Saturday, April 25, 2009

Wetfoot and Dryfoot Drills

Background

I've been struggling the last few weeks trying to find a way to train my dogs to run land-water-land (LWL) retrieves without long delays at the pick-up.

This morning, we practiced at Cheltenham and I'd give the dogs a D- on every retrieve, of which each dog had four. The only way they'd have done worse would be if they didn't come back at all.

This afternoon, instead of making another the two-hour round trip to Cheltenham, I decided to work with them on some additional LWL retrieves at the pond behind nearby Stadler's Nursery.

I selected a location for a start line (SL), and as the dogs ran around loose, I walked to a point along the shoreline requiring a 40-yard swim. I would have preferred a shorter swim for this work, which would have been available at Cheltenham. But given the debris on the shore and in the water at the pond we were using, the line I chose seemed the shortest available.

Discovering the Wetfoot Drill

As I looked down at the location where I intended to plant the ducks we'd be using to practice with, I had a sudden inspiration. Unlike many of the water locations where we train, this particular spot was on a gradual slope, with only an inch or two of depths next to dry land, and about six feet of wading depth before the swimline where the dog would need to push off and begin swimming.

In previous training, I would have placed the birds well up on the shore. I would have wanted the dog to climb up on dry land, shake off, pick up a bird, spin around 180°, immediately re-enter the water, and swim right back. What I would not have wanted, but what I probably would have gotten, was an endless ritual of shaking off, sniffing or licking the bird, picking it up, maybe carrying it in some random direction, and putting it back down again to start the cycle over again.

Of course, that doesn't happen in an open-water retrieve. If the bird is floating in swim-depth water, the dog grabs it, turns around, and swims back.

What this special terrain at today's pond gave me was a way to split the behavior: I would place the bird, not in open water, and not well up on dry land, but right at the edge of the shore, half in the shallow water, half on the wet beach.

Now the dogs and I returned to the SL, and while Lumi waited beside me, I sent Laddie for his first retrieve. At the moment he was about to reach the bird, I blew come-in whistle (CIW) and then immediately called "Fetch". It worked! Although Laddie was a bit slow in getting turned around and back in the water with the bird, he earned a grade of C, far superior to this morning's four LWL retrieves. As Laddie alligatored into the water and began to swim, I cheered and applauded, and when he returned with a great delivery and a shake-off, I gave him three bites of fried chicken liver.

I slipped on Laddie's lead and sent Lumi for the other bird I'd planted. Her performance was a carbon copy of Laddie's, another grade of C.

I felt almost certain I'd hit upon a solution to the problem that's been plaguing us. I would call this the Wetfoot Drill.

Making Progress

I planted four more ducks at the same location. Now isn't the time for generalizing location, I felt, now is the time for getting the behavior so that we can build reinforcement history for doing it the way we want it done.

I sent Laddie again, then Lumi, then Laddie, and finally Lumi one more time. Laddie's grades were B and A, while Lumi earned two solid Bs.

Laddie had worn himself out chasing a goose in the pond before we began our drill, and I though that his swimming form on the third retrieve showed a little fatigue. Lumi, on the other hand, is an exceptional swimmer and showed no hint of fatigue. So I planted two more ducks and ran her twice more. This time, a B and a B+. On the last retrieve, she was prompt but I felt she took a little more time than necessary getting back into the water.

Looking Ahead

I won't know for some time whether the Wetfoot Drill really is a solution to our problem. It's possible that no matter how much we practice in shallow water, we'll never be able to get a high-quality pick-up and water re-entry from further inland. It depends on various factors the dogs find reinforcing and aversive that I can only guess at.

But at least I have a plan. For now, I'll continue to plant the ducks as today. When the dogs are both consistently offering grade-A pick-ups and re-entries, I'll gradually fade all the cueing, even the CIW. At that stage, we'll have a high-quality LWL retrieve as a default behavior, though of course only in this one location and only when the bird is half-submerged.

Then we can begin to raise criteria. First, I'll move the bird a little further onto the shore, and re-introduce the whistle, and even verbal cueing, if needed. Assuming that we can get back to grade-A pick-ups and water re-entries without cueing, I'll move the bird further inland still. If the dogs are deriving intrinsic reinforcement from their high-quality retrieves, hopefully supplemented by extrinsic reinforcement from the rewards I'm delivering, it's possible that that accumulation of reinforcement will overwhelm any competing self-reinforcement available from incorrect versions of the behavior. That's the goal, in any case.

And if we succeed in stretching the behavior out so that someday we are consistently getting high-quality retrieves when the birds are five and ten yards inland, we'll be ready to begin generalizing the behavior for other locations.

Another Idea: The Dryfoot Drill

Before I close, it occurs to me that the exact opposite of the Wetfoot Drill might also be effective! As I think about it, if I were to plant a bird, say, 50 yards inland, I believe both dogs would swim across, run the 50 yards, pick the bird up, run back the 50 yards, and wade right into the water. It's never occurred to me before, but I believe that at this stage in their development, it's pick-up locations near the water that the dogs have trouble with, not open water at one extreme, and not well inland at the other.

Therefore, a possible alternative to the Wetfoot Drill might be the Dryfoot Drill, in which we would start with retrieves at sufficient distance inland to assure a high quality retrieve, and then gradually shrink that distance. I like the Wetfoot Drill better because it involves less wear-and-tear on the dogs, it reduces the possibility of some other factor coming into play, and it better suits the location we happen to have available.

But if the Wetfoot Drill begins to falter in effectiveness, perhaps we'll look for a location to try out the Dryfoot Drill instead.

No comments:

[Note that entries are displayed from newest to oldest.]