Wordless Weekend

Clicker Expo 2011 (Chicago): Helix Fairweather- Accelerating Success with Data Driven Training

There are two kinds of presentations at Clicker Expo: lectures and labs, which use attendee's dogs. I love this format, because it means that you get to see how a given idea works on a “real” dog (ie, not the presenter's). Helix Fairweather presented a lab on using data in training, and I really enjoyed the hands-on component. It took a topic that could have been dry and boring and transformed it into something fun and interesting.

This group- and dog- learned all about data-driven training.

 

Is your training data-driven? If you're like most people, probably not. Still, it's worth making the effort. Helix identified a number of reasons this is true: Data allows you to see progress in black and white: you know that your dog is doing better because the numbers prove it. Data helps you plan for your next session and can show you where there might be gaps in your training. Keeping data gives you a record that can help you learn how your dog learns.

So what data should you keep? There are two main things: rate of reinforcement, and success rates.

A sample training log. Click to embiggen.

Knowing your rate of reinforcement is especially helpful during the initial stages of shaping, because it lets you know if your criteria is too hard or too easy. Count out ten treats, then time how long it takes your dog to earn them. Divide by ten. For most tasks, the dog should be earning a click and treat every 6-10 seconds. Anything less means it's probably too easy, anything more means it's probably too hard.

Of course, these numbers may change if the behavior requires a lot of distance or duration, so repeat this step twice more. The average amount of time it takes the dog to earn ten treats should go down in each subsequent trial, because this indicates that the dog is learning. If it stays the same, or worse, goes up, try to figure out why.

During the lab, we all broke into groups. Each team would work with a dog on a given task three times, timing the rate of reinforcement. The conversations about why the numbers changed the way they did were very interesting. Everything from the way a treat got tossed to distractions in the room impacted those numbers. Figuring it out- and then compensating for it in the next set of trials- was almost like being a detective!

After you've attached a cue to the behavior, you can calculate how often the dog responds correctly to the cue. Count out ten treats. Cue the behavior. If the dog responds correctly, click and treat. If he doesn't, put one treat aside. Repeat until all ten treats are gone. Then look at how many treats you set aside and calculate your dog's success rate; each treat represents ten percent, so if you have two treats left, the incorrect response rate was 20%, and the success rate was 80%.

You can also keep track of how often your dog offers the behavior uncued. Ideally, your dog will only do the behavior when you've cued it, and keeping track of how often he does it without the cue will help you determine if you are achieving stimulus control or not.

No matter what you track, Helix said that it's a good idea to look at each data point separately as well as together. For example, it's a great idea to compare your success rate to your rate of reinforcement. A success rate of 100% with a really low rate of reinforcement is not good if you want quick, snappy responses to your cues. Similarly, a high rate of reinforcement with a high rate of uncued behaviors indicates that your dog doesn't quite understand the cue yet.

If you're thinking this is a lot of work, you're right! In the lab, we had teams of 4-5 people per dog to time, count, and evaluate what was going on. The punch line was that it's easy when there are lots of people to share the work, but not so easy when there's just you and the dog!

One of the best ways to track your data is by video recording your sessions. Doing so will allow you to concentrate fully on the dog, and not on counting or analyzing. Then, later on, you can go back and count and time things to calculate the various rates. It also allows you to see how things like treat delivery, body language or environmental distractions affects your dog. I have to admit, I hate watching myself train. I'm always really embarrassed by my sloppy handling skills, missed reinforcement opportunities, and unclear criteria. Still, I know that in the long run, this makes be a better trainer, too.

Helix also suggested using a voice recorder. That way you can talk as you're training (or immediately afterward) to summarize how things went. It's quicker and easier than writing during a session, although you will want to go back and transcribe those notes for future use.

I haven't actually started to keep data in training. I know I should, but I've been lazy about implementing it. I don't even have any excuses- at least not any good ones. But what about you guys? Do you take data, or record your training sessions in some way? What do find most helpful? What are your tricks and tips for getting the most out of it? If you don't take data, do you think it would help if you started? Why or why not?

Clicker Expo 2011 (Chicago): Cecilie Koste- Efficient (Clicker) Training

Ever feel like you're stuck in your training? Like you should be able to make more progress with your dog than you actually are? Afraid that you might be doing something to impede your dog's learning? Do you want to increase your skills with marking and reinforcing behaviors? Then settle in, because this is a summary of Norwegian clicker-trainer extraordinaire Cecilie Koste's presentation on how to harness the power of clicker training!

Cecilie talked about many things that good trainers do, but I think the secret lies with establishing a training routine. Cecilie has four stages in every training session: think- plan- do- pause (and think again). Before you does anything with your dog, you should think about your goals, both for the long-term and for the immediate session. What do you want to work on right now, in this moment? Once you've decided, it's time to plan the session. Planning the session is about more than just choosing the criteria; here are some things you'll want to think about:

Your Voice
I would venture a guess that Cecilie thinks one of the most important skills is to be quiet! She said that you should let your clicker, and your clicker alone, do the talking. The click is information, and when you talk during training, you blur the message.

This doesn't mean you can't praise your dog, though. In fact, Cecilie encourages praise; she said a good trainer is generous with reinforcement, and that you can increase the value of a treat by adding in praise, play or petting. I was glad to hear this- I have a very difficult time not talking to my dog. I just get so excited when she does something correctly! Cecilie just said that if you're going to talk, you should be sure to wait at a second or two after the click so you don't overshadow it.

Your Body Language
Being quiet applies to your body, too. A good trainer will think about how she is holding her body- especially her hands- because you don't want to confuse your dog with a lot of extraneous movement. I'm notoriously bad at this- I'm always using my body to prompt responses from Maisy! While this helps her in the short-term, I do think it hinders her from truly understanding the behavior in the long run. As Cecilie pointed out, it has caused her to be dependent on my body language. Worse yet, I'm not really conscious of what I'm doing, so I can't even replicate the body cues that Maisy is reading!

Similarly, be sure that you don't reach for a reinforcer until after the click. If you reach and click at the same time, your dog will probably pay more attention to your hands than your clicker. Some dogs will even interpret your hand movements as the marker! If this happens and you don't know it, your timing will be off, which could have a detrimental effect on the final behavior.

Your Teaching Method
Cecilie said you should shape, not lure, whenever possible. Shaping makes you a better trainer; it forces you to improve your observation and timing skills. It also makes your dog an active participant in the training. Cecilie wants the dog to initiate the session by offering a behavior, instead of the trainer having to prompt the behavior. Thus, you should think about how you could shape the behavior you want to teach, and lure only as a last resort.

Personally, I mix and match. Some behaviors are just easier to lure, and if I'm anything, it's a lazy trainer. While I appreciate that Cecilie's goal is a thinking, active dog, I certainly haven't had a problem with Maisy being willing to offer behaviors (and all of her initial training was taught through lure-reward). Still, I think it's wise to think about what method you'll use to teach a given behavior, and consider how it may impact the end result.

Your Reinforcement
Whether you are using play or food, you should always plan what the reinforcer will be and how you will present it. The way you reinforce the dog can reset him for the next repetition, can get his head or body in a certain position or facing a particular direction, and can even encourage proper chaining of behaviors. Play will increase arousal and put more intensity into the final product, while treats usually reduce arousal, and create a more restrained behavior. Decide what you want, and reward accordingly.

I try really hard to do this. For behaviors that require speed and enthusiasm, I will toss treats or throw her ball. For behaviors that require duration, I reward in place. I've definitely found that Maisy will anticipate the location of the reinforcer, and that I can use this to my advantage. It can also work to my disadvantage, so I know that I need to plan how to deliver that treat!


Now you're ready to train! The most important thing you can do during training is to be focused on your dog. Pretend that you're stepping into a bubble with your dog: all of your attention should be on him. During your training session, your dog should either be offering a behavior or being rewarded for a behavior. If you're distracted or not really paying attention, you will miss clickable moments, which might slow down your dog's learning.

It's okay to stop and take breaks. In fact, you should take breaks so you can evaluate your progress. Just be sure it's clear to your dog when he's supposed to be working, and when he's off duty! Cecilie recommended using a station, like a crate, mat or pedestal so your dog understands that he doesn't need to be offering a behavior. I really like this concept. I've tried using verbal cues to tell Maisy that she's on a break, but using a station seems to be so much more concrete and easy for her to understand. It also seems like this idea would help create an “off-switch,” especially if you used a crate as the station. This would especially be the case if you waited for signs of relaxation or self-control before doing another round of training.

Once you're done evaluating your progress, make a new plan, and train again! If you're planning well, you should see progress in each session. If you're not, then you need to figure out why that is. Are you using your voice or body in a way that interferes with your dog's learning? Is your reinforcement timing or placement affecting things? Is the teaching method slowing things down? Think about it, and then try again.

And that's the quick-and-dirty version of how Cecilie trains efficiently. Some of the material was familiar, and some of it was new to me. But what about you- can you identify any ways you slow down your dog's ability to learn things? Or maybe you do something awesome? Either way, please share your experiences! I'd love to hear what you do to ensure you're making progress towards your training goals!

Will Work for Cookies

Maisy is a complete food hound. She will go with anyone who has cookies, and will do just about anything to earn them. Case in point:

Maisy had never interacted with weave poles before, and yet there she is, happily going in and out of them, working her little heart out for the possibility of a cookie or two. The kidlet also sent her through tunnels, over jumps, and on a wobble board.

I really wish I had video of that last one- Maisy has a strong aversion to things that move under her feet- and yet she was completely willing to get on it in exchange for a cookie. She also promptly freaked out when it moved, prompting the kidlet to declare that Maisy “has issues.” Oh, honey. If only you knew how amazing it is that she did it for you at all.

Wordless Weekend

Who Are You, and What Have You Done with My Dog?

Maisy has improved so much that sometimes I barely recognize her. Last Sunday was one of those times.

Over the weekend, the Minnesota Mixed Breed Club held their semi-annual APDT rally trial. No, I didn't enter Maisy, but I did take the opportunity to expose Maisy to a trial environment. This isn't anything new; I have been working on gradually exposing her to busier environments, with the key word being gradual. Maisy's veterinary behaviorist has cautioned that I need to take things slowly so that I don't accidentally cause her to have a setback. I agreed, and I had every intention of taking Maisy to the trial, hanging around for a half-hour or so, and then going home.

Perhaps you've guessed that's not exactly what happened. In the end, we spent over three hours at the trial site, and she was content to be there. In fact, there were several times where I took her outside in order to give her a break, and instead of sniffing around, she strained and pulled in the direction of the building, clearly eager to go back inside.

She was very flirty. She said hi to every person she could get near with full on grins and helicopter tails. She even tried to say hi to a couple dogs, to the point where she rudely got in another dog's face (total handler fail on my part; I didn't see the dog under the table). When that dog (appropriately) snarked at her, Maisy jumped back, shook herself off, and was like, "Oh. Terribly sorry," and carried on with the business of flirting with people.

Maisy has (had?) two huge triggers: shepherdy looking dogs, and dogs running and playing. She used to be so bad when we did restrained recalls in class that she would become frantic even when we were out of sight in a back room. Just the sound of a handler cheerfully calling "come!" would send her over the edge. So imagine my surprise when she sat in my lap ringside, calm as could be, while a laekenois was sent running over a jump! She didn't even bat an eye when that dog came out of the ring and played tug!

I fed her plenty of treats, of course, and I was surprised by how soft her mouth was. Over the course of three hours, she never once took treats roughly, frantically, or with her teeth! This is pretty impressive since the first sign of stress she shows is a hard mouth. In fact, this is the most "stressed" she looked all weekend (and it's not stress at all, simply interest):

Despite all of this amazingness, I went home feeling a bit apprehensive. Maisy doesn't exactly have a history of recovering from stress well. In the past, she's been edgy and hyper-vigilant for several days after a trial. I expected to see some of that this week and... I didn't. Even on Tuesday night, when we went to our reactive dog class, she was fine. I expected barking and lunging and over-the-top theatrics, but it was business as usual.

So, look out, Minnesota: Maisy and I are back in business. We're ready to enter that ring and kick your butts... or at least have fun trying! Because Q or no Q, I will always be proud of my girl. Maisy has overcome adversity, and in doing so, she has proven herself to be a true champion.

Clicker Expo 2011 (Chicago): Cecilie Koste- Top OTCh: Skills for Top Obedience

Cecilie Koste is an amazing trainer. She's one of the top European obedience competitors, and the videos she showed were darn impressive. I enjoyed her presentation a lot, although like many of the sessions at Clicker Expo, I wish it had been longer. Talking about the skills you need to attain an OTCh in only 90 minutes is pretty much impossible, to the point that I kind of feel like her presentation was more about clicker training obedience skills for competition in general, and less about champion-level behaviors.

Without further ado, here are Cecilie's five steps for achieving great things in the obedience ring:

Step 1: Find the right puppy (and prepare it for training).
Declaring that good trainers deserve good dogs, Cecilie advised that we should take the time to find the right dog. After all, some dogs are better suited for competition than others. She didn't really say much about what the right dog is, which is too bad. While I understand this was not the focus of her talk, I would have loved to hear her perspective on what makes a good competition dog.

For those of you who already have the “wrong” dog for competition- it's okay! She said that any dog can be trained for obedience, it's simply more work. I'm not entirely sure I agree with this- a trial environment may be too stressful for many dogs. And while I certainly don't have a problem with picking a dog with a certain performance goal in mind, it bugs me just a little that this is listed as a step. It's certainly best to pick a dog that is suited to your life and desires, I also think there needs to be a spark- some relational compatibility. Maybe that's why she didn't talk about what the perfect puppy is- your perfect dog and my perfect dog are likely different, even if they both have the ability to attain high-level obedience titles.

Once you find the right puppy, you have three tasks. First, and most importantly, you need to socialize the heck out of that dog! Make sure he experiences as many different environments and as many different (nice) people as possible. (Interestingly, she didn't mention other dogs. I'm not sure if that was simply an oversight, or if she doesn't think that's as important.) Second, develop your reinforcers: Play. A lot. With as many different toys and objects as possible. And don't allow your puppy free access to food. Really work to build a reinforcing relationship with your dog. Finally, get your puppy used to rough handling. No, not too rough, but you don't want a sissy dog in the ring. This is especially important for clicker trainers since we tend to be hands off in training, and thus need to make a concerted effort to teach our dogs to accept handling.

Step 2: Teach the basic skills.
These are simple behaviors that form the building blocks of many different exercises. Cecilie called these the “letters” in the “doggie ABC.” Letters make up words, and words will make up sentences and paragraphs. Just as you can't write a book without knowing the letters, you can't achieve obedience titles without teaching your dog the basic skills.

The goal is for your dog to learn how to offer each of these basic skills without a cue, and without any help or luring from you. Once your dog knows them all, Cecilie said it's quite easy to teach him all of the obedience exercises you'll ever need because you simply wait for him to offer the behavior, then add the cue and develop them into exercises.

The basic skills you'll need are:
1. Look at you (focus)- Cecilie said your dog should be able to do this while standing in front of you, while you're walking away backwards, while in heel position, and with distractions.
2. Targeting- She recommended teaching a nose and paw target, and teaching both with duration.
3. Sit- The dog should be able to sit from a stand, while walking, and at a distance.
4. Rear end control- She teaches this mostly by teaching the dog to back up, but it can also be done with perches.
5. Sit at heel (finish)- She sends the dog to heel position from the front, back, left and right sides.
6. Gallop towards you- Your goal is a dog who runs at you with speed and enthusiasm.
7. Walk and look up (heeling)- Like most Scandinavian trainers, Cecilie recommends starting this by walking backwards and having the dog follow you, walking forward/toward you before teaching the dog to walk in heel position. (This site isn't Cecilie's but shows a good overview of the concept.)
8. Down- Like the sit, the dog should be able to down from a stand, while walking, and at a distance.
9. Stand- Cecilie teaches this while walking backwards away from the dog. The dog will simply stop walking and stay put. She also uses “reverse luring”- teaching the dog not to follow a food distraction.
10. Stay (remain in position)- The dog should stay in all positions: sit, down, and stand. Again, she uses reverse luring here.
11. Doggie zen- This is basically an uncued “leave it.” The dog should not chase or eat food he has not been told to take.
12. Hold- As in the dog holding objects in his mouth. The dog should be able to hold something while at heel, while sitting at front, while you lean over him, and while you touch the object.
13. Let go- As in, letting go of the object. I'm not quite sure how this can be an offered behavior if the dog is supposed to hold an object despite the distractions noted above.
14. Bark- The old “teach the dog to bark so you can teach him not to bark while working” idea, I think.
15. Jump- Including going away from the handler, towards the handler, and curving away (ie, a directed jump).
16. Scent discrimination- Cecilie recommended using duration targeting while teaching this so that the dog doesn't learn to depend on “tasting” the scents.
17. Tracking- This may not be needed, depending on your venue.
18. Go to person- This may not be needed, depending on your venue.

Now, if you're like me, you're probably thinking that sounds overwhelming. That is a lot of behaviors, and a lot of variations on each behavior, to have the dog offering uncued. Cecilie is adamant, though, that she wants voluntary, offered skills in order to fix the problems that surface during the course of trialing. She says it's easy to do this when you can go “back to basics.”

So how does she keep it from becoming chaotic? By utilizing something she called “the Windows principle,” which takes its name from the computer operating system. Basically, she thinks of each behavior as a file to be put in different “folders,” much like you would do on a computer. Each folder is based on the dog's location in relation to the handler. The dog will offer the behaviors that make sense based on context, and since there's only 4 or 5 behaviors he can do in each context, he'll quickly find the one you want. You'll click it, which will tell him to offer it again.

The basic folders are:
Folder 1- the dog is standing in front of the handler. His options are to sit, down, stand, or finish.
Folder 2- the handler is moving backwards. The dog's options are to sit, down, stand, or follow.
Folder 3- the dog is sitting in heel position. His options are to down, go out, or sit and look up.
Folder 4- object dependent. Targeting, jumping, or retrieving can only happen when targets, jumps or dumbbells are present.

Step 3: Perfect the separate parts of each exercise.
Once your dog knows all of the basic skills, you'll use the doggie ABCs to make words by developing each skill into parts of exercises. Split the complete exercises down into manageable parts- that's what you're working on here. Develop the maximum speed and precision you want before you name it, then add the cue and get stimulus control.

Step 4: Back chain the perfect parts into complete exercises.
Now take those words and make sentences by creating a chain. A behavior chain is two or more behaviors that are performed in a fixed order, with the reinforcer coming only after the last behavior in the chain. Back chaining works by teaching the last behavior first, which takes advantage of Premack and uses cues as reinforcers.

When you're doing this, if the dog doesn't perform one part of the chain, don't reinforce that mistake by giving the next cue. But don't be dismayed, either. Cecilie said it's good when your dog tests the behavior because that means he is figuring out exactly what you want. So, if your dog slows down in anticipation of the down cue in the drop on recall exercise, don't cue the drop! Go back to building up speed, and then reinforce the speed by giving the drop cue.

Step 5: Competition training.
The real challenge is to teach your dog to perform at a trial just as good as he does in training, which means your job is only half over. Your dog can (and will) mess up in a trial. This often happens because he doesn't recognize your cues. Maybe there are too many distractions for him to notice, or maybe he is dependent on something you've done in training. Or, maybe he's having an extinction burst, which shows up as increased variability in the behaviors performed. The problem is that bad behaviors are often reinforced in the ring because you allow him to continue the exercise anyway. If this happens just two or three times, you may end up with a ring-wise dog.

So what can you do to prepare for competition? First and foremost, don't compete before you're ready! Don't damage the behaviors and exercises you've spent so much time developing. Next, start training for competition in the beginning, while you're training the basic skills. You don't need to wait until you're training the full exercise to train with distractions like other dogs, an audience, a judge, etc. Then, begin working on a schedule of variable reinforcement, but do it gradually. And, as you decrease the quantity of reinforcement, increase the quality of reinforcement. Finally, Cecilie believes you shouldn't need warm ups, so you know you're ready for the ring when your dog can perform a perfectly backchained exercise on the required schedule of reinforcement on the first try.


As you can see, Cecilie spent most of her time talking about the basic skills needed and relatively little time talking about what she called “half the job”- polishing and proofing your dog for competition. I really feel like it's that polishing and proofing that will get you an OTCh, but I do appreciate that a solid foundation is necessary. I do think that Cecilie did a nice job of breaking down the exercises into the skills needed as well as the different variations on each skill. Just working on each of those little pieces- even if you put it on cue- would go a long way for most people, I think.

Still, despite her explanations, I continue to feel apprehensive about having all those behaviors being offered, not cued! I have so little on cue with Maisy that the idea of having even more behaviors thrown at me is a little scary. People who attended the lab said they felt the same way until they saw it in action with their own dog. They didn't expect that their dog would stop and offer a stand while they were walking away, they told me, and yet the dog did!

Anyway, it was an interesting session, and I definitely want to use some of the ideas. I can see where Maisy is missing some foundation skills (we definitely could work on more rear-end control, for example), and I definitely need to work on training around distractions more. I also love the idea of increasing quality of reinforcement when decreasing quantity, and will definitely implement that. (Maisy's excited about that.)

I'd love to hear your thoughts. Do you think Cecilie's basic skills cover all the bases for competition obedience? What would you add to or remove from the list? What do you think of having all those skills being offered, not cued? Does the "Windows principle" make any sense to you? Let me know!