Need some opinions on this method?

greytbigdreams

New Member
I am relatively new to clicker training, having learned it just recently through taking my dog obedience instructor/trainer course, but I am in love with it and my dog loves it. I have used clicker training thus far with my 9 year old toy poodle mix, to teach her to stop pulling on her leash and walk nicely beside me, the word "easy" which means she is about to get a tug on her leash for going too far ahead, and to slow down back to the proper position to avoid the correction (which she really learned well) as well as to sit-stay reliably, and to down-stay fairly well. She really loves it. The not pulling on her lead thing and the down-stay were really the only two things she didn't do and I figured she ought to, so now I am sort of scratching my head at what to teach her now.
She has a barking problem but I don't want to tackle that one just yet, but she does have an annoying habit of dropping whatever she is fetching about 2 feet from me and then coming the rest of the way. I want to get her to hold whatever is in her mouth, whatever she is retrieving, until I tell her to give it to me so I don't have to walk and get it every time to throw it again.

Since she already knows to fetch, or "go get it!" as is our cue phrase, I don't need to teach her that portion. But here is the program I came up with for this little behavior change and I'd like your thoughts on it.

Click for picking up the bone in her mouth after I have thrown it.
Click for picking it up and then carrying it 1 foot from where she picked it up.
Click for bringing it 1/4 of the way back before dropping.
Click for bringing it 2/4 of the way back before dropping.
Click for 3/4.
Click for bringing to my feet at least.
Click for bringing it to me and then holding it in her mouth for 1 second.
Click for her holding it for 3 seconds.
Click for 5 seconds.
Click for 10 seconds.
Click for 20 seconds.
Click for half a minute.
Click for holding it, and continuing to hold it when I move my hand under her mouth.
Click for holding it with my hand under her mouth for 2 seconds.
click for 5.
Click for 10.
Click for 20.
Click for half a minute.
Click for her giving it to me when I say "give it."
Click every other time when she does everything right and respondes to "give it.
Click every third time.
Click once in a blue moon.
Just use verbal praise.
**Clicker and treats are both phased out by this point**
And this would be where I could add a hand signal or something else if it were a different type of behavior, but since "give it" she already knows, and I just wanted her to associate it with the events preceding the command (bring it all the way to me and waiting for me to tell her to give it to me) I don't think there are any more steps needed.

I think it's a pretty good sounding plan. What do you think?
 

Jean Cote

Administrator
Staff member
That is an excellent plan, and very clearly broken down I might say! The only thing I might see happening down the road, is your dog learning to 'let go' of the bone when your hand goes under her mouth. This is because she will anticipate the "give it" command. Of course this can be avoided by testing her every so often, by placing your hand under her mouth and sometimes asking her to give you the bone, and other times to ignore your hand. :)

P.S. I would use verbal praise on all of your repetition, not just once clicking is over. :)
 

stormi

Well-Known Member
I have seen retrieve taught in a similar way to what you describe with very good results. Good Luck!
 

fickla

Experienced Member
that sounds like a fine plan for teaching the bring and hold! just know that the jump from bring to feet to hold for 1 second can actually be a very hard concept and will take awhile.

An alternative would be to first teach the hold. either with the dumbbell in your hand, or picking it up from the floor in front of you. I also taught one of my dogs to forcefully bump it into my hand before i started teaching any length of hold at all.

I only suggest this since the hold is the hardest part of all. if you teach it first, you are also back chaining which goes easier for most dogs. but then again if all you really want is a bring to hand, then you don't need much of a hold...
 

greytbigdreams

New Member
Thanks! It's good to hear I'm not off my rocker and that I am getting the hang of creating training plans for certain tricks.

Jean- Oh I always praise major whenever I click, I just didn't want to include that in each step because it was too much to write! Just like every click also has a treat, well, except at the very end of learning a behavior when I am weaning her off clicks and treats, I alternate. Sometimes she gets both, some time's she just get's a treat, sometimes she gets just verbal, sometimes all three. I know to absolutely never click and not reward so that one is never done on it's own though!

Fickla- Hey nice to see you are on here! I always love watching ferret dance videos and Lance and Vito training vids on youtube! Always so impressed! (I'm BakaniBoko over there.) We talked in PM's there one time, about training and theory and stuff I think! ^_^ And actually I have been learning in my training course about training dogs for open obedience and about the dumbbell, and I think I will change my method listed above so I can back chain. The more I train her the more back chaining really makes sense at least in this case. Get her always waiting to give me her toy, and then work on her doing that, but from farther and farther distances.
 

zoogal

Experienced Member
I have another suggestion for you if that doesn't work for you. The way I was taught (by Morgan Spector) ClickerSolutions Interviews -- Morgan Spector
was to back chain it (meaning start at the end and work your way back) I find this works GREAT for a lot of the film training I do but especially well with retrieves! What you do is start with the end result of the object in your hand. Start like your dog doesn't know how to retrieve and click for grabbing it (when you click the dog should drop it and if you click fast enough the drop should be in your hand. From there you progress to a hold and don't click until your hand is in position (make sure you never say commands like "sit" at the beginning of training as you want the dog to have the freedom to move. This way the natural response becomes to always release the object in your hand. Then practice raising and lowering you hand until your dog takes the object from both up high and low and always waits to release it until the click. Once you can put it on the ground and the dog takes it and waits until your hand is in position before releasing you can start movement (like you back up one step and dog needs to come one step so you can put your hand in position) or you can walk beside the dog while he holds and then randomly click when your hand is in position. You can also add other steps depending on how good your dog already retrieves and holds, if you get stuck I will list them.
 

snooks

Experienced Member
Very cute picture BTW. I think that's a fine job of shaping a behavior that is working well for you.

A couple of things to ponder though that now that behavior is established it's time to taper off and quit the clicker "for this learned" solid behavior, and slow the treats down taper to random over time. You don't want to live with a clicker in ur hand -at least I don't or billionz of treats everywhere. Not that I'm perfect-i still find last year's treats in coat pockets when it gets cold. :dogtongue:

The idea behind weaning from treats is that you actually get a dog that is at a higher rate of compliance because he's guessing is it this time? THIS TIME?? You know when you want them to stop something you previously thought was adorable and you really played into it and then you decide having your dog catapult off grandma isn't as funny as when she does it to you. You quit rewarding it and your dog goes nuts with the catapulting. It's called a pre-extinction burst. There is a balance with random unpredictable treats and uber compliance with tasks. On the other hand there is the dog that figures out why should I come now when I get a treat anyway I want to sniff this bush first.

Your method is perfect if it works for you and I use similar with great success. Zoogal makes a great point about backchaining-or teaching the final behavior first and working backward. IF you have trouble with that final hold Fickla describes as difficult might be easier if you backchain-it depends on you and your dog's style. Just food for thought since I have a great interest in different ways to teach things. My failure rate is variable but my success rate is eventually 100% if I explore enough, and both my dogs learn very differently.

One thing interesting I got from my Karen Pryor clicker class is that words of praise can dilute the click and it's razor precision. So in some cases it may serve you better to remain silent because the click IS the reward-it has been cemented as a primary reinforcer in the dogs mind click = food. :dogtongue2:

I notice if i say good when my dog goes for the correct item sometimes she breaks off early and comes back without it on a retrieve for example. So food for thought sometimes silence is Golden. I'm a big chatterbox praiser too, the silence trick helped me a lot with service dog work since it tends to be LONG chains of complex situationally adaptable (dog-thought-decision needed) tasks.

Impressive work with ur pup--don't take this as criticism-just gravy. I love to see people do work like this with their dogs and admire it. Very nice work, I love Poodles too, my first dog was a Poodle so they have a special spot in my heart. Also like to share and learn from/with others esp when they're asking.

:dogwink:
 
Top