Clicker training "hold"

 ok I'm no dog trainer here, but what if you actually assisted him in holding it for longer periods, then slowly phase out the assist? just an idea dunno if i would work or not-or if it would make problems in other areas. ( i'm still a clicker newb)

akyramoto82

 ok I'm no dog trainer here, but what if you actually assisted him in holding it for longer periods, then slowly phase out the assist? just an idea dunno if i would work or not-or if it would make problems in other areas. ( i'm still a clicker newb)

I kind of experimented with this but the result was too much "chewing" which is really bad.  The entire obedience phase is 100 points and 40 of that is *just* retrieves.

We did a really short session last night.  I think he's starting to understand that he needs to actually hold it until I say not to.  I'm clicking less, and he's getting less frustrated and more focused when he drops it prematurely and doesn't get clicked.  He will hold it for a full second ("one one thousand...";) which is hardly anything but still seems like an eternity compared to tapping it with his nose!

That sounds like good progress. Are you trying to use a fixed interval or variable interval reinforcement schedule at this point? I'll look at your video tonight--I don't have the right software at work and we're not allowed to install upgrades, etc. (Go figure!) 

Right now I am clicking and rewarding (with food at this point, will eventually drop a ball from my chin) for every good hold.  What was a good hold two days ago might no longer be a good hold today, so there are a few that I'm not clicking for, but for the good ones I'm still always clicking.  Since the behavior is so crucial I've been told to work on just the hold part for several months and not let the dog near a dumbell until he is correctly holding the pipe.

Sounds good--you'll be able to focus on length of hold and set up your schedule (e.g. clicking for a 3 second hold but not a 2) as you see improvement. And, you may decide you want to go for variable ratio eventually (where you focus on an average (e.g. 5 seconds) and reinforce different lengths of hold that average your target (Sorry! I imagine you know this already...it's the special ed teacher educator in me coming out :-)

My main goal is no chewing.  Now that he understands what I'm looking for, I no longer click for *any* chewing or dropping the pipe prematurely (he is allowed to drop the pipe when I click, not his fault if I can't catch it or I drop it).  I think the chewing is the hardest thing to fix so I've decided to make that the number one criteria.  My trainer has a SchH3 Malinois he is competing with at a high level but sometimes they have a chewing problem.  Sucks to be "fixing" this with a SchH3 dog capable of international competition.

However, tonight I'm wondering about moving on to another goal - initially I assumed it was length of the hold.  We're veeery slowly adding time.  I did a short video clip from tonight that just shows a few attempts.  The last one you can see is clearly the longest.  I gave the dog the "jackpot" reward and ended the session (with some ball play) because at the time I figured I was going for length (since he is not chewing).  But as you can see during that longer hold, he looks away at DH making a noise upstairs and then back at me.  That leads me to wonder if focus should be as important or even more important than length?  I still clicked and awarded the jackpot because he quickly focused back on me without me correcting or saying anything and I really really wanted to reinforce that longer hold.

 http://www.vimeo.com/8608666

So now my questions are, for those that have done this and compete with it, should I be really strict on focus and take a step backward as far as length of hold until the focus is unwavering?  Or should I make sure there are no distractions like DH trying to watch us from above and bring in distractions later?  Also, should I build length while reinforcing every time, and then start the variable schedule when I have the length I want (with no chewing and solid focus)?  Or should I start varying the reinforcement schedule as I'm building length?

I would be really strict on focus even if it means that you may have to step back length of the hold.  IMO, if the focus isn't there, then the retireve isn't good, no matter how nice the hold is.

I would work on making sure that he's focusing on you all of the time.  I would gradually add length to the hold, & when you are sure that he fully understands both the focus & hold, then I would start varying the length of the hold, as well as adding distractions.

I would be concerned about adding too many different criteria at once... If you want length of hold, no chewing, and focus in the face of heavy distraction, I think you should work on each separately, and then add them together... And when you DO add one, decrease your expectations of another slightly. If you think focus is most important, I would work holds and reward heavily for good focus, and not reward if the dog looks away or seems distracted. You could jackpot instances where he is focused and holding for longer periods, but I would not HOLD OUT for focus + length at this point.

Then you can add no chewing. Click and treat for a very short hold with no chewing and acceptable focus. Increase your expectations until he is never chewing and giving great focus.

Then add length... click for longer holds with acceptable focus and "acceptable" chewing, and slowly up the expectations til he is offering long enough holds with acceptable focus and no chewing. Then I would start increasing the pressure by only rewarding holds with good length AND great focus AND no chewing.

Once he is offering all three, I would then up the criteria by adding more distractions, but decrease the expectation when it comes to length of time. I would still expect focus and no chewing, but would use short sessions at first when in a new area and/or in the face of new distractions he has not worked this command around (so new dogs, a bouncing ball, new people, sounds of traffic...)

I hope that made sense!!

I would not use a variable reward schedule until the behavior is absolutely solid in all ways. Right now I would have an extremely high rate of reinforcement.

Thanks Amanda, makes sense.  I'll reward for focus with a higher priority than length of the hold.  No chewing HAS to be the number one criteria.  Chewing is a common problem, hard to fix it it's not dealt with right away, and looses a lot of points.  Focus is important too, but usually not as much of a challenge, at least in SchH.  Distractions are not as much of an issue.  Training in our own home is actually far more distracting for him than training a more difficult exercise on the SchH field because of the amount of distractions possible and the proximity.  Training on the SchH field and being in a trial is usually a pretty controlled environment, which is good and bad.  However if I want to do UKC/AKC obedience, focus and distractions will be more important because of the environment.

Liesje

No chewing HAS to be the number one criteria. 

I totally agree with this. 

Bevo & Shooter's retrieves were trained differently due to Bevo being a natural retriever, & Shooter not being a retriever.  Bevo was backchained.  He was easy to train, &  has/had a very nice retrieve with minimal work.  With Shooter, I played with him with the dumbbell, because I felt like I needed to get him excited by it before I could actually backchain the retrieve.  Even in the initial "touch" phase, Shooter had an intense desire to chew the dumbbell.  As we progressed, he continued to chew while he worked.  I allowed the chewing & my thought process was that once he got the retrieve down, I would "fix" the chewing issue.  He's been retrieving, reliably, for about 2 years, & he still chews the hell out of his dumbbell.  It's so bad that I have attempted to retrain the entire sequence while not rewarding for chewing.  It didn't work.  He's the type of dog who is happy to work for his rewards, but when I stopped rewarding for the chewing, he stopped being willing to work.  In hindsight, I should have handled Shooter's retrieve just as I did Bevo's.  It would have taken much, much longer, but it would have been worth it.  You live & learn.

tenna

I would not use a variable reward schedule until the behavior is absolutely solid in all ways. Right now I would have an extremely high rate of reinforcement.

I agree--originally, we weren't talking about a focus on 3 different behavioral components--the focus was on the duration dimension. The new video showed a lot of info so she could see what additional dimensions she needs/wants to work on.

How's the training going?

BEVOLASVEGAS

Even in the initial "touch" phase, Shooter had an intense desire to chew the dumbbell.  As we progressed, he continued to chew while he worked. 

Hi 

This is one of the reasons that for obedience i work hard on just one criterea which is firmness of hold right at the beginning. I don't mind how i capture that. With my old Lab i got that by doing some brief tugs with it and getting a great hold with it. I want my dogs to hang on to the dumb bell so hard until they are told to give that i could pick them up off the ground if i had too.

Have you tried taking the chewing to extincion? I.E. Just stand there with high value treats and as soon as you get a non chew start throwing the food? Sometimes you might need three or four dumb bells :)

I only do one thing different between retrieving and non retrieving dogs. In some ways getting a great retrieve out of a natural retreiving dog is harder than with a non retriever.

poodleOwned

Have you tried taking the chewing to extincion? I.E. Just stand there with high value treats and as soon as you get a non chew start throwing the food? Sometimes you might need three or four dumb bells :)

I haven't tried this, although it sounds like it is worth a try.  I appreciate the idea, because I've drawn a blank, & even our trainer is unsure how to fix it.  I suppose that I deserve the problems for trying to take a shortcut.  I should have known better, but after training with a dobe, & then moving to a doxie, I felt like he was moving at a snail's pace, & wanted to speed it up a bit.  Live & learn, I suppose.

FrisbyPI

How's the training going?

Thanks for asking.  I took several days off for personal reasons, but we started again last night and tonight.  I'm also doing perch work with a clicker, just to get him doing more shaping with a clicker.  Hopefully using the same method elsewhere might keep his mind sharp and move the retrieve along.  He is doing better than I thought possible, with one new caveat: tonight we did the "hold" after doing some intense obedience with the ball, so he was panting, and that made him a slight bit chewy.  Not real chewing, but not as tight a grip and the panting of the tongue was tapping the pipe.  However I made a barely audible "look look" command while he held the pipe (which comes out as a very soft "kluck" sound almost) and that got him focusing on my face, thus shutting his mouth tighter and not allowing the panting to cause chewing.  I got a few good ones and quit since he was already tired.