Contact: LDRidgeway at gmail dot com

Saturday, June 28, 2014

Building motivation in a retriever

[The material in this post was extracted and edited from a reply I sent during a private email conversation.]

. . . As far as your dog's inconsistent motivation goes, I haven’t had a dog quite like that, so I'll be writing from my own experience, but not with any strategy proven for a dog like yours.

To start, I might mention that my primary focus above all other training objectives when Laddie was a puppy, and to some extent to this day, is building motivation. We play a lot of tug, we play a lot of fetch with various toys inside and outdoors, and I would never, for example, do anything that he might perceive as punishment for picking up a bird, even if I would have liked to discourage something happening at the same time, such as slipping a whistle. As an example, when I decided to try to fix Laddie’s vocalizing on water casts (which he’s done since he was a puppy), I saw almost immediately that if he took a cast while vocalizing and I stopped him in an attempt to discourage the vocalizing, it also discouraged his taking the cast. Needless to say, I dropped that “training” strategy the instant I saw that happening.

Thus you can see that if I were to try to work with a dog who had motivational issues, I’d probably pretty much stop normal field work for awhile and work on finding games that he loved, and I’d play those games with him for weeks or even months. Fetch with objects he particularly likes (including clipwing pigeons if you have them available), chase games where you run away as he’s coming back to you, “find it”, tug, water retrieves — those are the kinds of things I’d try since those are things my dogs love. I’d avoid all physical discomfort at least at first, for example hot days or cold water. In Lumi’s case, I’d avoid stick ponds, because although she’s crazy about swimming, she’s somewhat scared of stick ponds.

One more point. How long you train your dog per session could be a major factor in his motivation. A cardinal rule of positive trainers is always to stop while the dog is at peak motivation and wanting more, so that that’s how he spends his idle time before the next session, wanting more. For a young puppy, this means just a single retrieve per session. I have a friend who trains Schutzhund and she uses what she calls “commercial training”: She puts the young dog alone in a room while she’s watching TV, and when a commercial comes on, she runs in to train the dog for a short time, making sure that she’s back in her chair by the time the commercial ends. She told me that approach builds motivation rapidly.


Overall, my idea for a dog like yours would be to build a huge, unalloyed positive reinforcement history for active games, and I’d be looking for a gradual ending of his tendency to go into “I don’t want to.” If a few weeks passed and he was still exhibiting avoidance behaviors for activities I thought he loved, I guess I’d probably think about retiring him from competition. But if I found that, by playing only games that he loved, and by becoming skillful at stopping before he started to lose motivation, he gradually stopped bringing out his “I don’t want to” behavior, then I would feel some hope and my next step would be to begin shaping the games we were playing so that, while they remained highly motivating, they also began to introduce the skills I wanted him to learn for competition. All of my training with such a dog would be based on maintaining and building motivation. But again, I think that’s largely true of how I trained Lumi and even more so (because I was more experienced), Laddie.

Sunday, February 23, 2014

Anti-popping: The Fake Throw Drill

For the last year or two, I've been trying to repair Laddie's tendency to pop, that is, turn to look at me from mid-field after the send on a mark or blind, even though I haven't blown a whistle.

Popping is not necessarily a disqualification in retriever Hunt Tests and Field Trials, but it's at least a minor fault, and it certainly doesn't improve the dog's score. In the case of field trials, it could cost you a position in the final placements. In either venue, especially if combined with other issues, it could knock the dog out of an event.

However, training a dog not to pop is difficult, even for experienced field trainers using traditional methods, much less someone like me, both relatively inexperienced and using positive training methods for which no book or video exists.

I think part of the problem is that we're not really sure why a particular dog pops, and it might vary from dog to dog and from incident to incident. But the other problem is that even if we correctly guess the reason for a particular dog and particular scenario, it may not be easy to see how to counter-condition it. I've heard of people trying many methods without success, and the same applies to me.

Yesterday, a fellow trainer with an all-age field dog whose popping had become so frequent that he had decided to retire her, told me that he'd recently learned of an approach that he was having great success with: Hide from the dog after sending her. That way, if she looks back for help, no one is there to help her, and she receives no reinforcement for popping. Ideally, after a few tries, the unreinforced behavior extinguishes. But my friend is not sure yet how well her new anti-popping training will stand up in competition.

Meanwhile, over the last few months, I've begun trying to consistently use another approach: At the instant Laddie pops, I call out "sit". Then I walk out to him, slip on his lead, gently walk him back to the start line, and send him again. I've never seen him pop a second time when I do that, whether the bird is re-thrown or not. Still, I'm not sure whether this method will ever get rid of the first pop of the session. It hasn't yet, though fortunately many sessions have no pops in them these days, thank goodness.

At yesterday's session, I noticed something else, however. On the one mark where Laddie popped, the most difficult of the day, the throw was invisible during most of its flight, and only appeared for a split second just before it landed. It was also a 250+ yard mark with lots of hills and bales of hay in the picture, so while I knew exactly where to look and still didn't see it for most of its arc, Laddie might not have seen the flight of the bumper at all. (We were throwing a bumper to a bird, because the thrown bird was even more invisible.)

This has led me to wonder whether a common thread in Laddie's popping is confusion, and a common thread in confusion is not seeing the throw. I'm not saying that that explains all popping, but it might explain some of it, and optimistically, a side effect of fixing it might even be to cause other popping to become less likely as well, on the theory that confidence begets more confidence.

Following from that analysis, it seems that a fairly straight-forward drill can be used to help the dog learn to deal with that particular kind of confusion and thus, hopefully, popping in that situation. Assuming the dog has a good memory and is otherwise a good marker with good hunting skills, simply have the thrower fake the throw to a pre-positioned article. Everything else is normal: the duck call (if you're training for Hunt Tests), the gunshot, the throwing motion, the thrower taking a seat after the "throw". Then you as the handler line your dog up and send him as if nothing unusual were involved. When the dog gets out there, hunting if necessary, and finds that lo and behold an article is out there right where the thrower seemed to have been aiming his fake throw, you've just added reinforcement history to the dog's behavior of running to a fall even though he never saw the throw, and you've added to the dog's confidence that he can handle such a situation without popping, that is, without asking for help.

I guess it goes without saying that the Fake Throw drill can be run with incremental levels of difficulty. At the easy end of the scale, for initial training, the mark would be fairly short and simple, the gunner would stay out rather than retiring, and the the mark would be thrown as a single. At the difficult end of the scale, you'd use a fake throw on a long mark that would be difficult anyway because of obstacles, wind, and diversions, you'd have the gunner retire, and you'd include the mark as part of an otherwise difficult triple. In fact, you could even fake two or three of the throws in such a triple, not just one. When the dog has worked up to the point where he can consistently run such a triple without popping, even though some hunting might be required, you may have eliminated a significant share of the dog's potential for popping.

What about if the dog does pop in this drill? Well, in some of my other experiments, I've actually wanted Laddie to pop occasionally so that he could learn that it doesn't do him any good. For the Fake Throw drill, I don't see any advantage in that. Rather, I would say that if the dog pops, it means you've increased the difficulty too quickly and should scale back your subsequent setups accordingly, working back to that level of difficulty again eventually, but more gradually.

I've found that when Laddie hopes he's going the right direction, for example after a cast on a blind, but isn't yet certain, and then he suddenly finds the bird or bumper, it seems to add even more to his already over-the-top motivation. Perhaps the Fake Drill works the same way, and is actually a fun drill for the dog, in much the same way as people -- at least some people -- enjoy surprises. Fixing a problem (we hope), while making the game even more fun for the dog, is a great combination.