Do Dogs Understand Speed? - Page 1

It has been my general observations that rewarding for speed (in basically any exercise, from simply sitting fast to sprinting back on the retrieve) in and of itself does not necessarily produce speed. Rather, the dog's motivational state seems to be the key, i.e. higher motivation produces a faster response. That's all fine and good, and seems simple enough, but then problems seem to crop up with anticpation. The dog begins anticpate the fast reward, and becomes so focused on the reward, they seem to almost forget to actually do the action. Am I just not balancing varying lengths of how soon I reward the dog after the action well enough? Am I missing something in my foundation? Are my observations themselves off base?

I do flyball (among Schutzhund and many other things) so obviously speed is a huge factor overall, not just one or a few exercises in a larger scheme of a trial. IMO, the biggest reason a dog "forgets to do the action" is that it just isn't proofed enough. Everything we do is backchained, so when the dog starts messing up, we back up a step or two (or as far back as we need to go for success) and restart the training from there.  If the dog is not performing what is required, it's just got to be dealt with. NRMs (no reward markers) or waiting for a dog to get it right and only rewarding *that* have never worked for my dogs; the training *must* set the dog up for success and as soon as my dog makes a mistake twice, I back up.  Anticipation and high drive cannot be an excuse.  There are plenty of crazy fast dogs that are consistent to the point that their only "mistakes" in a tournament are handler errors (early passes).

Thanks VK! So, to clarify, you're saying that if a dog begins to anticipate, just back up? That certainly makes sense for complex exercises, but what's kind of getting me "stuck" is the sort of thing where the dog will "sit," but his butt isn't totally on the ground, or he's not settled into the sit, because he's waiting for the release so intently. The dog is fast and does the action, but only 90% of the way before he gets so locked in to the antipation for the reward that he doesn't complete the exercise. I'm not sure I'm describing that terribly well, but does that make sense? And in that scenario, it seems like rewarding him as soon as his butt touches the ground is as much as you can break it up/back it up. I'm guessing the root of the problem is that I rewarded him too early too many times, not realizing he was hovering, and now it's a bad habit. So I suppose it would make sense to just reward a complete sit from here on out and make that the new habit...

For me it depends on the dog.  If we're talking IPO type obediece, I add some -R right away, even with a younger dog, but ALWAYS paired with release and rewards.  The play, engagement, and rewarding always makes up about 90% of our time "working" on obedience.  I like training GSDs because I can use some pressure in my training (both physical pressure and some pressure from my own body language and how I handle my dogs).  If the dog has a decent temperament, he can work through this pretty quick and is still begging to go out and train. I am not talking about putting e-collars on puppies or cranking and yanking a dog off his feet by a choke chain.  A few well-timed pops used in conjunction with verbal affirmation/praise that my dog understands and some sort of release and reward goes a long way. 

I tend to use corrections (actually negative reinforcement) to help shape the behavior I want up front more often than correcting the dog after the fact.  For example, I once had a dog (now a multi-SchH3 national level dog) that absolutely WOULD NOT open his mouth and take an object in his mouth so I could start backchaining formal retrieves.  I don't train a forced retrieve (not because I am against them, but all of my dogs have enjoyed retrieving so I have done well simply spending a few months backchaining the behavior) but I had to use some force to get this dog to open his mouth.  Force applied, mouth opened, object in, force stopped.  The dog realized in less than 5 minutes what I wanted and never had a problem after that.  He has excellent retrieves, very clean and fast, calm hold and no outing problems.  That use of moderate force for one training session made it clear to him and we were able to continue.

I have a dog that would hover-sit.  He sat pretty fast and correct, but would often hover his butt a bit rather than go all the way down. What worked there was a quick pop on the prong collar but me correcting *forward* (so I put the prong on turned 180*, with the leash attached under the chin). The quick forward correction seemed to help trigger and oppositional reflext to sit all the way down backward.YMMV, this was just the trick that worked for me to get over that hump in our training.

Another thing to consider: is your reward in sight?  Like do you wear a treat pouch or is there a toy under your armpit? Some dogs just get too loaded and locked up (literally) when the object of their motivation is so close.

I agree, working a slow or hover sit can almost be MORE difficult and frustrating than training a more complex behavior chain!  I do like to try things to fix it right away.  It has not been my experience that letting dogs make wrong choices and then only rewarding the 50% or less time they do it right has any effect. If I see the same mistake or undesired behavior a few times in a row, I want to do something to fix it even if it takes me trying a few methods to find something that works.  Only rewarding the good while the dog is still doing it wrong some times doesn't work for me.

good response VK