Tonight, Laev showed me how our training is really starting to stick.
We set up two blinds in the barn, and I set up Laev before the first. Now, we'd done this for the first time in four months or so last Saturday, and Laev had made a single mistake -- she'd lost control of herself during the hold and bark and taken a shallow grab at the sleeve, for which she was removed (P-) and taken away for a restart. The worst punishment ever, losing the bite! Aaargh! She did not make another mistake on Saturday.
If I'd really thought about how we hadn't done a blind search in four months, I probably wouldn't have graduated her to doing it offleash, too. But Laev did a fantastic job of pretending that she'd been off the long line for a long time, never hinting that she would skip the first blind to go directly to the second, and stopping when I asked before going on. Beautiful, absolutely beautiful.
Tonight, I sent her around the first blind and called, "Laev, sit" as she whipped around it. Nice, tight turns, probably because we're indoors, but very, very pretty. Laev skidded to a halt and managed to give me eye contact, and I sent her on to the second blind. Yay!
The second time, Laev tested parameters, deliberately bumping the sleeve with her nose during the hold and bark. Her mouth was closed, but that's still outside my definition of a correct hold and bark, so I came around the blind, took the tab on her collar, and removed her. Even though the P- was late by 10 seconds or so, Laev apparently understood it, because it did not happen again in her first or second session tonight!
And then the funny instance, where I called for a sit as Laev came around the blind and she CHEATED, running right by me to the second blind. I made an annoyed "aart!" sound, marking the moment she failed and alerting the helper that she could not be reinforced. I collected her and returned for another try.
(Note: Laev does not, by all appearances through her lifetime, find such a sound to be punishing of itself. It can, however, serve as a marker for P-, after conditioning; a sort of anti-clicker, if you will. If someone wants to call this P+, I won't argue, but I would say that it's been conditioned to be punishing rather than being a primary punisher.)
Correct response, the next time. And the next. Whee! Laev is a one-rep learner in many things, and it seems that she was also learning in one rep which actions were mistakes which would not pay off.
And here's where it gets funny. Laev whipped around the blind and I called her to front. Laev started to blow by me, and I marked the error. Laev skidded to a halt, sliding in the wood chips, and BACKED UP to face me. Picture this -- a Schutzhund dog, offleash, in full drive for bitework, interrupting herself and backing from the helper to the handler. Whee! I was so proud of her. I marked the corrected behavior and sent her on for a hold and bark and then a bite.
My mistake tonight was that I sent her around me in only one direction, so I'm going to have to remember to alternate directions next time.
We followed with some escape bites, Laev's first; these confused her slightly ("I'm allowed to break position and go after the bad guy?") but she did all right, and she'll gain confidence with time. We also did some obedience for the bite, which was much less impressive than her blind searches, but she hadn't seen that in several months, either. I've got to come up with a training plan to get that cleaner.
(Glossary:
blind search -- dog circles blind as directed, returns to handler, is sent to next blind
escape bite -- dog is left in position, helper attempts to flee, dog stops him without cue from handler
hold and bark -- dog searches blind, finds helper, alerts by barking intensely but may not bite helper. Laev was trained without P+ or R- for this, which just might be a first; see her blog entry "Look, Mom, No P+!" for details.
obedience for the bite -- dog is fully charged and ready for action, handler cues incompatible behavior, dog is rewarded by being sent to helper; this is great for teaching reliability and focus under heavy distraction)
