My baby dog Arie is not quite 2 years old, and started agility competition a few months ago. I taught her a 2on/2off contact behavior, using tons of positive reinforcement, and she was pretty good at it, though needing more training to get additional distance and independence, especially when working ahead of me (sends). I dealt with that by not putting her in a position she wouldn't be able to deal with. Prior to trialling, I went to a couple show-n-go's, and there it happened for the first time - Arie overran the bottom of the dogwalk. My instructor differentiates between "self-releases" as being a "naughty" impulse control issue, and overrunning, which, for a young dog is basically an honest mistake. So following the guidelines, I downed her briefly, and continued the course (down has been positively reinforced, a lot). Had it been a self release, she would not have been allowed to continue the course. This is negative punishment, taking away the opportunity to do more agilit, and is pretty much SOP for agility training, at least around here. She overran one more time at her third trial, and again I downed her and then continued.
Then she started not driving into position, and sometimes wasn't coming all the way down, instead doing a 4on at the end. This could be a result of her not having sends as solid as her run-bys and recalls, but it happened with those as well, though not all the time. Could it be the result of punishing her contacts? Not sure, but I backed off on working her independence until I had the behavior I wanted back, and got her driving into position again, and reliably and enthusiastically doing 2o/2o in the correct position. While doing this, I stopped asking for prolonged waits, unless I was tugging with her. I can't help wondering if we aren't punishing the dog's behavior when we work on proofing prolonged waiting times before we give the release cue. I also reinforced her best efforts with releasing her very quickly, followed by vigorous tug and verbal praise, or continuing on the course, also with verbal praise.
While working through this, I went to a camp with a top level trainer, whose opinion I value highly. I watched one of her students have to remove her dog several times from contacts, for which she was praised for maintaining her criteria. She also advised me to hold Arie on her contacts longer before releasing, to give my releases more loudly, and that at one point I should have removed her because she didn't wait for my release (she had released on "good girl" which I didn't realize at the time). While I am willing to wait 2-3 seconds, I really don't like doing it longer, as discussed above. The argument for giving my release more loudly is that she might not make the effort to listen, and start releasing off something else, some body movement. This is already something I work to avoid, by rotating among moving releases, releasing while still and then moving after she starts moving, and starting to move and then releasing (avoiding releasing with start of motion). Since dogs have such good hearing, it seems odd to me that I should need to speak loudly, and I am not a loud person -except when I'm working as the Gate Steward ;-) so this would add one more element that I need to concentrate on at a time when I have plenty of other things on my mind. If anything, I would think speaking softly would encourage her to focus on listening for the release. Anyone have any experience to back either opinion? As for releasing on the "good girl," I went home and replicated this easily, so took it off the equipment, worked holding her position (down or sit) through this phrase, reinforcing with the release cue and tug when she waits, not reinforcing when she got up before the release- this is something I should have worked through a long time ago, and simply neglected- bad trainer! Then I transferred it onto the equipment fairly easily, she did great on her first session. When I worked her the next day and she released on the "good girl" I took her back to kindergarten and worked it off the equipment, then back on the equipment. So far this has held up, and now I'm working some other phrases also, using the same progression.
I've decided I really don't want to follow SOP in this manner, though of course I've seen plenty of dogs do really well in agility who do. But could they have done better- faster, more enthusiastic- if they hadn't used negative punishment? And of course, I do realize the danger of allowing the dog to continue is to allow +R of a behavior I don't want. But had I been satisfied with using negative punishment, I might not have spent the time training the "good girl" issue, and getting all that +R in. So I'm wondering what others' opinions are regarding the use of negative punishment and possible fallout, whether in agility or elsewhere.
