What do I do when my dog ignores a command?

Stormie's handover is where the sequence is stated most completely and most usably, because it is being taught to a family who will have to run it without help. The rule comes with its exception attached in the same breath - the e-collar always comes after, never before, and there is exactly one time you may use it without a word, which is if she pulls on the leash, and only if you can time it to the instant she hits the end of it. It also comes with a fallback nobody else supplies: if she is still confused after the stim, give a gentle pull towards you rather than repeating. And the handover opens with the piece that has to happen before any of it works - you have to charge the word up, because a word that does not reliably predict anything is a word in a language she does not speak. Later in the same lesson the whole thing is run as a drill with the owner holding the remote, and it is re-taught on the spot when a press lands late.

Deuce is a ninety-six pound adolescent who does not know his own strength, and his file is where the rule is treated as non-negotiable and where its boundaries get drawn most precisely. The rule itself is absolute - the e-collar never comes before - and the reason given is that the dog has to be told which of two kinds of thing is arriving, which is why there are two marker words rather than one. The timing is given as a window: within about two seconds, and as close together as you can get without being simultaneous. Part 2 of the same visit supplies the other half from the dog's point of view, correcting an owner who pressed the button with nothing outstanding: there was no context, he did not hit the end of the leash, he did not blatantly disobey. Two other sessions bound the collar's role from the remaining sides. In one the trainer says he does not always mark with the collar - he marks with everything else except the collar. In another, doing leash-pressure conditioning, he deliberately leaves the marker out, because what he wants the dog to learn is that hitting the end of the line is itself the thing, with nothing in between. And on the day the marker is introduced he pays it every single time, because until then it is only a sound.

Dutchess has a bite history with dogs she lives with, and her lesson is where the rule stops being about obedience and becomes a safety rule. The trainer says he does not usually drill this, drills it anyway, and gives the reason in the plainest terms available: if you say nothing, panic, crank it up and hit the button - which does work - she concludes the shot came from the other dog, and the answer to that is to get bigger. The word is what tells her where it came from. He adds a ranking nobody else gives: of all the moments to remember to say it, this is the one. And then the lesson stops being a lecture, because two of the household's own dogs flare up in front of the owners and the whole thing is demonstrated live, with the extra criterion that the word and the collar have to be quick enough to read as one event.

Lola is a nervous dog, and her file makes the strongest version of the argument: the order does not just decide whether the correction is fair, it decides what the correction is. Fire the collar at a dog who is starting to fixate, with nothing said, and you build an association with people themselves. Say let's go first and then fire it, and it is a penalty for not listening. The trainer calls this word order and says outright that the two are very different worlds - and this is the error that makes a nervous dog worse rather than better. Upstream of it sits the attention law, taught in the first lesson: dogs pay attention to the first thing in the chain, which is why a dog who sits beautifully when you reach into your pocket does not know the word at all. Downstream sits the half-second - say it, wait, and only then apply pressure, because that gap is his chance to do it before feeling anything. And the go-home lesson supplies the vivid version of what happens when timing slips: a dog who decided a patch of grass was electric, and the observation underneath it, that the dog rather than the handler decides what counts as a punishment.

Zeus's lesson supplies the principle the rest of it hangs on, and the trainer's own phrase for it: learning has to be legible. Whatever you do or say first is what means something, so the cue comes first, then processing time, then the marker, then the consequence. The reason a consequence with no warning teaches nothing is given as an analogy any adult recognises - after you have been slapped you are not thinking about what the person said, you are thinking about being slapped. And the same rule catches the failure owners never notice: bend down and point before you say the word, do it every time, and the bending is the command. Say the word on its own later and the dog will simply look at you, because the word never meant anything. At the go-home lesson it is compressed into an instruction for the family: the word always right after, never right before, so it is you the dog is responding to.

Dolce is a protection dog whose commands are in German, and her first owner lesson is where the ladder is stated as three fixed steps with one absolute prohibition. Ask the command; one verbal correction if she ignores it; the e-collar if she ignores the correction. The prohibition is that the collar is never used to produce the behaviour - and the reason given answers the commonest objection to the tool: a dog corrected the other way round ends up only performing when she can feel it. The teaching device is counting to three with a child, told with an ice cream in it, and it works because the family already runs that system. A session six weeks earlier shows a junior handler running the same ladder without narrating it, which is what makes it a house method rather than one person's habit.

Bo's follow-up lesson gives the rule its clearest reason, and it is about attribution rather than fairness. The tool only ever comes after something the handler said - you tell him nope, or you tell him come and then nope if he does not come - and never on its own. Then the trainer holds up the remote and says the thing that makes it land: the dog does not know what this is, it does not exist to him, he has never looked at it. The only things that make any difference to him are the two people in the room. Which means the conclusion the dog draws afterwards is not that the collar got him. It is that he should have listened to mum.

Bullet's evaluation is where the ladder is made countable for owners, and where the reason the early rungs must be free is spelled out. Marker plus the command repeated is step one; putting him back in position is step two; the collar tap is step three, for the dog who blows off the warning outright. The analogy is counting to three with children, and the important half of it is the part about one: if I say one and he listens, no big deal. A separate session gives the concept's outer boundary in a small, exact scene - the trainer shakes a stranger's hand weakly because he has a finger on the remote, and explains that a stray press would be non-contingent, the dog would have no idea why, and what would go home with him is a dread of handshakes.

Sage's file gives the domestic version - the whole unit, front to back, as something a family rehearses. Say her name normally and then the command; if she does not do it, the marker and the command again, with a tap if you have to repeat; then okay, which the trainer spells out as O and then a K; then love on her. The trainer calls it the basic sequence for any command. Her go-home lesson adds the waiting: do not say it again straight away - count one Mississippi through four before repeating. And the rule is stated as an absolute in the in-program lesson: verbal first always, consequence after.

Luna's lesson is where the two marker words are contrasted live, and it is the clearest demonstration that the choice of word matters as much as the order. The soft marker was used with the collar as she came out of a sit, and it meant: you left the position, get back into it. Had he said nope in the same moment, the trainer explains, she would have concluded she may never sit down again. The same session shows the lever being chosen to suit the dog rather than the dial being raised - when she fought the sit, he locked the leash so she lost access to exploring, because exploring is what she actually wants.

Finn's go-home lesson is the exception in the record and it belongs on this page rather than buried in his file. He is a major flight risk, his recall is described as by far the most important command he has, and the family are told the opposite of the rule for that one word: do not press the e-collar after you have said come. Do not say come and then correct him for not coming. Say it, wait - one Mississippi, two Mississippi - and let the pressure attach to something other than the word itself. The passage where he says exactly what the collar should attach to is garbled in the transcript and only the sense survives; what is clean is the prohibition and the wait. The same stretch carries the other half of the recall rule, which is that when the dog finally does arrive you cannot be angry, and the trainer says he made that mistake himself a long time ago.

How long to wait after the command before doing anything about it.

Not addressed by anyone on camera, and the two may be measuring different gaps: Lola's half-second is between the command and the pressure, Sage's four Mississippi is between the command and repeating the command, and Deuce's window of about two seconds is between the marker and the collar rather than between the command and anything. But a family that has watched two of these lessons has been given two different numbers for what feels like the same question, and nobody reconciles them. King's file contains the reason it matters: a trainer reviewing his own footage found he was not leaving enough time between the command and the correction - see `correction-that-failed`.

Whether the collar may follow the recall word.

Both are said to families, in handovers, about dogs going home. The cases differ - Finn is a major flight risk for whom the recall is described as the single most important command by a long way, and the cost of a poisoned recall on that dog is not a training problem but a safety one - but no trainer on camera addresses the other position, and the sentence in Finn's lesson that would explain exactly what the collar should attach to instead is garbled in the transcript. An owner who has watched both lessons has been told two different things about the same word. See `dont-poison-the-command` and `never-punish-the-return`.

Whether the rule holds when the behaviour is serious.

These are close to compatible - one is about a dog-to-dog flare-up where the misattribution makes the fight worse, the other about stopping a dog injuring a person, where nothing matters except stopping it. But they are the two emergencies an owner is most likely to face, they are given opposite instructions, and no session puts them side by side.

Why can't I just press the button when he does something wrong?

Because the dog has to be able to tell where the feeling came from, and if nothing was said he will attach it to whatever he happened to be attending to. The trainers give three examples of that going wrong, all of them ordinary. A dog in a scuffle concludes the other dog did it, and escalates - which is how an e-collar can make a dog fight worse rather than better. A dog corrected at the wrong instant in a park decides that one patch of grass is electric and will not step on it again. And a trainer with a finger on the remote shakes a stranger's hand badly on purpose, because a stray press would be non-contingent, the dog would have no idea why, and what would go home is a dog who dreads handshakes. The word is what points at the cause. Without it, the correction still happens; it just teaches something nobody chose.

What exactly do I say, and when?

Say the command once, in a normal voice, with your body still. If he does not do it, say your marker word - the trainers use eh-eh, ah-ah, ant or nope depending on which of two things they mean - and say the command again. That is his free chance. If he ignores that, the collar. One trainer adds a fallback for the moment after: if she is still confused once the stim has happened, give a gentle pull towards you rather than repeating anything. And there is a fourth step families are often not told about: end the command with your release word - one trainer spells it out to a family as O and then a K - and then make a fuss of the dog. That last part is the whole reason the first part is worth doing.

How long do I wait before I do anything?

This is the one place the record does not agree with itself, and you should know that rather than be given a false certainty. One trainer, teaching the difference between what he calls forwards and backwards conditioning, says half a second to a second, and that the gap matters because it is the dog's opportunity to comply before feeling anything. Another, handing a dog back to her family, says do not repeat it straight away - count one Mississippi through four, and if she still has not done it, then tell her again. A third, on a recall, says wait a second, give him a minute, one Mississippi two Mississippi. Some of that is different gaps being measured - command to pressure, command to repeat, marker to collar - but not all of it. The safe reading is that the pause is real, that it exists for the dog's benefit rather than yours, and that being early is the commoner mistake: a trainer in this corpus caught himself doing exactly that by reviewing his own footage.

Won't this make my dog afraid of the word?

It will if you do it in the other order, and that is precisely what the rule exists to prevent. A trainer describes watching owners demonstrate their method - sit, and the stimulation in the same breath - and says what they have just taught the dog is that sit means you are getting punished. A dog who has learned that starts backing away from a command he used to know, and one file in this corpus puts the cost of undoing it at two and a half to four weeks. The order protects the word: the command is the question, the marker says you did not answer it, and the collar is what follows the marker rather than the command. It is also why the two marker words matter. Using the hard one as a dog leaves a position tells her the position was wrong; the soft one tells her to get back into it.

Is there ever a time I use the collar without saying anything?

The trainers name exactly one for owners, and they put a condition on it: a dog hitting the end of the leash, and only if you can time the press to the instant she hits it. The reason the exception exists is that the thing being taught there is not obedience to a word at all - it is that the end of the line is unpleasant - so a word in between would get in the way. A trainer doing that exercise says he deliberately does not prioritise markers for it. There is a second case in the record, from a lesson about a reactive dog: a full bark-and-lunge can be penalised on its own, because the behaviour itself is what is being answered, while a look that has not become anything gets a command first. Both are narrow. Neither is a general permission, and the same lesson that states the leash exception states the rule as an absolute in the same sentence.

My marker word doesn't do anything. He just ignores it.

Then it has not been charged, and the fix is mechanical rather than louder. A cue word means nothing until it reliably predicts what follows it - the analogy a trainer uses is a word in a language you do not speak, which is noise until you learn that it means something is about to happen. So for a while you follow it every single time, without exception, until the word alone produces the response. The same thing is done from the reward side when the marker is first introduced: every time the dog follows it, he gets paid. And the way a charged marker goes flat is by being said and not backed, over and over - see `correction-that-failed` for the version of that with a battery in it.

Pressing the button with nothing said and nothing outstanding, so the dog has no way to know what it was about.

Pairing the collar with the command itself - saying sit and stimulating in the same moment - which teaches the dog that the word is what the bad feeling is about.

Panicking during a scuffle and cranking the collar with no word, which teaches the dog that the other dog is the source of the pain.

Moving before speaking - reaching for the pocket, bending, pointing, stepping in - so the movement becomes the real command and the word never means anything.

Using the collar to make the behaviour happen rather than as the consequence of ignoring a correction, which produces a dog who only performs when he can feel it.

Reaching for the marker word before it has been charged, so it is a noise with nothing behind it.

Using the hard marker where the soft one belongs - saying nope as the dog leaves a position, which tells her the position itself was wrong.

He does the command when you reach for the treats, or when you bend down, and not when you just say the word.

He has started backing away or looking uneasy when you say a word he used to know.

He only listens when he is wearing the collar, and you have started to think of the collar as the thing that makes him behave.

Your dogs squabble and it has been getting worse rather than better since you started correcting it.

There is a particular place, object or person your dog now avoids and you cannot work out why.

Your marker word - the eh-eh or the nope - gets no reaction at all, because it has never been backed by anything.