Re: How to Train a Dog to Come when Called

[Michael D’Abruzzo]

Great post! It it has a lot of questions that I guess many other trainers would have that are new to the remote. I’ll break down the answer by your numbers:

1. First thing to mention here is that the level of correction is always relative to many factors: the individual dog’s pain tolerance, the situation, collar placement, fur thickness, etc, etc.. Depending on these factors, level 50 may be too low or inhumanely too high. Sometimes dogs respond to levels that are so low I can barely detect them on my fingertips – other times it is hard for me to fathom that they can basically “blow off” levels that I think would be unbearable to myself. Dog’s pain tolerance has a great range depending on individual and situation. For instance, if you were about to get into a bar brawl you may barely notice the level 50 yourself. This is why level 50 may be inhumane if you were just calling your dog from chasing a butterfly, but may be appropriate if he was chasing a dog and preparing for a possible fight. If it is too low you will know because the dog will not respond and if the level is too high you will know because the dog may yelp or otherwise show body language which reflects fear or stress (this can also happen on low levels if the dog does not understand).
The best way to avoid overcorrection while still reaching the minimum level needed to motivate the dog is by using an “escalating correction”. I do have plans of making a video on this concept, but for now I will explain it. What you do is start with the remote on an estimate (when in doubt go on low side) of the level for the current situation. If a correction is needed you pulsate (do not use continuous). As you pulsate turn the dial slow and steady at the same time. The moment that the dog responds you stop turning the dial. It is not necessary to watch the display since what matters is not the number. What matters is what the dog responds to, so you watch the dog. With practice (the remote not on the dog) you can get good with this with one hand. One client put a cut piece of rubber tubing around the dial to make it easier to turn using only his thumb. This works great and ultimately allows for lower corrections. The dog eventually realizes that the really low correction will only strengthen so they basically say “why bother challenging” if you are consistent. The extra beauty to this is that if you always say “no” before the correction consistently, the “no” becomes a conditioned punisher (in operant conditioning terms). What this means is that if you are consistent the word “no” should be the only correction that you ever need.

2.The remote works better than the pinch collar for aggression because it does not resemble a “bite” like the pinch collar and you can stay calmer while giving a less aggressive but more motivational correction, but just like any of the training collars you don’t want to use the remote in most cases to correct aggression – instead you use it to enforce obedience. So you can use it to enforce the command “come” when the dog is at a fence acting aggressively or running down the road in an aggressive manner. There are always limitations to the remote. For instance it may not work for a “come” command if the dog is actively in a fight – especially if the other dog is fighting back. The fear of the fight and not defending themselves may be more powerful than the thought to avoid a remote correction. If that is the case it is better to have a good plan for breaking up the fight if it gets to that point. There are other techniques for using the remote to break up a fight – but thats for another post…

3. I sort of answered half this question already above – as far as what level to use when the dog turns around is a judgement call. Generally if the dog turns quick I use the same level – if the dog is closer to me and the adrenaline seems to have subsided I turn it down. Again, the dog will let you know if the level is too high or too low. It is best to practice this in controlled situations so that you are prepared for an emergency. If you are consistent with discouraging the turns – they happen less over time.

4. The question about when to withdraw the remote and still have control over the dog during heavy distractions depends on a lot of factors. The biggest one is wether or not you also use positive reinforcement in your training. If you are not than it is similar to a clicker trainer asking when can i withdraw treats (or other rewards) from the training. If they did eventually the obedience will get worse. Using compulsion in training is similar. If you base your training mainly on the ecollar – once the obedience can not be enforced it will slowly unwind.
If you use a balanced approach (both reward and correction) you stack the odds in your favor that you don’t have to rely on the remote as much. But in reality you can not get around science. And the more you understand what I’m about to write here the better for maintaining a good recall (or any obedience):

If someone trains using only positive reinforcement the best way to have reliable behavior is using a variable ratio reward schedule. In layman’s terms this means that the dog gets a reward at random times. It is the slot machine method. A person still pulls the lever on a slot machine even when they dont get rewarded for unpredticable amounts of times because they know eventually it will spit out money.
Compared to a vending machine that always rewards when you enter money (a continuous reward ratio) – a person will stop putting in money much quicker when not rewarded (machine breaks). This is why if someone trains with positive reinforcement they make a plan so that they dont reward every time a dog responds – because if they are in a situation with no rewards the dog stops responding much faster. If they treated randomly the dog will still respond for a much longer time before giving up.

Training with an ecollar or any compulsion is opposite. When training purely with compulsion the best punishment schedule (sounds bad but that would be the scientific term) would be a continuous punishment schedule. This means that every time the dog did not respond to the obedience the dog always was punished (corrected). Compare this too speeding in a car. What if every time you broke the speed limit 100% of the time you got a ticket. This obviously doesnt happen in real life, but if it did – not only would people speed less they would get less tickets. The catch to this is that this would only work if the drivers new these super police were on the road that day.

Compare this to remote training. If you always enforce disobedience with a correction 100% of the time – it will be very rare that the dog disobeys. But the catch is, just like the super police on the road, the dog will need to know you have the ability to enforce the obedience with the ecollar. So by the dog wearing the collar – the dog doesn’t disobey and doesnt get corrected.

In reality most people are not this consistent and by the dog getting the chance to disobey on a variable ratio schedule (like the slot machine) it gives them a reason to challenge the collar more. This is like a child that throws tantrums to get what they want because it works one out of 5 times randomly so it is worth trying. The child who consistently gets punished for the tantrum doesnt throw tantrums.

Therefore, the most reliable obedience without having to use primary reinforcers (treats) or primary punishers (ecollar or other collar) comes from a combination of the best of both worlds. Rewarding the obedience on a variable ratio schedule and punishing the disobedience on a continuous punishment schedule. If you say “good boy/girl” before you treat and you say “no” before you correct you can have a dog obey very consistently without having to treat or correct with the ecollar around heavy distractions. But, the catch is that if you dont have some treats in your pocket to treat randomly or the collar on to remind the dog that you have the ability to enforce this ideal eventually unwinds. Those are just the facts of science and anything less than that is a compromise to the most reliable obedience.

A real life example are the guide dogs for the blind that are trained on the same city streets that I train my dogs on. The trainers use choke collars to correct the dogs for disobedience (they also treat). These dogs are also bred to be some of the most compliant dogs in the world. But, even after they are finished training they go to their new homes and lead around a blind person while still wearing the choker collar. Why? It is a reminder to the dog that they still can be corrected for disobedience. It is still very rare that you will see a trained guide dog be corrected with the choker collar once it is finished training – but part of that is because the dog knows that it can be corrected.

The problem with remote collars is that no one wants their dog to be seen wearing one. If a dog is trained with a choker collar it is different. A guide dog can be lead down the street wearing one and no one will notice – you can even compete in most obedience trials with your dog wearing a choker collar. Try that with an ecollar – will never happen. But basically the same concept – it is just another training collar.

I have a lot of youtube videos up. In most of my videos where the dogs are working under heavy distractions they are all wearing ecollars – yet in none of the videos do you see them get corrected. At most I say “no” and that is their correction. I don’t even need to hold the remote. It is in my pocket or side or whatever. I have an old bulldog that is 12 years old. And after i trained her with a remote I didnt have to correct her the rest of her life with it – but I would have her wear it if I was going to have her off-leash around distractions like in protection demos off-leash at pet fairs around other barking dogs. Still never had to use it.

Another example are invisible fence systems. When trained correctly the dog may get only one correction their whole life – then because the invisible fence is on autopilot and gives a warning 100% of the time that is backed up 100% of the time the dog goes the rest of their life not challenging the property line. The catch there is that the dog must wear the collar.

So sorry for the complicated answer but it is a kind of complicated question with a lot of twists – not to mention to factor in the individual dog’s temperament.

Hope this helps. If too confusing feel free to ask questions!