I can like reinforce it

Watch at 3:47 ↗

I hope that he at least a few times doesn't listen to you right now I'm hoping for it because that's the most instructive thing

Watch at 59:57 ↗

you don't need to do it too many times in a row otherwise you end up with problems

Watch at 27:47 ↗

Common owner mistakes

  • Managing the behaviour with the leash or your body so it never happens, which leaves nothing to answer.
    On film

    all right, like let me like manage her with the leash or with other things or with my body and like make sure it doesn't happen

    Watch at 13:08 ↗
  • Correcting before anything has happened - a look, or a passer-by who has not yet been barked at.
    On film

    punish it i can't punish it if he doesn't make it that's a mistake that a lot of people make they make um a big deal out of something like looking at another person

    Watch at 47:48 ↗

    next time what happens is you want him to choose whether or not he wants to bark or be nice

    Watch at 14:01 ↗
  • Distracting the dog with a stream of commands when a trigger appears, instead of letting her choose.
    On film

    estamos distraendo eso no funciona no funciona

    Watch at 10:03 ↗
  • Repeating the error setup over and over in a row.
    On film

    you don't need to do it too many times in a row otherwise you end up with problems

    Watch at 27:47 ↗

Questions owners ask

Isn't it kinder to prevent the mistake?

In daily life, managing is sometimes right - the trainers set a dog back from a door so the door can close on the attempt. But in training, prevention alone leaves nothing to teach: the trainers say they cannot correct a choice the dog never makes, and a dog that has never refused has never been tested. The Roxy household was told to stop managing the jump and let it happen, then praise or answer it.

On film

um a major correction because he needs to make the choice the bad choice in order for me to punish it

Watch at 47:41 ↗

If the door opens and they you can't like put them back. They were already gone.

Watch at 2:17 ↗

My dog is good all the time. Why would I want him to fail?

Because the failure will come later anyway, when nobody is ready for it. Zeus's trainer hoped the dog would disobey the owner in front of him so she could practise the answer; Sputnik's hoped for one dumb thing at handover, after being phoned the next day by other families.

On film

I hope that he at least a few times doesn't listen to you right now I'm hoping for it because that's the most instructive thing

Watch at 59:57 ↗

Please. Just do a dumb thing.

Watch at 28:21 ↗

yeah and so like what i would be hoping for is that she goes after the cat i want her to go after the cat right

Watch at 50:49 ↗

moments with him a lot of times with this kind of thing it's like you need like I'm searching for it i'm like i need to see it you need to actually do the behavior for me to be able to correct it

Watch at 1:25 ↗

Should I set up my dog to meet dogs he's fought with?

Not on your own. The trainers stage those meetings where professionals can contain them, and one moved a dog's introduction to the dogs she had problems with to a day with time to make it foolproof.

On film

I'm always happy like I want that to happen with us not at home yes you have to do it you have to try it

Watch at 12:08 ↗

sometimes when there's bigger stuff like that i like to make sure everything is really nice and tight that way like it's very foolproof

Watch at 9:40 ↗

How do I proof a recall if my dog always comes?

Make leaving likely. Finn's trainer says the dog must want to be somewhere else and must actually get away from you, or there is nothing to practise; Bo is let wander on a long line until he finds a smell, then called; Baxter's owner is told to find something he wants more than her.

On film

in order for me to proof it, I need him to be motivated to go somewhere else other than me and I need him to blow me off.

Watch at 7:29 ↗

I think I'm going to let him wander around a little bit now and get distracted so that I can do a recall with him.

Watch at 7:50 ↗

Did give me one more opportunity to flip the recall with the e-collar and like didn't come back right away

Watch at 0:13 ↗

The lesson

The goal

Create a real chance for the dog to make the wrong choice in training, answer it, and reward the right one - within safe limits.

Learn first

The drill

As run for Bo: switch to the long line, let it drag, let the dog wander until he finds something he wants, then call; answer a refusal and release a return. As run for Roxy: stand still where she looks like she wants to jump, give her the opportunity, and praise the sit if she chooses it. As run for King: leave the dog in a down beside the thing he most wants (the open truck) and see if he holds.

Watch out for

  • Blocking the behaviour so it never happens.
  • Correcting a look or an intention the dog has not acted on.
  • Repeating the setup too many times in a row.
  • Staging a dangerous failure without a professional.

Check yourself

When did your dog last have a real chance to get this wrong - and what happened when he did?

ONE TEAM. MORE WAYS TO UNDERSTAND THE WORK.

Explore Angier K9.

Published by the Angier K9 team. For your dog, start with a free consultation.

Find Angier K9 on YouTube, Instagram, TikTok, Facebook, Reddit, Thumbtack, X and Pinterest.