I considered leaving the agility field and not asking for the teeter again. I knew what she was showing me was a "warning" signal she was going to break down. However, I asked for the teeter again, and she did come off and start refusing. I got the leash back out and insisted she do the teeter. After she did it with the leash, I took the leash off and asked for a couple of reps without it and then we were done with that success.
It wasn't until later when I realized that the glancing over the side was a similar signal to me as Devon's head flick to the left right before she's going to break down in pile work. In pile work (field training) where the dog goes out to get bumpers from a distant pile, the goal of the exercise is two-fold: 1) sending to a distant pile for a bumper that isn't thrown and 2) repetition to build work ethic by doing something when asked by the handler (even if they don't want to). In pile work, Devon will start to break down around bumper #6. If we don't push through to 10 or 12 bumpers, she starts breaking down earlier and earlier until she won't even go on bumper #1. At that point we have a huge battle. However, if I correct the problem in the first training session when I see it, we never get to the huge battle.
Last night when I was talking with Kim, she brought up another training issue: doing too much and then failing when the dog has had too much. I've done this way more times than I can count! I've had 2-3 successes, so I push for more instead of getting out of it on a good note and then ending up with a battle I don't need. The proper procedure in this case is to go back to where you were successful, be it fewer reps or and easier requested behavior, and then stop with success. I totally agree with this.
I guess for me it's all about reading the dog and where you are in the behavior. In my example with the teeter, the head checking went from rep 4 or 5 to rep 1 in the next session. That told me we were going to have to deal with this issue and stopping earlier in our reps was only going to delay addressing the behavior (like the pile work). Second, Devon has been doing the teeter at this height successfully for two days before a one-day break. She's been happily banging the teeter down, wagging her tail and standing straight up at the pivot. It's not like it was the first day she'd seen this height where she might be unsure and worried.
In my analysis, she just didn't want to do it. She doesn't like the teeter. She'd rather do any other obstacle out there; kind of like pile work. It's not fun, but doing that skill builds on other skills that are fun and that she's getting to do now. If we didn't have solid pile work in her foundation skills, she wouldn't be able to do blinds, and we are both having a lot of fun learning that skill.
I also have to admit she's likely getting a little bored. We've been doing isolated obstacle work for about 2.5 weeks now, and I'm sure it's getting boring. However, I need her to be confident on the equipment before we can start sequencing. This is the foundation work I didn't do last fall that bit us in the butt when I moved into sequencing and trialing way too quickly.
It's about time for me to go out and do agility training again today. I'll be interested to see what kind of teeter performance we have today!
2 comments:
I just thought of something that may help you work with Devon on the pilework, and on the teeter. Have you tried playing with your ratio of reinforcement? You say she breaks down typically after bumper #6. Have you done sessions were you stop at a lower number of repetitions of fetch, and do a mark (or something highly reinfocing for her)? Then other sessions will stop after a higher number.
This is one way of building duration in stationary behaviors using operant conditioning. For example, a stay on the table. The reward comes after 2 second, 5 seconds, 1 second, 10 seconds, 7 seconds, 4 seconds, 13 seconds, etc...on *average* you're building towards your goal duration each time you ask for the behavior, but the reward doesn't always come after the same number of seconds. In a similar fashion, you might try rewarding Devon after a variable number of bumpers or teeter attempts. Changing the ratio of reinforcement works for the same reason: because the dog figures out that eventually the work will pay off--it just never knows exactly when the reward will come, so it keeps working.
Actually, thinking a little more, I'd reward every teeter attempt, but vary the number you ask for in each training session. The bumpers are a different story for her since she has a different opinion about field work. It sounds like you're safe to vary your ratio of reinforcement to attempts there.
GREAT IDEA! I love the idea of the varying the number of times we do it each session. I was just looking at my training outline trying to figure out how to make it less boring but still getting the work in. I can also maybe do one teeter, then run over and do one set of jump work, then back to the teeter, then the weave poles, etc. so I'm running between the exercises instead of doing all weaves, then teeter, then jumping.
THANKS!! BTW, the pile work fixed itself once I insisted she do the behavior with a "reminder" from the collar. It extinguished completely when I went to an agility trial and gosh, darn if she still had to do pile work and that stupid collar went with us! Plus she doesn't really do "pile work" anymore without it being rolled into something else, like blinds, that's more fun. Pile work (going to a pile of bumpers) is the foundation for doing a blind (a planted bird that they do not see fall and they are sent to retrieve not knowing where the bird is) and a lot of stuff we do. Sort of like one jump work is the foundation for sequencing.
Post a Comment