A friend invited me to write a guest article for her monthly musical freestyle column, so I decided to try it in the form of a list of Do's and Don'ts for motivating the freestyle dog. By motivation, I mean training an enthusiastic, high-energy, focused freestyle performance.
Here's the list I've come up with, culled from years of reading advice on this subject from experts:
DO set your dog up for success before you practice. If she's just eaten, or just finished a rousing play session, or just cuddled with you all afternoon, she's not hungry for those reinforcers. Motivation is all about hunger. Of course stay gentle and loving all day, but bring a dog to practice who is hungry for whatever it is you'll be offering as reinforcement.
DO train enthusiasm as a specific criterion. For example, if you asked for a weave and your dog did a spin, would you click and treat? Probably not. So when you train the weave, gradually stop reinforcing weaves that are physically correct but do not show the upbeat attitude you are looking for. Shaping enthusiasm is just like shaping any other behavior. Start by reinforcing anything that looks like the beginning of what you want, but gradually extinguish the weakest 20-30% of responses. As those weaker responses disappear, raise the criteria again, always reinforcing only the strongest 70-80%. As with any other criteria, only raise one criterion at a time, and relax other criteria at first when focusing on a particular one, in this case, enthusiasm. In other words, if you get an unusually high-energy version of a move, but that move isn't as accurate as you've been seeing, that's fine. That energy is exactly what you want, so be sure to let your dog know it with a reward. Once you've got that consistent attitude, you can go back and work on accuracy as needed. Also, don't be surprised if performance for that criterion occasionally goes the wrong way at times before turning back and reaching new heights.
DO retrain a move from scratch if you don't have an enthusiastic version of it now.
DO use the reinforcers that work for your dog. It might be nice if all dogs would work for praise, or if all dogs loved rough-housing, but that's not the case. It's simple to determine whether an intended reinforcer is helping you build enthusiasm from one rep to the next: After one enthusiastic response, are you seeing that enthusiasm again the next time you cue the behavior? If the answer is no, the dog is not being reinforced by whatever "reinforcer" you're using, whether you think she should be or not.
DO use high-value reinforcers exclusively. If your dog will work for kibble but isn't wild about it, use string cheese or tuna brownies or liverwurst instead, whatever it is that your dog loves. If your dog will play tug-of-war but isn't thrilled about it, use a chase toy or a scratch behind the ears or a cheer instead, again, whatever it is that your dog loves. Better still, combine reinforcers, for example using a cookie and a word of enthusiastic praise. If you want your dog to love freestyle and show it, make it great fun for her.
DO save your reinforcers for behavior you want to reinforce. For example, let's say the dog has just finished practice with a beautiful closing move, and now you and she are walking to the sidelines. Should you give her a cookie just as she completes that beautiful move, or should you wait until you're both halfway across the floor? If your answer is to wait, then expect great, enthusiastic exits and not so great moves.
DO give reinforcers in response to your dog's enthusiastic performance, rather than before she performs a routine or a particular move. You may occasionally want to let her know what scrumptious cookies you have in store for her — known as sampling — but once she knows how great her training sessions always are, the risks of dulling appetite or reinforcing the wrong behavior may be higher than any motivational advantage of sampling.
DO save your cheers for the behavior you want to reinforce. Let's say you've found your dog loves it when you say encouraging words, so you find yourself saying encouraging words throughout your practice sessions, even when your dog is not performing the way you want her to and the way you know she can. This is sometimes called cheerleading. Your intent is to cheer your dog on, but instead, your words of encouragement are actually reinforcing versions of behavior you do not want. "Oh, I see," thinks your dog, "when I move all pokey like this, I get sweet words from my best friend."
DO practice moves until they're fluent before combining them with other moves into sequences. For example, let's say that you'd like to use a sequence of moves A-B-C-D, but sometimes when you cue C, you don't get an enthusiastic response. When you see that happening, go back to training C by itself until you consistently get the response you want. Then gradually begin combining it with other moves, again re-training the individual move if the quality of the response declines.
DO use reinforcement after every correct response when training a new move. If the dog doesn't know what you want yet, and you nonetheless use intermittent reinforcement by reinforcing some correct responses and not others (for example, "two-fers"), how will she know which of her responses are correct? If you are consistent in reinforcing the behaviors you want, and in not reinforcing the behaviors you don't want, the dog will be able to determine what works to get what she wants.
DO consider switching to intermittent reinforcement when practicing a move that has become fluent. If introduced gradually and randomly, you may find that the sense of anticipation will increase your dog's focus and enthusiasm even higher.
DO maintain an appropriate rate of reinforcement. As you begin to introduce intermittent reinforcement, always watch to see whether you are continuing to get the responses you want. If not, you may be seeing ratio strain, where the rate of reinforcement is no longer high enough to maintain the behavior. If that happens, you may have dropped the rate of reinforcement too quickly. Go back to more frequent reinforcement, finding the level where your dog is showing you the responses you want.
DO surprise your dog constantly with good things. Whatever reinforcers you are using, whether food, play breaks, petting, praise, celebration, or anything else that you know from experience works with your dog, present those reinforcers at the most unpredictable times. Of course, a great time for a party is when your dog has shown a particularly good version of the behavior, but don't limit yourself that way. Use a variety of reinforcers, use them in varying quantities, and present them at completely different times from one practice session to another. A motivated dog is a dog who has no idea what wonderful thing is coming next, but knows from experience that it always might be around the next corner as long as she keeps up her end of the bargain.
DO keep your training sessions short. Of course, the term short depends on the dog, but if you want a high level of enthusiasm, you don't want your dog learning to pace herself. For some dogs, it's best to quit after a single great execution of a particular move. Others thrive on repetition since that means lots of cookies, as long as they don't become physically uncomfortable. Watch your dog and see what produces the best results. Experts generally suggest you err on the side of sessions that may seem too short to you. That single key often makes all the difference in motivation.
Well, there's my list of Do's and Don'ts for building motivation. Did you read it and think you're already doing most of those things, but you'd still like to see more enthusiasm in your dog's performance? That's quite normal. Somewhere on that list may be a single item that you're not using -- for example, maybe you're cheerleading, or maybe you've let the rate of reinforcement drop too low -- and that single item is undercutting all the other things you're doing right.
The good news is that your dog isn't lying to you. Her behavior is exactly reflecting your training technique. Read that reflection, believe her, and make the changes she needs you to make so she can learn to give you the enthusiasm you want.
As for the list, sorry there weren't any Don'ts. I hope you didn't miss them.
Sunday, April 8, 2007
Sunday, March 25, 2007
The Chance to Swim
Lumi and I often train near water, because I've found that nothing reinforces a behavior more for Lumi than the chance to swim afterwards.
Today, we worked on a variety of skills in a new location we've never worked before, with a pond giving us the special opportunity to set up a water retrieve as the reinforcer for correct responses.
TRACKING
I started the session by leaving Lumi in the van and laying a track for her, about 120 paces. Besides meandering a great deal, the track also contained some challenges: it went over a pile of rocks, through underbrush, and across and then down an old gravel road, in addition to stretches of weedy grass. I baited most of the track, but had a few stretches as much as ten paces, always on curves, with no bait.
Lumi did fine with this track, and gave a good down indication when she got to my wallet. I'm not sure if either of us learned anything new, but we both had good practice. Lumi continued to learn persistence in the face of challenges, and I continued to learn what Lumi looks like when she's off-scent. I feel that's a key skill for me, so that I'll be able to recognize it in case it happens someday when we're testing on a track where I don't know where the turns are.
After the track, we ran to the pond so that Lumi could swim a little before we started our field exercises.
RETRIEVING IN A "T"
For our first field training, we worked on whistle sits, backs, and overs using a "T" formation. I set it up so that as Lumi was completing the retrieve, I was running away from her and toward the pond. She would race to catch me and give me the dummy. Then I'd send her into the pond, cue "left", "right", or "go out" (straight), and throw the dummy to reinforce her response. When she'd return to me, she'd swing to left heel, I'd take the dummy, cue "shake", then toss the dummy on the grass or offer it for tug or both, and finally toss it near the pile of "used" dummies so we could head for the next rep of what we were working on.
That game may sound simple, but for Lumi, it seems to be an exciting sequence. As a result, she can hardly wait for the chance to make her next retrieve in the "T", knowing what's about to follow.
We set the "T" up with three dummies at left, three at right, and four straight ahead. The whistle sit was always in the middle, of course, and we used "back" spinning half the time to the left, half the time to the right. I tried to make the sequence as random as possible. In previous sessions, Lumi has begun to flag after a smaller number of dummies, but today, she was high the entire time. I'm pretty sure it was the swimming at the end that made the difference.
CASTING
Next, we worked on casting, that is, straight retrieves out and in, rather than bowed paths. We worked both in the water, and also across a little inlet of the pond with Lumi swimming across the inlet, retrieving the dummy on the land on the other side, and then swimming back. Because the inlet wasn't very wide, she could get to the dummy faster by running around on shore, so the challenge was to train her to take the straight path.
We made progress both in the water and across the inlet, and since we've been only working on casting for a few sessions, I'm pleased with where we are on this. At our last field training day, I had the opportunity to see a very experienced Lab doing this kind of thing, so it's given us something to aim for.
LONG RETRIEVES
Next we did a number of long retrieves. Since I wasn't familiar with the area yet, I had trouble finding a good place to set up a 200+ yard retrieve, so we started with two 100 yarders, one through the woods, and one over grass and then across an old gravel road. Lumi always seems most motivated by retrieving through woods, so I try to set those up when possible, but it's harder to find a place to set up long, straight distances.
At last, we walked about 200 yards down a mostly straight section of the gravel road, and where it curved, I threw Lumi's dummy straight ahead into the meadowy area in front of us. Then we walked back to the starting point and I sent Lumi to the retrieve. After a few seconds, she disappeared around a bend, and then she was gone from sight. But judging by the interval, she must have been at a full run the entire time, because soon she came tearing back around the bend with her dummy in her mouth, and I turned away and ran to the pon. She caught me before I got there, and we ran the rest of the way together. Then I sent her out into the water and threw the dummy out beyond her, leading her straight out into the middle.
Great fun, and a great way to end the session at our new pond.
Today, we worked on a variety of skills in a new location we've never worked before, with a pond giving us the special opportunity to set up a water retrieve as the reinforcer for correct responses.
TRACKING
I started the session by leaving Lumi in the van and laying a track for her, about 120 paces. Besides meandering a great deal, the track also contained some challenges: it went over a pile of rocks, through underbrush, and across and then down an old gravel road, in addition to stretches of weedy grass. I baited most of the track, but had a few stretches as much as ten paces, always on curves, with no bait.
Lumi did fine with this track, and gave a good down indication when she got to my wallet. I'm not sure if either of us learned anything new, but we both had good practice. Lumi continued to learn persistence in the face of challenges, and I continued to learn what Lumi looks like when she's off-scent. I feel that's a key skill for me, so that I'll be able to recognize it in case it happens someday when we're testing on a track where I don't know where the turns are.
After the track, we ran to the pond so that Lumi could swim a little before we started our field exercises.
RETRIEVING IN A "T"
For our first field training, we worked on whistle sits, backs, and overs using a "T" formation. I set it up so that as Lumi was completing the retrieve, I was running away from her and toward the pond. She would race to catch me and give me the dummy. Then I'd send her into the pond, cue "left", "right", or "go out" (straight), and throw the dummy to reinforce her response. When she'd return to me, she'd swing to left heel, I'd take the dummy, cue "shake", then toss the dummy on the grass or offer it for tug or both, and finally toss it near the pile of "used" dummies so we could head for the next rep of what we were working on.
That game may sound simple, but for Lumi, it seems to be an exciting sequence. As a result, she can hardly wait for the chance to make her next retrieve in the "T", knowing what's about to follow.
We set the "T" up with three dummies at left, three at right, and four straight ahead. The whistle sit was always in the middle, of course, and we used "back" spinning half the time to the left, half the time to the right. I tried to make the sequence as random as possible. In previous sessions, Lumi has begun to flag after a smaller number of dummies, but today, she was high the entire time. I'm pretty sure it was the swimming at the end that made the difference.
CASTING
Next, we worked on casting, that is, straight retrieves out and in, rather than bowed paths. We worked both in the water, and also across a little inlet of the pond with Lumi swimming across the inlet, retrieving the dummy on the land on the other side, and then swimming back. Because the inlet wasn't very wide, she could get to the dummy faster by running around on shore, so the challenge was to train her to take the straight path.
We made progress both in the water and across the inlet, and since we've been only working on casting for a few sessions, I'm pleased with where we are on this. At our last field training day, I had the opportunity to see a very experienced Lab doing this kind of thing, so it's given us something to aim for.
LONG RETRIEVES
Next we did a number of long retrieves. Since I wasn't familiar with the area yet, I had trouble finding a good place to set up a 200+ yard retrieve, so we started with two 100 yarders, one through the woods, and one over grass and then across an old gravel road. Lumi always seems most motivated by retrieving through woods, so I try to set those up when possible, but it's harder to find a place to set up long, straight distances.
At last, we walked about 200 yards down a mostly straight section of the gravel road, and where it curved, I threw Lumi's dummy straight ahead into the meadowy area in front of us. Then we walked back to the starting point and I sent Lumi to the retrieve. After a few seconds, she disappeared around a bend, and then she was gone from sight. But judging by the interval, she must have been at a full run the entire time, because soon she came tearing back around the bend with her dummy in her mouth, and I turned away and ran to the pon. She caught me before I got there, and we ran the rest of the way together. Then I sent her out into the water and threw the dummy out beyond her, leading her straight out into the middle.
Great fun, and a great way to end the session at our new pond.
Monday, March 19, 2007
Rust
Based on a diagnosis last November of arthritis in Lumi's hips and left wrist, we stopped doing agility entirely at that time. I had considered Lumi to be retired from the sport.
But at our last visit to Carol Lundquist, our holistic vet, she said that Lumi does not necessarily need to stop agility entirely. She suggested that we introduce one obstacle per month, similar to a food allergy test, watching to see whether a particular obstacle shows a return of Lumi's wrist pain. For this month's obstacle, I chose the weave poles.
Because of weather and competing priorities, we've only had time for two sessions so far. Yet I've been surprised by how little Lumi's long lay-off has affected her performance. Yesterday with our steel-based weaves, Lumi ran 12 poles in 2.85 seconds.
Though she's had faster times at lakeside with swimming as the reward, that's one of the fastest times Lumi ever had in our training field, where she's working for her tennis ball.
It seems that if there was any rust after four months, it flaked off quickly.
But at our last visit to Carol Lundquist, our holistic vet, she said that Lumi does not necessarily need to stop agility entirely. She suggested that we introduce one obstacle per month, similar to a food allergy test, watching to see whether a particular obstacle shows a return of Lumi's wrist pain. For this month's obstacle, I chose the weave poles.
Because of weather and competing priorities, we've only had time for two sessions so far. Yet I've been surprised by how little Lumi's long lay-off has affected her performance. Yesterday with our steel-based weaves, Lumi ran 12 poles in 2.85 seconds.
Though she's had faster times at lakeside with swimming as the reward, that's one of the fastest times Lumi ever had in our training field, where she's working for her tennis ball.
It seems that if there was any rust after four months, it flaked off quickly.
Thursday, March 15, 2007
A Gathering in the Woods
After several posts on the DogTrek list on the subject of desensitizing Lumi to the sound of gunfire, Amy Cook joined in with what she called her "unsolicited two cents", which of course was really very much welcome. Her suggestions gave me some assurance for the first time that Lumi could actually become comfortable with gunfire, provided I went about the training properly.
The key that Amy emphasized over and above all others was sequence: Gunfire predicts good things, not the other way around.
Amy, of course, was not the only one on DogTrek offering helpful insights this training. But for some reason, the easy authority with which she explained what we needed to do was especially helpful. This really came through when I posted about an unsuccessful attempt to apply her advice yesterday, and she replied with yet another clear, relaxed post that set aside my concerns and again showed the path we needed to follow.
So this morning, Lumi and I went out for another session, armed with a crystal ball provided by Amy that showed success in our future. Confidence is a great thing.
Based on what little I had learned from previous sessions, and guided by Amy's suggestions, here's how today's session went. For each point, I've also added my subjective evaluation about what mattered with respect to our D&CC training, and what did not:
The key that Amy emphasized over and above all others was sequence: Gunfire predicts good things, not the other way around.
Amy, of course, was not the only one on DogTrek offering helpful insights this training. But for some reason, the easy authority with which she explained what we needed to do was especially helpful. This really came through when I posted about an unsuccessful attempt to apply her advice yesterday, and she replied with yet another clear, relaxed post that set aside my concerns and again showed the path we needed to follow.
So this morning, Lumi and I went out for another session, armed with a crystal ball provided by Amy that showed success in our future. Confidence is a great thing.
Based on what little I had learned from previous sessions, and guided by Amy's suggestions, here's how today's session went. For each point, I've also added my subjective evaluation about what mattered with respect to our D&CC training, and what did not:
- We went to a location where hunting is permitted, rather than continuing to hope that no one would hear us and report us at one of the other places we had tried. Our new location happened to be a muddy and heavily wooded area surrounding a "river", actually no more than a stream. Do I think the location mattered? Not really, but I guess it's possible Lumi was more comfortable in the woods than some of the more open areas we had tried before.
- Lumi and I had a lot of company. Dave and his girlfriend Heather, Brookie the Cookie, and Dave's MinPin Athena all took one route through the woods, while Lumi tethered to my belt and Gabriel on a Flexi took another. Do I think all that company mattered? Absolutely. Gunshot after gunshot, Lumi could not help but notice that no one else was bothered by the sound in the least. The more people and dogs showing no reaction, the more normal no reaction must have seemed.
- I didn't bother to bring a toy this time. Instead, I came armed with a variety of treats, especially several sticks of string cheese (mozzarella). Do I think the choice of food versus play, as what the gunfire would predict, mattered? It was crucial. There's no way Lumi would have played after the gunfire during this session, at least in the early going, and even a lower value treat would have failed to interest her. But Lumi loves cheese and this meant that I was able to feed her within a second after each shot, about three little bites of string cheese. I also fed Gabriel, not because he needed it, but to show Lumi that if she didn't eat the cheese, someone else was available to eat her share. I don't know whether that mattered at all, but I don't think it hurt.
- Another reason that I'm glad I tried food rather than toys is because play is about arousal, while eating is about calmness. I think Lumi needed calmness, not arousal, at this stage. Do I think feeding Lumi after the gunfire, rather than before or during, mattered? Based on my observations today, I couldn't say. But based on my faith in Amy's experience in these matters, I have no doubt of it.
- I kept Lumi on leash the entire time. In our previous attempts, she was either on a long line or off lead, and I wish we could have done this work today with Lumi off lead. But I needed her to stay with me, and it would have been too much work, and unpleasant for both of us, if I had to keep calling her back to me. Do I think having Lumi on lead mattered? Without question. It made things so much simpler and easier, especially the fact that it was a short lead. It took the possibility of flight or any other independent action out of the equation, and allowed Lumi to settle in with me the entire time.
- Lumi and I were separated from the gunfire by about 250 yards at first. Not as far as earlier experiments, and with less ambient noise, but I wanted to try it because the woods were so thick and the ground was so muddy. Do I think the distance mattered? Yes, but remarkably enough, in the inverse! After the first couple of reps, I tried moving us about 25 yards closer to the others for the next rep, and to my surprise, Lumi seemed more comfortable at that new distance. After that, we moved about the same distance closer after each rep, and the closer we got, the more comfortable Lumi became. The last time Dave fired the pistol, Lumi and I were standing only about 20 feet away. Lumi started at the gunfire, but immediately welcomed her pieces of cheese and made no effort to move away from the direction of the gunfire. As with all the other reps, Lumi tried harder than normal to press in close to me immediately after the shot, but aside from that seemed her normal self, alert and energetic the entire session.
- I didn't keep Gabriel with Lumi and me for all the conditioning. The closer we got to our other family members, the more that both Gabriel and, yes, Lumi too were pulling on their leads toward the clan, gun and all. At last I called Cookie on the walkie-talkie to come and take Gabriel, so that for the last few reps, Lumi and I were in our own little group and everyone else was with the gun. Do I think it mattered that Gabriel was with Lumi's and my group at first? I don't think so, but I'm not sure. It certainly didn't seem to matter after a few reps, though. If anything, I think it helped to normalize our work to have Gabriel off in the distance with the rest of our gathering, completely at ease and waiting for us to join them.
Sunday, March 11, 2007
Pistols and Labs
Yesterday was our second visit to a field training day for gundogs. Lumi made some progress in our D&CC work for gunfire, but she needs a lot of sessions, not just two, and that place isn't doing any more training days this year.
TRAINING PISTOLS
Although Lumi already has some gundog instincts and trained skills, her noise sensitivity trumps everything else. In the big picture of gundog training, it's a small matter, but for Lumi and me, it's the top priority, the only priority, because nothing else matters if we can't solve it.
So I've put a lot of thought into how to work on gunfire D&CC. Yet it didn't occur to me until today that I might be able to order a pistol that shoots blanks online.
Sure enough, you can buy them from online stores for gundog supplies, and other places as well. I found one that was pretty inexpensive: about $22. It should be here in a few days.
Then all I need is someone to stand about a mile away and fire it for me while I play with Lumi. You'd think it would be easy to find someone to do that, wouldn't you? We'll see.
WATCHING THE LABS
Meanwhile, as suggested to me by private correspondence, I didn't spend all my time at yesterday's training day working with Lumi. At one point, I put her in the van and went off to watch the pros working with their dogs, or their clients' dogs, for awhile. I don't know why the difference between marks and blinds was so confusing to me before, there's really nothing to it once you've watched for a few minutes.
Watching serious field-trained Labs was an education in many ways. It goes without saying that they are nothing like your typical neighborhood Lab. One of the yellows looked like a body builder, muscles bulging up down his legs. He also had no hair around his elbows, perhaps from hunting hazards? I didn't ask why.
The skill of one of the black Labs in particular was also amazing, cutting off small triangles of water hazards to run a perfectly straight mark. As he approached each pond, he flew into the air and leapt in wildly. What drive!
Lumi, of course, will never look or act like one of those FC Labs, but she still looks athletic as Goldens go. I saw other Goldens there, and none of them had Lumi's lean, muscular physique. On the other hand, none of them had her fear of gunfire, either.
MORE PHEASANT WORK
We have so many things we could work on these days, but yesterday's field training inspired me to work with Lumi on field work some more today.
We still have two dead birds from last week to train with. They won't last forever. As I understand it, we can eventually look forward to maggots, red ants, and other niceties. I don't know where we'll get more birds, but at least we have these two for now.
We were training with our pheasant today at the lake for about half an hour. I'd throw Lumi's dummy into the lake, and then when she brought it to me, happily swinging into heel position, I'd reinforce that by throwing the pheasant for her up onto the hillside, with Lumi in a "wait" at heel. When I'd release her after the pheasant had rolled to a stop, she'd break from her sit like a shot, run to the bird, and race back with it. I'd reinforce that with another throw of the dummy into the water, and so the cycle continued, a continuous stream of self-reinforcing behaviors feeding upon one another.
By the end, Lumi seemed to be enjoying the bird more and more. As we were walking up the hill toward the van, I tossed the dummy one time and Lumi just looked at me for a second. "Hey, where's the bird? That's just a dummy!"
"You better get used to seeing more dummies, Sweetie," I told her. "Those birds won't last forever. You might not mind maggots and red ants, but I sure do."
TRAINING PISTOLS
Although Lumi already has some gundog instincts and trained skills, her noise sensitivity trumps everything else. In the big picture of gundog training, it's a small matter, but for Lumi and me, it's the top priority, the only priority, because nothing else matters if we can't solve it.
So I've put a lot of thought into how to work on gunfire D&CC. Yet it didn't occur to me until today that I might be able to order a pistol that shoots blanks online.
Sure enough, you can buy them from online stores for gundog supplies, and other places as well. I found one that was pretty inexpensive: about $22. It should be here in a few days.
Then all I need is someone to stand about a mile away and fire it for me while I play with Lumi. You'd think it would be easy to find someone to do that, wouldn't you? We'll see.
WATCHING THE LABS
Meanwhile, as suggested to me by private correspondence, I didn't spend all my time at yesterday's training day working with Lumi. At one point, I put her in the van and went off to watch the pros working with their dogs, or their clients' dogs, for awhile. I don't know why the difference between marks and blinds was so confusing to me before, there's really nothing to it once you've watched for a few minutes.
Watching serious field-trained Labs was an education in many ways. It goes without saying that they are nothing like your typical neighborhood Lab. One of the yellows looked like a body builder, muscles bulging up down his legs. He also had no hair around his elbows, perhaps from hunting hazards? I didn't ask why.
The skill of one of the black Labs in particular was also amazing, cutting off small triangles of water hazards to run a perfectly straight mark. As he approached each pond, he flew into the air and leapt in wildly. What drive!
Lumi, of course, will never look or act like one of those FC Labs, but she still looks athletic as Goldens go. I saw other Goldens there, and none of them had Lumi's lean, muscular physique. On the other hand, none of them had her fear of gunfire, either.
MORE PHEASANT WORK
We have so many things we could work on these days, but yesterday's field training inspired me to work with Lumi on field work some more today.
We still have two dead birds from last week to train with. They won't last forever. As I understand it, we can eventually look forward to maggots, red ants, and other niceties. I don't know where we'll get more birds, but at least we have these two for now.
We were training with our pheasant today at the lake for about half an hour. I'd throw Lumi's dummy into the lake, and then when she brought it to me, happily swinging into heel position, I'd reinforce that by throwing the pheasant for her up onto the hillside, with Lumi in a "wait" at heel. When I'd release her after the pheasant had rolled to a stop, she'd break from her sit like a shot, run to the bird, and race back with it. I'd reinforce that with another throw of the dummy into the water, and so the cycle continued, a continuous stream of self-reinforcing behaviors feeding upon one another.
By the end, Lumi seemed to be enjoying the bird more and more. As we were walking up the hill toward the van, I tossed the dummy one time and Lumi just looked at me for a second. "Hey, where's the bird? That's just a dummy!"
"You better get used to seeing more dummies, Sweetie," I told her. "Those birds won't last forever. You might not mind maggots and red ants, but I sure do."
Thursday, March 8, 2007
Enthusiasm and Intermittent Reinforcement
When I began trying to learn how to use intermittent reinforcement for Lumi's training recently, my objective was to prepare Lumi better for freestyle trials, in which she needs to go about three minutes without food, toys, or play breaks. That is, she needs to go without any primary positive reinforcement (+R), unless performing the freestyle routine itself is an intrinsic reinforcer to some extent.
PARADIGM SHIFT
My reason for beginning to experiment with intermittent reinforcement was based loosely on a suggestion made to me on DogTrek. I speculated that the dog's experience with intermittent reinforcement is to some extent a learning paradigm, like clicker training. According to that concept, a dog accustomed to intermittent reinforcement for some behaviors would also be more comfortable with intermittent reinforcement for others.
So I chose a few behaviors — on doctor's orders, we weren't working on freestyle at the time — and began gradually putting them on intermittent reinforcement, specifically, a variable schedule of reinforcement (VSR).
RATIO STRAIN
This turned out to be more difficult with Lumi than I expected it to be. After only a small number of unreinforced reps, Lumi would seem to become confused by cues she had previously had reliable responses to. She might freeze and just look at me, or she might offer some completely different behavior.
One of my correspondents on DogTrek stated that this showed that Lumi did not have those behaviors on "stimulus control", so I researched the term a bit. What I learned is that stimulus control isn't an absolute, it's the measurable tendency of an organism to respond to a stimulus. According to that understanding, those behaviors and a hundred others that Lumi offers in response to cues are on stimulus control. They just aren't proofed for extinction when not being reinforced as often as needed.
Which makes Lumi like every other trained dog. That's because every trained dog has behaviors that are not intrinsically reinforcing, behaviors that she does instead because she has been conditioned to do them by positive or negative reinforcement. Those behaviors vary in their resistance to extinction, but they'll all extinguish eventually if not reinforced for a large enough number of reps. No one would suggest that those behaviors are not on stimulus control simply because they won't last forever without reinforcement. The same applies to Lumi's behaviors.
Instead, it was simply a case of ratio strain, exceeding the dog's capacity for performing without reinforcement. According to Pam Reid in Ex-Celerated Learning, dogs vary by individual in how long a behavior will persist without reinforcement, and I'm guessing some of it has to do with how much prior experience the dog has with ratios as well. Lumi has little experience with VSRs, and she also may have less intrinsic tolerance for them anyway. Pam speaks of a BC who might go 30 reps for the chance to chase a tennis ball, while some Saluki might not go 30 reps no matter what reward is available. I'd guess Lumi is somewhere in the middle of that spectrum.
UNEXPECTED RESULT
While putting a few of Lumi's behaviors on intermittent reinforcement proved to be more difficult than I expected, I had another surprise in store as well. Lumi's enthusiasm for the behaviors, as long as we didn't hit ratio strain, seemed to go up during the training, not down. Since motivation affects the amplitude of a response, it seems that intermittent reinforcement was increasing Lumi's motivation. I had heard of this phenomenon, but I guess it was counter-intuitive, because I still didn't expect it.
I've since discovered that I'm not the only one. Other trainers, too, make no connection between intermittent reinforcement and motivation. This seems especially true of those who for one reason or another have little use for VSRs in their training.
ON FURTHER REFLECTION
In thinking about it, I realized that it's actually inevitable that intermittent reinforcement would result in more vigorous, enthusiastic responses. The reason can be given in two words: "extinction burst".
An extinction burst is the phenomenon that sometimes occurs when an organism finds that a previously reinforced behavior is no longer being reinforced. Eventually, the behavior will extinguish (stop occuring), but first, the subject is likely to attempt more vigorous versions of the behavior in hopes that those versions will obtain the reinforcement that the less energetic version did not earn.
To the subject, intermittent reinforcement looks exactly like extinction until the next reinforcement occurs. Looked at in that light, it seems only natural to expect higher motivation for behaviors being trained on an intermittent schedule.
SLIPPING AWAY
That would seem to make intermittent reinforcement a pretty attractive training tool, and truth be told, I am attracted to it for that reason. But just as an extinction burst can quickly slip into extinction, a VSR can quickly slip into extinction, too. That's ratio strain, what I was seeing with Lumi.
In addition, just as an extinction burst can result in other kinds of behavior variability besides energy level, I've found it difficult to apply intermittent reinforcement without losing some precision in some of Lumi's behaviors.
But I think I've found one way to take advantage of this motivational boost without those disadvantages, and that is by means of non-contingent variable-value reinforcement.
DISPENSING TREATS
When you read about lab experiments, it seems as though behavioral scientists generally dispense a single-value reinforcement after each correct response, sometimes using automatic, mechanical dispensing equipment. When dog trainers are training behaviors, they often try to simulate that equipment in various ways: maintaining a neutral composure, giving the treats with precise timing, and dispensing uniform amounts.
In addition, many trainers also sometimes give what they call "jackpots". By this, they do not mean the random payoffs that occur in casino gambling, but extra rewards when the dog seems to show a breakthrough in understanding or higher than usual enthusiasm in her response.
That is not the kind of variable reinforcement I'm talking about for increasing motivation. The kind I'm talking about is actually more like the casino terminology, in that it is random and is not contingent on the quality of the correct response
Let's say the dog gives five correct responses in a row. Instead of giving an equal reinforcement for each response, say a single treat 7 units in weight, you might give the dog several treats each weighing 2 units. But the number of those treats would vary randomly from one rep to the next: 3 small treats for the first rep, then 1, then 4, then 7, then 3 again, for example. Both dogs received about the same number quantity of food, but the second dog also experienced the element of surprise from one rep to the next.
DOES IT WORK?
From my experience, many trainers have never tried that approach to giving reinforcement and are convinced that it would not have any effect on the dog's acquisition of the skill being trained. For them, trying to understand why it would work is beside the point because they are convinced it would not.
And for some dogs in some circumstances, I'm sure they're right. If the dog is particularly hungry for example, as the typical laboratory subject is, then variations in motivation, if they exist at all, may be too slight to observe. Also, motivation isn't the same as speed of acquisition. Two dogs may learn the same skill but with very different attitudes toward the training process and the skill itself, and no data may be collected to show those differences. It's comfortable to assume that if a scientist says he has not seen improved results with a particular approach, it would a waste of time for you to try it yourself on your own dog.
Nonetheless, at the recommendation of my friend Lee Baragona, a trainer with national championships in multiple sports with multiple dogs, I began experimenting with randomly varying numbers of treats for sequences of correct responses as described above. The result was immediate and striking: a distinct boost in Lumi's level of excitement for the game. I could even feel it affecting me as the treat giver. For example, I might think, "Here's a smaller amount than last time, Lumi, but you know what that means, don't you? Maybe more next time!"
VARIATIONS ON A THEME
Once you have the experience that that way of giving reinforcement does affect the dog's attitude, then you find yourself wondering about an explanation. For me, the explanation is that intermittent reinforcement is really just a special case of random non-contingent variations in reinforcement.
That is, in a VSR, the number of treats is still varying randomly from one correct rep to another. In a VSR, the number of treats is either 0 or 1, whereas in the more general case that I'm describing, the number of treats includes other possibilities. You might vary among counts of 1 through 7, or you might also respond to some correct responses with no reward. You would only do that, I guess, if the dog has enough reinforcement history for the correct response that she won't take zero to mean that she had performed the behavior incorrectly.
FOLLOWING FOOTSTEPS
In all of the above, I may well be following in the footsteps of researchers and authors who have explored learning and motivation in the past. Some may have reached the same conclusions, some the opposite. Nonetheless, it's interesting to experiment and try to understand what I'm seeing on my own.
It's one of the things that makes dog training so rewarding.
PARADIGM SHIFT
My reason for beginning to experiment with intermittent reinforcement was based loosely on a suggestion made to me on DogTrek. I speculated that the dog's experience with intermittent reinforcement is to some extent a learning paradigm, like clicker training. According to that concept, a dog accustomed to intermittent reinforcement for some behaviors would also be more comfortable with intermittent reinforcement for others.
So I chose a few behaviors — on doctor's orders, we weren't working on freestyle at the time — and began gradually putting them on intermittent reinforcement, specifically, a variable schedule of reinforcement (VSR).
RATIO STRAIN
This turned out to be more difficult with Lumi than I expected it to be. After only a small number of unreinforced reps, Lumi would seem to become confused by cues she had previously had reliable responses to. She might freeze and just look at me, or she might offer some completely different behavior.
One of my correspondents on DogTrek stated that this showed that Lumi did not have those behaviors on "stimulus control", so I researched the term a bit. What I learned is that stimulus control isn't an absolute, it's the measurable tendency of an organism to respond to a stimulus. According to that understanding, those behaviors and a hundred others that Lumi offers in response to cues are on stimulus control. They just aren't proofed for extinction when not being reinforced as often as needed.
Which makes Lumi like every other trained dog. That's because every trained dog has behaviors that are not intrinsically reinforcing, behaviors that she does instead because she has been conditioned to do them by positive or negative reinforcement. Those behaviors vary in their resistance to extinction, but they'll all extinguish eventually if not reinforced for a large enough number of reps. No one would suggest that those behaviors are not on stimulus control simply because they won't last forever without reinforcement. The same applies to Lumi's behaviors.
Instead, it was simply a case of ratio strain, exceeding the dog's capacity for performing without reinforcement. According to Pam Reid in Ex-Celerated Learning, dogs vary by individual in how long a behavior will persist without reinforcement, and I'm guessing some of it has to do with how much prior experience the dog has with ratios as well. Lumi has little experience with VSRs, and she also may have less intrinsic tolerance for them anyway. Pam speaks of a BC who might go 30 reps for the chance to chase a tennis ball, while some Saluki might not go 30 reps no matter what reward is available. I'd guess Lumi is somewhere in the middle of that spectrum.
UNEXPECTED RESULT
While putting a few of Lumi's behaviors on intermittent reinforcement proved to be more difficult than I expected, I had another surprise in store as well. Lumi's enthusiasm for the behaviors, as long as we didn't hit ratio strain, seemed to go up during the training, not down. Since motivation affects the amplitude of a response, it seems that intermittent reinforcement was increasing Lumi's motivation. I had heard of this phenomenon, but I guess it was counter-intuitive, because I still didn't expect it.
I've since discovered that I'm not the only one. Other trainers, too, make no connection between intermittent reinforcement and motivation. This seems especially true of those who for one reason or another have little use for VSRs in their training.
ON FURTHER REFLECTION
In thinking about it, I realized that it's actually inevitable that intermittent reinforcement would result in more vigorous, enthusiastic responses. The reason can be given in two words: "extinction burst".
An extinction burst is the phenomenon that sometimes occurs when an organism finds that a previously reinforced behavior is no longer being reinforced. Eventually, the behavior will extinguish (stop occuring), but first, the subject is likely to attempt more vigorous versions of the behavior in hopes that those versions will obtain the reinforcement that the less energetic version did not earn.
To the subject, intermittent reinforcement looks exactly like extinction until the next reinforcement occurs. Looked at in that light, it seems only natural to expect higher motivation for behaviors being trained on an intermittent schedule.
SLIPPING AWAY
That would seem to make intermittent reinforcement a pretty attractive training tool, and truth be told, I am attracted to it for that reason. But just as an extinction burst can quickly slip into extinction, a VSR can quickly slip into extinction, too. That's ratio strain, what I was seeing with Lumi.
In addition, just as an extinction burst can result in other kinds of behavior variability besides energy level, I've found it difficult to apply intermittent reinforcement without losing some precision in some of Lumi's behaviors.
But I think I've found one way to take advantage of this motivational boost without those disadvantages, and that is by means of non-contingent variable-value reinforcement.
DISPENSING TREATS
When you read about lab experiments, it seems as though behavioral scientists generally dispense a single-value reinforcement after each correct response, sometimes using automatic, mechanical dispensing equipment. When dog trainers are training behaviors, they often try to simulate that equipment in various ways: maintaining a neutral composure, giving the treats with precise timing, and dispensing uniform amounts.
In addition, many trainers also sometimes give what they call "jackpots". By this, they do not mean the random payoffs that occur in casino gambling, but extra rewards when the dog seems to show a breakthrough in understanding or higher than usual enthusiasm in her response.
That is not the kind of variable reinforcement I'm talking about for increasing motivation. The kind I'm talking about is actually more like the casino terminology, in that it is random and is not contingent on the quality of the correct response
Let's say the dog gives five correct responses in a row. Instead of giving an equal reinforcement for each response, say a single treat 7 units in weight, you might give the dog several treats each weighing 2 units. But the number of those treats would vary randomly from one rep to the next: 3 small treats for the first rep, then 1, then 4, then 7, then 3 again, for example. Both dogs received about the same number quantity of food, but the second dog also experienced the element of surprise from one rep to the next.
DOES IT WORK?
From my experience, many trainers have never tried that approach to giving reinforcement and are convinced that it would not have any effect on the dog's acquisition of the skill being trained. For them, trying to understand why it would work is beside the point because they are convinced it would not.
And for some dogs in some circumstances, I'm sure they're right. If the dog is particularly hungry for example, as the typical laboratory subject is, then variations in motivation, if they exist at all, may be too slight to observe. Also, motivation isn't the same as speed of acquisition. Two dogs may learn the same skill but with very different attitudes toward the training process and the skill itself, and no data may be collected to show those differences. It's comfortable to assume that if a scientist says he has not seen improved results with a particular approach, it would a waste of time for you to try it yourself on your own dog.
Nonetheless, at the recommendation of my friend Lee Baragona, a trainer with national championships in multiple sports with multiple dogs, I began experimenting with randomly varying numbers of treats for sequences of correct responses as described above. The result was immediate and striking: a distinct boost in Lumi's level of excitement for the game. I could even feel it affecting me as the treat giver. For example, I might think, "Here's a smaller amount than last time, Lumi, but you know what that means, don't you? Maybe more next time!"
VARIATIONS ON A THEME
Once you have the experience that that way of giving reinforcement does affect the dog's attitude, then you find yourself wondering about an explanation. For me, the explanation is that intermittent reinforcement is really just a special case of random non-contingent variations in reinforcement.
That is, in a VSR, the number of treats is still varying randomly from one correct rep to another. In a VSR, the number of treats is either 0 or 1, whereas in the more general case that I'm describing, the number of treats includes other possibilities. You might vary among counts of 1 through 7, or you might also respond to some correct responses with no reward. You would only do that, I guess, if the dog has enough reinforcement history for the correct response that she won't take zero to mean that she had performed the behavior incorrectly.
FOLLOWING FOOTSTEPS
In all of the above, I may well be following in the footsteps of researchers and authors who have explored learning and motivation in the past. Some may have reached the same conclusions, some the opposite. Nonetheless, it's interesting to experiment and try to understand what I'm seeing on my own.
It's one of the things that makes dog training so rewarding.
Wednesday, March 7, 2007
Backing Together
I mentioned in yesterday's story that Lumi and I were doing some remedial work on a basic freestyle move: backing together with Lumi at right heel. I described the training as a repeating two-behavior chain:
That may have been a good way to perform this training over the last few days, but this evening I saw that Lumi was anticipating the behavior that was being reinforced — her move backward — and not waiting for me to step back. The result was that she would get too far behind me unless I cued "wait" every time.
If we're going to get a smooth version of this chain, I decided that we also need Lumi to have a strong reinforcement history for the wait, not only for her stepping back.
So I tried doing it this way. First I say "wait" and step back, then reinforce Lumi for waiting with a spoonful of her raw food.
Secondly, I say "OK" as a release from the wait, and Lumi backs into heel position again. I reinforce that, not with a primary reinforcer, but with another "wait". I believe the wait cue acts as a secondary reinforcer because it's an opportunity to earn another cookie if executed correctly.
I was pleased with the results, at least at the slow speed we were working. To earn her cookie, Lumi remained calmly in place as I took a step back, then quickly came back into heel as soon as I released her. Her level of motivation for the game played this way remained as high as with yesterday's version.
The overall backing move, as the chain is repeated over and over, is still too choppy. But it already has a nice measured feeling that hopefully will eventually be as fast, or slow, as required for a particular musical phrase.
- With Lumi in a wait, I step back so that I'm even with Lumi's lower back
- Then she steps back into heel position
That may have been a good way to perform this training over the last few days, but this evening I saw that Lumi was anticipating the behavior that was being reinforced — her move backward — and not waiting for me to step back. The result was that she would get too far behind me unless I cued "wait" every time.
If we're going to get a smooth version of this chain, I decided that we also need Lumi to have a strong reinforcement history for the wait, not only for her stepping back.
So I tried doing it this way. First I say "wait" and step back, then reinforce Lumi for waiting with a spoonful of her raw food.
Secondly, I say "OK" as a release from the wait, and Lumi backs into heel position again. I reinforce that, not with a primary reinforcer, but with another "wait". I believe the wait cue acts as a secondary reinforcer because it's an opportunity to earn another cookie if executed correctly.
I was pleased with the results, at least at the slow speed we were working. To earn her cookie, Lumi remained calmly in place as I took a step back, then quickly came back into heel as soon as I released her. Her level of motivation for the game played this way remained as high as with yesterday's version.
The overall backing move, as the chain is repeated over and over, is still too choppy. But it already has a nice measured feeling that hopefully will eventually be as fast, or slow, as required for a particular musical phrase.
Subscribe to:
Posts (Atom)
