Happy, confident dog (and happy, confident handler!) performing reliable, precise, independent behaviors. (And to build this efficiently, without a lot of angst or wasted time!)
How do we get there?
By applying the science of learning and behavior through good training practices:
- Reinforce the behavior we want
- Don’t reinforce the behavior we DON’T want
I know you already know that part, but of course, there is a big difference between knowing the principle and knowing exactly how to implement it, and be ABLE to implement it. What exactly does that LOOK like? How do we make it happen? What do we need to be able to DO to make it happen?
A large part of what determines the success of a training session is dependent on how cleverly we set up that session so that it is easy for the dog to perform the correct behavior, hard for the dog to perform wrong behaviors, and easy for us (as trainers) reinforce what we want and reset when we get an error.
What are “good training practices”?
Good training practices mean following standard guidelines that are proven to be most efficient and effective at achieving the desired outcome. For our purposes, it’s everything about how we systematically set up a training session and mindfully interact with the dog.
These are the features of good training, and they apply to both you and your dog:
- Predictable, consistent movement
What we want to do is create a stable structure for learning that is independent of a specific method or technique. There are always going to be many different behaviors to train and each of those behaviors will have many possible methods for teaching them. We can make the learning easier on everyone by slotting it into a recognizable format that makes it easy to concentrate on the important stuff. Once you have established a format, it’s easy to layer in whatever behavior or exercise you are teaching.
Think of a book, for example. A book can contain many different types of content, but you pretty much know what to expect when you pick that book up – cover, title page, table of contents, the body of the material, index, and back cover. That basic structure is similar whether it’s a book about the history of Rome or how to sew a sock unicorn.
In a workbook, the format and structure make it clear when you are reading along and when you stop reading and fill in the blank with your own answer.
Use the words below to complete the sentence:
|Retrieve Pivot Jump Down|
- She’s standing with her hand on her belly button, so I should ________ into heel position.
- I should _________ the dumbbell she just threw.
- Time to lie ________ on my mat.
- On my way back, I need to _________ over the jump.
Whatever the subject of discussion, it’s the way the sentence and blank are laid out that make it clear when you should respond with a behavior (writing in a word). We want to structure our sessions to make it clear to the dog when he should respond by offering a behavior, and when to follow along.
It’s also extremely helpful to have a ready-made set of behaviors that the dog can offer to use as starting points for a new “sentence”.
In a training session, our movements and how we arrange the environment, including the room we are in, our own position, any props that may be involved, and when/how we reinforce provide that structure.
We must make sure that we are consistent within that structure so the dog knows exactly when he is expected to offer a behavior to earn reinforcement, when he should follow along with your movement, and when he should wait for a cue.
Clear boundaries to those expectations relieve a lot of the conflict and stress, because the dog isn’t left wondering “should I move now? How about now?” and minimizes the chance that you’ll accidentally incorporate undesired behavior into your session, as well as miss a chance to reinforce the desired behavior if it happens before you are ready. (It’s hard to decide which of those is more tragic!)
What makes a great training session?
To borrow from my friends Emelie and Eva:
Intensity and Seamlessness
It’s not about stressing out or rushing around like a chicken with your head cut off. It’s about having purpose and deliberate direction in every part of your training session. In the video of an ideal training session, we should be able to freeze the playback at any moment and say that what is happening in that moment is benefitting the purpose of the session. Each action, each movement in that session is moving us closer to our goal (or at the very least not moving us farther away from it!). The dog is either performing an approximation of the desired behavior, receiving reinforcement, or preparing for the next repetition.
That’s not always easy! But that is the picture we are striving for because we want to minimize gaps, which lead to uncertainty and stress, and prevent “junk” behaviors from creeping in on us.
This is where having a set of trained behaviors or “anchor patterns” as tools to support learning new behaviors becomes really helpful. We can use these behaviors/games/procedures/patterns to structure the training session to stay focused on the purpose and keep it moving efficiently in the right direction.
Great Training Session structure:
- Clear start
- Clean transport
- Clean behavior-reinforcement loops
- Clear end
What might that look like?
Station – Transport – Behavior – Reinforcement – Behavior – Reinforcement – Behavior – Reinforcement – Station
What does it NOT look like?
Wander into training yard – bark – circle handler – handler gets out food – behavior 1 – behavior 2 – reinforcement – behavior 2 – reinforcement – behavior 2 – reinforcement – handler looks for bag to reload – run around yard – poke – bark – bark – behavior 1 – behavior 2 – reinforcement…
Ok, so you can see how much cleaner the first example is compared to the second. Less stuff going on that doesn’t have anything to do with our training goals means a more efficient session in the same amount of time. We want that first one. None of us has unlimited time to get that training done and accomplish our goals, and even if we did, dogs just don’t live long enough (their only fault, really). So we want efficient.
Efficient session means everyone knows what they are doing when. There is no energy (mental or physical) being wasted in activity that doesn’t match our goal. And there are no gaps where time is wasted waiting for something to happen (unless that’s a deliberate part of your training plan, which is sometimes the case).
Here is an example of what that might look like:
Our anchor patterns give us tools to fill in the gaps and keep the session on course. They let us put the logistical parts of the session that aren’t the goal behavior on automatic, so that we can focus all our energy on the purpose of the training session, and make that bit easy.
Our anchors serve to fill three roles:
- Parking Spot – “wait here”
- Transport – “follow this”
- Reinforce-Reset – “after the treat…”
These behaviors should be so easy the dog can do it without thinking and without any overt cues from you. They should also be automatic for you, too! The point is to automate those logistical actions that are needed to structure the training session.
Think of it like a ski lift… The lift carries you to the top of the hill, no work, no decisions. At the top, you ski down (only one direction to go!), and then when you get to the bottom, you hop back on that lift and scoot on up. You only have to do work and use skills on the slope, practicing your skiing, and not wasting energy hiking back up.
I think of them as anchor patterns because they keep our training session from floating away (ship metaphor) or falling off the wall (hanging-picture-on-drywall metaphor).
Of course, the ironic part is that we’ll be spending this whole workshop thinking a great deal about these structural behaviors or anchor patterns… but with the eventual goal of not having to think about them again in the future.
The cool part is that if we do a good job building these tools, we’ll be able to think a lot less about them in the future AND the behaviors we are trying to shape will happen a lot more easily, too. Win-win!
The dog needs to know that the training session has begun and that reinforcement is available. We need a clear way to communicate “time to work!” so that the dog snaps into focused work-mode instantly and goes straight into behaviors that will benefit your purpose without having to wait around or beg for attention.
We also need a way to end the session, and even a way to hit the pause button briefly if we need to reload or move equipment (or just take a moment to breathe and regroup!). So for example, if you need to change a jump height, get more treats, or move a platform, positioning the dog on his mat during that time eliminates gaps in the session, and prevents accidental punishment by “disconnection”, or bad habits from developing.
My favorite way to make the start of the session very clear is to start the dog on a “station” of some kind. A station anchors the dog and makes a very clear boundary in space and time between working and not working.
A “station” often means a mat or dog bed, but it doesn’t have to.
The station could be a:
- Baby gate
Sometimes we may “station” a dog by placing them in a position like a down, but I do recommend having a target of some kind (like a mat) whenever possible, and a physical barrier (like a crate) is even better. The easier it is for the dog to tell the difference between “on the station” and “off the station”, and the less attention and brain power it takes for you to maintain that behavior, the more effectively it will function in training.
Because of that, a crate is often the simplest. A crate has walls and a door, so it’s easy to tell if the dog is in the crate or not. You can close the door until you are ready to begin the session, and not have to worry about the dog leaving the station when you aren’t looking or otherwise maintaining criteria on that station behavior. It’s not always convenient to drag a crate around, so a mat is a convenient alternative, but does take more training and attention. For our purposes, we will be working with both a crate and mat option. I train and use both with my own dogs, so we’ll practice both.
No matter what the station itself is, it is important that the dog offers to pop on/into the station voluntarily when it’s in proximity. You can ALSO have a cue to send the dog to the station, but it is important that the dog default to popping on the station without a cue when it’s right in front of him.
For the purposes of this workshop, we’ll be working exclusively with offered (ie. uncued) station behavior. In later workshops, we’ll address under what conditions we might want to use a cue for the station. But for right now, we want the dog to default to hopping on/in the station if it’s nearby.
Here is an example of using the mat to pause a session (in this case on cue), while I track down more treats.
For some training sessions, we’ll be able to station the dog right next to where we will be training the goal behavior. But most of the time, we’ll need to move the dog through space in some way to set him up to begin the behavior we are hoping to train.
We want to be able to move the dog through space in a way that:
- Doesn’t use up any of his (or your) mental energy
- Positions him at a good starting point to make that first rep likely to be successful
There are several ways to potentially transport a dog through space. You can use a toy, a trained hand in the collar, or a food lure. In this workshop, we’ll be focusing on the food lure variety, which we call a “treat magnet”. (Mainly because it’s easiest to talk about, but you can apply the principles with any other strategy in the future.)
We do want to train this behavior, even though it’s already a very easy one, because we don’t want to take anything for granted. There is a bit of skill and communication involved for both you and the dog.
For our purposes, the treat magnet behavior begins when you present the treat with your hand in a treat magnet posture, and ends when you release the treat into the dog’s mouth. We may use several consecutive treat magnets in a row, if we need to maintain it for a long distance or time. But in general, when the dog swallows the treat, that ends the transport and begins the next behavior.
The treat isn’t the end of the training session, and it certainly isn’t the end of the learning. We need to have a plan for what happens after that treat, because that is the third place where we are most likely to have a gap. Remember, the behavior cycle isn’t complete until the dog is back in position (and YOU are back in position) to do the next behavior. We need to be specific about what we want to happen, and how to get there.
This is going to vary a lot depending on exactly what you are training, but it is still something we need to account for. Otherwise, we risk the possibility of the end of the reinforcement becoming a cue for some unwanted behavior (checking out, running off with the toy, etc.)
The first question we need to ask is:
After getting to the reinforcer, is the dog in position to repeat the behavior?
If yes, great! Sometimes the dog can repeat the behavior again right away after swallowing the treat. For example, if you cue the dog to sit, click and drop a treat, and the dog stands up to eat the treat, then he is in position to perform “sit” again right away, and you don’t need to change anything.
If no, then we need a plan. What if the behavior leaves the dog on the landing side of a jump or other obstacle? Maybe he can turn around and go right back, or maybe we need to transport him back to the starting point on the other side of the obstacle.
It is extremely helpful to have a set of trained patterns to pull from depending on what makes sense for the behavior you are training.
Teach the game first, then add the behavior to the game.
For our purposes, we will be looking at 3 patterns that are helpful in training a variety of behaviors.
- Loop Around
In each case, you’ll have a defined program that you run as soon as the treat is delivered. Let’s look at each option.
This pattern is great for working with an object or obstacle where the dog can perform the behavior (or part of the behavior) in either direction. What is nice about this pattern, is that we are starting with movement – even before we add any props to the picture. So it almost completely eliminates the “waiting for the dog to interact with the object” problem.
I love this for shaping any kind of jump (that can be performed safely in both directions), tunnels, platforms, or introducing a new object or distraction (like a human). Once the pattern is learned (which happens VERY quickly), it’s easy to layer in obstacles or other elements. So much easier than trying to start with a stationary dog and shape movement in the face of an obstacle.
Here is an example of putting jump (with a landing side distraction) in the middle of the back-and-forth pattern.
This pattern mimics many obedience tasks that we have to train for where the dog must leave the handler, perform some task, and then return to the handler for reinforcement or the next cue.
Any retrieve follows this pattern:
- Run out to the object (dumbbell, glove, etc)
- Pick it up
- Return to the trainer
This pattern seems almost a given, and it really is a very easy one to teach. The way we usually train dogs, and the fact that we nearly always are physically holding the reinforcers, makes moving or at least turning toward the handler a very likely behavior to occur. I tend to orient toward the refrigerator whenever I enter the kitchen… that’s the way reinforcement history tends to work!
We can optimize this pattern though by training it deliberately. It’s great for any behavior where we want movement toward the handler to happen as part of that behavior. The pattern defines the path of movement, and we can layer in behaviors anywhere along that path.
This pattern is a little harder. It requires both you and the dog to flip between transport strategy and offering behavior and back again. This strategy is useful for times when you need to cover some space between the end of one behavior, and the beginning of the next.
The video above of the A-frame session is a good example. In that case, while the A-frame could be navigated either direction, the rest of my props (remote feeder and contact guide) were only on one side. So I needed a quick way to set him up on the starting side.
That example also included a wait in position, to allow me to arrange myself in MY starting position (since that was a big part of what we were training). That’s not required for every session, but you do want to decide in advance whether you want the dog to wait or to offer, depending on what your goals are for that session.
Here’s an example where I don’t have to move myself quite as far, and we are using exclusively offered behavior.
And with a different behavior, also offered, where I am not moving at all once the dog is ready to start the behavior.
To prepare for the workshop, get a jump start on these behaviors. The instructions *deliberately* have you introducing the earliest stages of the behavior. I want you to play with just these steps on your own between now and the workshop, and at that point we’ll build on them.
You’ll notice that all of the exercises (except treat magnet) have a similar rhythm, and are all making use of the same set of mechanical skills on your part. That’s also deliberate! (A method to my madness! muahahaha!) If it feels really similar, just with different backgrounds and props, then you are doing it right.
Each exercise should take you no more than 2-3 minutes TOPS!
- Parking Spot – We want to make sure your dog has at least some history of offering to get on/in a station before we add to it. (Worry about duration separately.)
- Offering crate
- Click as the dog steps into the crate
- Deliver 1 treat in the crate and immediately click and deliver a second treat out of the crate.
- Offering mat
- Click as the dog steps A SINGLE PAW onto the mat
- Deliver 1 treat on the mat, and immediately click and deliver a second treat off the mat
- Treat Magnet – We will rehearse the Treat Magnet game first without the dog to get your part right, then with the dog, to get his part right.
Hold the treat against the second joint on your fingers with your thumb. The dog should be pressing his nose and chin into the flat part of your hand.
- First stationary.
- Hold the treat in your thumb and second knuckle. Let the dog lick the treat, when you feel tongue, release the treat.
- Now add duration. Let the dog lick the treat for 1 Mississippi then release the treat.
- Now add movement.
- Let the dog lick the treat while you lure him to one side of your body, release the treat. Repeat in the other direction.
- Add steps.
- Add turns.
- Switch sides.
- Send-Return – This exercise can be done either standing or sitting. I recommend practicing with both!
- Drop a treat at your feet
- Click when the dog eats the treat
- Hand a second treat to the dog from your hand
- Drop another treat at your feet, clicking when the dog eats that treat.
- Toss the treat a bit farther (just a foot or two)
- Click when the dog eats the treat
- Deliver a second treat from your hand
- Toss another treat a little farther
- Back-and-Forth – Sit with a bowl on your right and a bowl on your left. You’ll want to be able to seamlessly drop a treat in either bowl, so keep the distances quite short.
- Without clicking
- Drop a treat in one bowl
- As soon as the dog eats the first treat, drop a treat in the second bowl
- Repeat, going from bowl to bowl.
- Make sure to return your hands to neutral (and still) between each treat.
- Add your click
- Click just as the dog is eating the first treat (before the head comes out of the bowl)
- Click just as the dog’s head is lifting out of the bowl, but before any feet move
- Focus on returning your hands to neutral between each treat.
And here’s the same game with a different dog, and slightly rougher session.