If At First You Don’t Succeed, Do Something Different
When my dog Remy was a puppy, many years ago, I set about teaching him a sit-stay.
As any trainer can attest, this is usually a fairly straightforward exercise. I was still in veterinary school at that time, so I had less training experience than I do now – but as a student assistant in OSU’s puppy kindergarten classes twice a week, I had taught dozens of pups the basics of how to sit and stay without any difficulty.
My typical “go-to” plan for teaching this skill is as follows: Ask the dog to sit. Give a hand signal and verbal cue to stay, and immediately reward before he has a chance to move. Praise and release.
Once this is going well, I can begin to pause briefly after giving the cue – then reward. First one second, then two, then three. Etc. I add distance very gradually once the dog is solid for several seconds with me standing in front.
For most dogs, this is a fun and effective way to build the beginnings of a good stay behavior. Done correctly, in the early stages, there’s almost no opportunity for the dog to make a mistake – he’s too busy being rewarded for being right. If he does happen to get distracted and pop up too early, I simply reset and try again. Most of the pups in my classes catch on very quickly, and have a good start on this skill by the end of the first session.
Easy-peasy, right?
Only, Remy wasn’t getting it.
At first, I thought he was just taking a bit longer than expected to catch on. I couldn’t seem to build any duration – the first few repetitions went all right, but once I began adding the pause, he had a tendency to get up before I had a chance to reward. I resolved to be quicker with my treat delivery, and redoubled my efforts.
The next day, I was hopeful that things would go better. Many dogs show improvement in a skill after having a chance to “sleep on it” and let the previous day’s lesson sink in, so I was eager to try again and check his progress.
I asked him to sit, and mentally made a note to go ahead and reward him right away on the first couple of reps to get him started on the right foot. I raised my hand, and said “stay!”…
…and watched, flabbergasted, as he immediately stood up and took a step backwards.
I reset him and tried again, to confirm it wasn’t a fluke. As before, the moment I gave the cue to stay, he stood and backed up.
Not quite the response I was looking for.
******
My training plan was a good one, and had a solid track record of working well with other dogs in the past. And yet – my pup was decidedly *not* learning to stay.
So what was the problem?
As I eventually worked out, there was something about my treat delivery – leaning forward and down, with my hand reaching quickly towards his face – that he found uncomfortable. The more I tried to be quicker with the reward, the worse things got. Remy is a sensitive guy. So for him, it was simply too much pressure in his personal space.
Of course, I didn’t know that then. But I could tell that what I was doing wasn’t working, so I scrapped that plan and decided to try something else.
Instead of rewarding him in place, which is usually my preference for teaching a stationary behavior like “stay,” I opted to use a clicker to mark the behavior – then tossed the treat a few feet away, so he could get up and collect it himself. This way, I could avoid leaning over him or reaching towards him at all.
I started with a simple sequence of sit -> click -> toss the treat, which immediately went much smoother than our previous efforts. Using the same technique, I was able to add some distance fairly quickly. First one step back, then click and reward. Then two steps, then three. I never went back to reward him in place (in contrast to my usual method!), but always clicked from a distance and then tossed the reward on the floor.
Within a few short sessions, I had a solid sit-stay with several steps of distance and a good 10-15 seconds duration. And more importantly, a happy and confident dog. 🙂
******
The lesson here isn’t especially complicated or profound, but it’s one that most of us (myself included!) still need to be reminded of occasionally. Sometimes, for whatever reason, your training plan doesn’t work. Even if it’s a good plan, and usually works with other dogs!
It doesn’t mean you’re a bad trainer, or that there’s something wrong with your pup. Dogs are just like people when it comes to learning new things – what makes perfect sense to one student, may be confusing or stressful for another.
The bottom line?
If you’re not making progress, try something else.
Don’t beat your head against the wall, trying to make Plan A work when it’s clear that things aren’t going well. Training is not a “one size fits all” proposition. There’s no shame in going back to the drawing board when needed.
Your dog will tell you, once you’ve got it right.
10 thoughts on “If At First You Don’t Succeed, Do Something Different”
This is great! I had rather a similar experience with my dog Summer. Her early training was at a dog obedience club. We used food, and by all appearances we taught “stay” in a positive manner. However, we were taught to use a very vigorous hand signal. Imagine a hand moving toward your face very fast and stopping an inch away. Summer did learn to stay, but it was never a happy behavior for her. I figured out years later why and retrained it, changing the verbal cue and dropping the hand in the face thing. She was much happier.
A seemingly small thing for the human was very large and unpleasant for the dog. I’m glad you wrote about this because it can really be hard to assess the problem sometimes.
“but it was never a happy behavior for her” – I love this – so important to notice! Sometimes, I find myself trying to “force” something, like shoving a square peg into a round hole. Sure, it’ll go – eventually. Probably. But there may be easier, smoother, and, above all, happier ways to get to the same place.
I think maybe it’s time to re-evaluate some of the techniques I’ve been using with my adult rescue Dane. Training has seemed to move so slowly with him. :/
Thanks, Eileen! I agree, Summer’s experience sounds very similar.
It’s easy to forget that even if we’re training something “positively” (i.e. with treats, and no physical corrections), we may still be doing things that our dogs find stressful.
Love this! There is more than one right answer to achieve the same outcome. 2+2=4 but so does 1+3 and 4+0.
I love that analogy Amy. I was explaining to a client this morning that there are many ways (R+) to teach each behavior and that my client’s two Italian Greyhounds may enjoying learning in different ways but we can get the same end behaviors. Thank you Dr Jen for keeping the happiness of our dogs at the forefront of our interactions with them!
So true! There are many paths to the same result. It’s our job as trainers to be able to come up with alternatives, if things aren’t going the way we want. 🙂
Hi There! I’m a trainer also. I’ve learned over the last 18 years to separate the verbal cue from the hand-signal. I want the dog to learn both. However, when people use both together, the dog will get very good at doing the behavior ONLY with the hand-signal. I suspect that’s because hand-signals are body language, and that’s how dogs communicate. So, I’ll teach the hand-signal first. Once the dog does well with that, I’ll teach the verbal by saying it, hesitating for a count of two, and then I give the hand-signal without speaking. It will take several reps, and then the dog begins to understand what the verbal cue means. You can tell that the dog is trying to figure it out because he’ll begin to move (or not move) until he finally does the verbal and gets a nice jackpot of treats! That’s what works for me. Using a verbal and a hand-signal at the same time is like speaking English and another foreign language at the same time. It all gets garbled.
Thanks, Renee! You’re absolutely right – this is a great explanation of overshadowing, in behavior science terms. 🙂 When we present two cues at the same time, the dog tends to pay more attention to the one that seems more relevant to them – usually a hand signal or some type of body language, vs. a verbal command. Hence, the verbal cue may be “overshadowed” and not be learned at all. This is a common source of frustration for many pet owners in my classes!
It is indeed far more effective to add the verbal cue later, in the way you’ve described here. (Verbal cue -> hand signal -> behavior -> reward.)
I love your analogy about speaking two different languages at the same time, making it hard for the learner to decipher what’s important.
This makes me wonder if there is a similar reason my dog has never enjoyed ‘fetch.’ She will bring back an item once or twice if a high-reward treat is involved, but obviously gets a bigger thrill from running away with the item (and chewing it to pieces, if at all possible). This bothers me because she is high energy and wants to run and play, but never picked up on that basic game.
Thanks Dr. Jen! As always your posts are so very helpful! I have used this same type of training with both of my dogs and both of my cats! Often times I find that the cats get what I am teaching them faster than the dogs! Hahaha! I find that as long as I use their name to get their attention first, then tell them or use the hand signal followed by the treat, they all learn the commands easily.
Comments are closed.