Transcript of Episode 222: Get Your Dog To Respond Faster: Latency In Dog Training

Speaker Key

SG Susan Garrett

Transcript

00:00:00

When you ask your dog to perform a behavior, does he do it like in lightning time or more of a leisurely pace? Or maybe some behaviors he can do quicker and others there is a little bit of a lag time?

Well in today's podcast episode we're going to talk about why you may have that lag time in your behaviors and how you can fix it, if in fact you do want to fix it.

00:00:37

Hi, I am Susan Garrett. Welcome to Shaped by Dog. What you're seeing in your dog's behavior if they're not responding to the cues quickly is something called latency.

Latency is the actual lag in the behavior between when the dog gets a cue and when they get around to doing it.

Now, when I design a training plan for my dogs, latency is built in, which means I want it pretty well instant. I cue you; you do something. And so, it's a criteria. And when the dog meets all the criteria of behavior, that's called a fluent behavior. So, the dog has been trained to fluency. 

00:01:21

Now, why might a dog have this long lag time with some behaviors? First of all, and the thing you should always consider is the dog in pain?

For example, my 12-and-a-half-year-old Border Collie Swagger, he started sitting slower and slower at the door and turns out he had some arthritis. So, we just stopped waiting for him to sit at the door to be released out. He just gets to go on a cue. He just stands and waits now instead of sitting and waits.

So, pain is something you always want to rule out, even with a young dog. There may be a reason why they're not responding in the way that you are expecting them to respond. 

00:01:58

Number two reason is, yeah, sorry, it's training. Chances are you have built that long latency into the behavior as a part of the training. Now, very likely you did it unconsciously and how would that happen?

It would happen if you weren't aware of the idea of latency to begin with, so, you didn't actually put any concern to what you were reinforcing, or more likely how quickly the dog did something that you were reinforcing.

So, truth is reinforcement builds behavior and if your dog is consistently a little more leisurely when they do a prompted behavior, it's likely been trained and it's definitely been reinforced. 

00:02:39

So, it could be pain, it could be dog training design, it could be behavior specific based on your environment. So, you might be in an environment that's completely overwhelming for the dog, so they might be a little bit afraid, worried, lack of confidence, there’s so many distractions. So that could be a reason. Or they're overstimulated by the environment.

For example, take a dog that loves to swim and ask them to sit right near the pond. They might not be able to do that. You might get a long latency, they might get super, super slow. And some people think, “Oh, it's a ‘don't wanna, don't hafta’ moment where the dog is just saying, no, I'm not going to sit. I'm too excited.” But in actual fact, dogs don't think that way.

00:03:21

It is a latency problem based on the stimulation of the environment. So, the goal for my dogs is that when I cue a behavior, I would like them to perform it as fast as possible for them.

And so how I get that is the conscious awareness of how fast that is in the training process, how I can use arousal states aka getting them more excited so that their focus is for the behavior.

How can I use arousal states to decrease the latency in the training and to be super aware of how long it does take and to not reinforce the longer ones?

So, what would people typically do that when the first time they see a dog that sits slower, they might physically correct the dog and that could add stress to the behavior, which means they're going to get even slower the next time possibly. At the very least, it means there'll be more corrections in that dog's future which, let's not go there. 

00:04:24

Number two, what people might do is they might help the dog with getting out a cookie lure or repeating the cue. Either one, you have just reinforced the lag time, you have just reinforced the dog for not sitting promptly or hand touching promptly or whatever the behavior is that they're doing slowly. So, you don't want to help them.

Number three, what they might do is they just wait them out and so what previously might've taken a half a second now takes five seconds.

Think of things like asking a dog in agility to sit at the start line when they're super excited or waiting for a dog if you've taught them focus forward, so to focus on work for obedience or for agility, and they are slowly going into that position and then you reinforce, you've actually built that weight into the behavior because reinforcement builds behavior. 

00:05:20

And if you have to wait for it the dog is telling you, “What you taught wasn't what I learned.” And so how are you going to fix this? You just have to go back and teach what you really wanted to teach. So, what are we going to do to fix this?

Number one, self-awareness. Always be aware of when latency is creeping in to your dog's behavior. Here's the good news, when you fix one slow behavior, dogs are pretty good at generalizing speed to other behaviors. But you need to have that self-awareness because here's what happens. 

00:05:57

Let's say it's a sit at the door. The dog is slow to sit at the door and you're focused on what you're doing next. “Oh, I'm going out to train.” So I'll wait and wait and wait and you finally get the sit, you go. You've just reinforced that slow drawn-out sit. And guess what, it's going to creep into other behaviors with that dog.

So, self-awareness is so important. That's point number one. Have the self-awareness to realize, “Oh, that got a little slow.” Break it off. Go on back in. Try it again.

00:06:30

And then you're going to write in your journal. Note to self: I need to work on latency with behavior X. Do not wait until it gets to be a chronic problem.

I look around and I say, “Why don't I get that latency issue that a lot of my students get?” I think the biggest reason is I am very aware the first time it happens, and I want to make sure that's the last time it happens. 

00:06:53

Number two, when we're fixing this latency problem is we're going to take the problem away from the behavior. In that if it is a slow sit in agility, we're not going to work it anywhere near agility equipment. If it is a slow sit at the front door, we're not going to work it near the front door.

We're going to go to another environment.

I love, love, love taking the problem away from the environment, training it in a new environment, getting it to the place I want, and then slowly bringing it back to the original environment. So, get that behavior out of the environment. 

00:07:29

Point number three, in this new environment, you are going to ask for ten repetitions of the behavior. And ideally, you're videoing this ten and you're going to reinforce unless they're painfully slow, you're going to reinforce all ten of them. If you get one that's really exceptionally slow, just break it off. In your head, what I would do is I would be counting.

Counting to see if I, “Oh, it took two seconds to do that one. Oh, it took three seconds. Oh, it took two seconds. Oh, it took one.” If you get one that takes ten seconds or five seconds, break it off. But what we're looking for is to find out what is the average of ten behaviors. 

00:08:06

What is the average of ten responses? What is the average latency when you cue a behavior in this new environment away from where you need it, how long does it take your dog to get it?

Now we have your average. And I would recommend you confirm that average time by going back and looking at your video, timing each of the ten. 

But now we know what the average latency is in this environment. Let's call it environment one. Now what we're going to do is we're going to do things in sets of five, and we're going to go to what I described in podcast episode 184, Jean Donaldson's Push, Stick, or Drop.

00:08:44

So, in five sets, we're going to know that you were either average or better.

For my first set of five, I am going to do reinforce average or better for that first set. Do I get to a push situation?

If I get to a push situation, meaning my dog was correct four of those five times got me average or better in my latency, I'm going to decrease that time of average or better by 10%.

So, whatever I got for those five, let's say now I got a two seconds. I'm going to drop that by 10%, so it's now going to be 1.8 seconds is what my average will be. I'm pushing them a little bit. Do those five repetitions, you're going to keep doing it until you get that behavior pretty darn fast.

00:09:32

You're going to repeat this several times a day. Maybe over the course of two days you're going to do it in the same behavior, the same way using push, stick, or drop, increasing by 10% every time.

By the end of your second day, you should have a really, really pretty darn fast response to the behavior.

And now what we're going to do is going to move to another environment or another ‘H’ as I call it, for our DASH, the H being the habitat.

We are moving it to another behavior, another five reps. You're going to keep doing this until the behavior has generalized in at least five different environments.

I mean it could be in the kitchen, and then in the living room, and then in the living room with like a bowl of cheezies on the table or something else that might distract the dog a little bit, and then back in the kitchen with another dog in the kitchen.

00:10:22

Whatever it is that's going to change the environment for you just a little bit, five more repetitions. When you get a fast response to your cue in every environment that you put that dog in, take the behavior back to your original environment.

So, if it's a sit at the door, put it back into in front of a door. I would put it in front of an interior door before I would go back to a front door. Just, it's similar, but it's still away from the original problem.

And when you get it fast at an interior door, you're going to go back to your original door. The moment your dog sits fast, you're going to say “search” and toss a cookie behind them. That allows them to eat the cookie, come back to the door, and show you another fast sit. 

00:11:06

I might do that three or four times and guess what, with Tater Salad, who is of all of the dogs the one that might give you a higher latency in a sit at the door, I will do this response with him I don't know, once a month. I'll just give him like four or five reinforcements before I allow him to sit and release him out the door.

So, the ultimate sit is the release out the door, which is a head scratcher to me. Because they really, really want to go out the door so why do they sit so slowly?

But once you reinforce that with the game of ‘search, come on, drive back, search, come on, drive back,’ you will find that latency on your sit at the door gets quite low. 

00:11:44

And here's the really cool thing. As I mentioned before, this latency generalizes really, really well to other behaviors. So, you're going to now go to another behavior that you know you've got a long lag time between the cue and the response, and you're going to work on the new one.

There's only one little asterisk I'm going to say with this.

The new behavior needs to be one that is trained to fluency.

Even the fluency of a slow response. You don't want to do this with a behavior that's still in process. Those behaviors, you're just going to go back to shaping and get that quick response with your cue. So, try this with one more behavior and one more behavior.

00:12:25 

And if you do this, I promise it will generalize to all the behaviors pretty fast likely after only two or three behaviors.

Give it a go. Let me know if it works the way I think it's going to work for you. Leave me a comment. Jump over to YouTube and while you're there hey, go ahead and give us a like on this video because that like goes a long way to letting other dog lovers like yourself know, Susan Garrett's got some pretty cool ideas about dog training. 

I'll see you next time right here on Shaped by Dog.