Saturday, January 5, 2013

Shaping the broad jump for obedience


When I asked my obedience coach how she teaches the broad jump for obedience, her answer was "it depends."  :)    I've taught it for agility, but am not up on current methods, and was looking to expand my training repertoire.   So, we're going to shape Fizz to jump the broad jump.

Yesterday was our first attempt, and we worked her in 2 short sessions of roughly 5 minutes each, and approximately 10 minutes apart (you can see the clock in the background).    We started with two boards, and the criteria was simply "pass over the b.j. "    Early in Fizz's life, we did a lot of work with proprioception, which included ground pole work (and avoidance of stepping on them) as well as getting up on any thing that I went near and didn't ask for another behavior.   We worked a lot with wiggly, unsteady, non-level objects.    I knew that Fizz would try in some fashion to get ON the broad jump, and she did not disappoint. :)

Unfortunately, I forgot to start the video, so the first minute or so of interaction is missing, however, it's not very different from what you see on the video.   The very first thing she did was stand on the 2nd board, all 4 feet on, and hold her balance.   As soon as she came down off of it, having crossed over the full jump, I clicked and treated, and away we went! :)  

About half way through our first session, I stopped clicking for walking through the jump, and increased the criteria to "jumping the jump."   Fizz has a very high threshhold for frustration and does not give up easily.   She understands, and seems to enjoy, free shaping, and I am often able to take fairly big steps in raising criteria with her.  






You could almost see her mentally tiring, so we put her up in her crate for about 10 minutes to process what she had learned.    She was very excited to come back to the work.    In reviewing the videos, I see much less walking through the jump in Session 2, and more jumping.   She's understanding the criteria and showing me that we didn't raise it too quickly or too much.    One comment here is how poor my placement of reward is.   I thought I had cut hunks of cheese that were big enough, but nope..........they were small, bouncy, and not tossing very well.   Optimally, I would have been tossing cheese such that where she went to get it for reward set her up correctly for the next rep.   However, she is frequently to the side of the jump and coming to it on an angle.   

Personally, 'wrong' does not bother me, and I do NOT view it as 'bad.'   It is a choice the dog has made, and each new choice brings information.   She is refining her understanding of what brings reward and what does not.   By figuring it out for herself, her understanding is more clear, and in my opinion, more solid.    The same reasoning that causes a parent to say to a child "figure it out for yourself........put up 4 fingers on this hand, and 3 on that one, and tell me how many you have".....when asked "what is 4 + 3?"   In my limited experience, shaping is FAR FAR more powerful (and lasting) of a teacher than luring.   It requires patience, and the ability to read one's dog (and know if you've pushed too far/too fast).....but the rewards are great! :)

Here is session 2:




Wednesday, January 2, 2013

Scent discrimination, 2 Jan 2013, part 2

In thinking through this whole process, and trying to decipher which scent the dog is keying on, I got to wondering about something.   Interestingly, this same subject came up while discussing our progress with my friend / Fizz's breeder.   When one re-uses a hot/scented article, and puts it back in the pile for the dog to find, a new scent is added (heavily!) to it:  the dog's own scent, via saliva.    But I want her finding MY scent.   Therefore, I need a fresh article for each repetition.   B/c once she's chosen and grabbed the hot article, it's contaminated with her saliva/scent.   So......more canning rings received sharpie marks today, and now I have 6 to work with as my hot article. 

We worked 2 sessions like this today (versus previous sessions where I was just re-using the same hot article).   The first rep of the first session, she was quite slow, worked for a long time, and chose the wrong one.   After that, she was correct each time (5 more reps).  However, she was still working slowly, and I could tell she recognized a difference in the game.






2 hours later, we came back for our 3rd session of the day, and our 2nd with new hot articles for each search.   This time, she was fast, confident and correct.   I kept adding more rings to the pile, I think at one point, we were up to 14 in the pile.   I could really hear her sniffing out the pile, checking and looking for the right one.   And when she found it, she confidently grabbed it.  YAY! :)   I also added in some favored 'shopping' items from around the house, things she dearly loves to hunt and carry.   She didn't even consider looking at them.

Our final rep was a bit of a stretch, and I knew it as I set it up.  However, I also know that Fizz is not easily over-faced, failure is not 'bad' to her (or to me!), and I wasn't worried that 'things go wrong' would be some unrecoverable incident......so forward I went.   We had played tug with an old jute tug that she really likes.  I put it down, near the pile, and between Fizz and the pile.  She would have to cross over it to get to the pile.  I also sat her back from the pile, and then called her in, while I stood near it (instead of me sitting right at it).   She didn't even look at that tug as she bee-lined for the pile, got in and worked it confidently, gave me the correct ring, and then we had a tug party!  :)  :)   When I tried to toss aside the ring so I could hold the tug better, she went and brought it back! HA!  GOOOD girlie! :)




Scent discrimination 1/2/13


Continuing on, I decided to increase the number of ring, and add in some distraction.  Fizz LOVES James' matchbox cars.  I wanted something that was also metal, so cars are the distraction.  Normally, she loves to find them, and then trot around with them, showing off her "prize."   They seemed to bother her not one whit in her work.  Was VERY happy to see her choosing to work, vs. taking something that is self-rewarding (carrying a car around).   :)   

This session has 18 reps and 2 errors.   In one set, her errors are very convicted, she gives me a full indication (eye contact, presenting me the ring, etc).   I simply took it and looked back at the pile.  No correction and no non-reward marker.   Lack of reward is message enough, and I liked how she handled this, and that there was still the same intensity in her work. :)




Tuesday, January 1, 2013

Scent Discrimination

After watching Denise Fenzi's blog post regarding her puppy Lyra (who at the time, was 13.5 weeks old), I decided it was time to get started with Fizz on scent discrimination.   You can see Denise's blog here

I purchased some children's puzzles at the local dollar store, because they came in metal tins (and didn't have the smell that altoid tins have).   Dumped the puzzles out into bags for the kids, and started working with our tins.

Our first day went very well.  It didn't take me long to get her from 1 tin with food, to 2 tins, one with food and one without.   She was 100% right.   I removed the tin with food, put in a new '2nd tin' that only had my scent and allowed her to work.  Still 100% correct.  Her natural behavior is to pick the tin up and give it to me, so that's what I accepted.   I am NOT looking for a formal retreive, this is not currently about the retreive, but rather about her choosing the correct scent.   We worked up to 4 tins, and only a handful of mistakes were made.   I did not record that session, and thus, couldn't data-keep.

However, being one who LOVES data, I recorded this session, and then sat and made myself a data spreadsheet.   It may not tell me anything now, and it may not tell me anything later.   But hopefully, when a "big" problem crops up one day, I will be able to look back at my data and see where the issue started, and then have a more clear idea of how to resolve it.    But truth be told, I'm a science geek and I love data. :)

We start with our puzzle tins again.  No food in them, simply using my scent on the Dora puzzle.    We start with 4 tins, and quickly move to 5 tins.  On the 8th rep, I replaced the Dora tin with a new tin that had my scent.  It took her  2 seconds longer on this rep, but she was still confident in her indication and selection (and she was correct).   After that, we changed to canning rings.   This was new, and again, her first rep took 5 seconds, instead of her usual 3 seconds.   After that, her next 2 reps were 3 and 2 seconds respectively.

I noted the following parameters:  Rep #, number of objects, if she was correct on her first presentation, where the hot object is located relative to others, and to her approach, the time she was on the pile working it, and the time she presented me with her choice.   If there were any other oddities, they were noted too.   Overall, this session went 8.5minutes.    The first error does not show up until 5:23, it took her 6 seconds to choose the wrong one, and then 5 seconds to find the correct one. :)

The hot canning ring is marked with blue sharpie, from long long ago.  The sharpie should not be a noticeable smell to any effect.

My goal is to have Fizz working a full Utility pile of metal & leather articles by the end of January.   I will be curious to see how her percentage correct changes, and if her time in finding the right article improves or not.

Of note in this session:  Rep #4, she mouths the correct tin, then goes to check the others before bringing me the correct one.   Rep #12, I flung a canning ring too far, it was about 3ft from the pile (off camera).  She comes in, scoops it up and brings it to the pile, and then immediately puts it down and brings me the right one.  Her presentation of her choice is very clear:  she makes strong eye contact with me while distinctly  presenting  me with the object.  In scooping up the wayward ring, she never made eye contact, and never showed me any interest.  It was very typical of her 'around the house' carrying behavior, and clearly NOT a presentation to me of choice.   At Rep #14, she makes a wrong choice, and very clearly presents me with the incorrect ring.    Big difference from the 'wayward ring.'

In Rep #16, there's an interesting behavior of piling together the incorrect rings.  Never once does she make eye contact with me, or any indication that she thinks those are 'right.'   She's just busily moving them, and then grabs the right one and brings it to me.   Dogs fascinate me in their behaviors.   We took a play break at that point with an old SchH tug from Frol-dawg. :)    On Rep #17, she presents me with 2 incorrect rings.  Interestingly, I had just moved those rings, and touched them more solidly than the others that I moved around.  :)

The video shuts off at rep 21, I know we did at least 5 or 6 more, and they were all correct.   In the last recorded rep, she starts to grab the right one, bonks it under my leg, and then seems unable to find it.  I waited this out, and she did figure it out.  One other missing rep, she has 2 rings in her mouth (b/c they were touching).....recognizes that she has 2, puts them both down and only presents me with the correct one.   GOOD girlie!! :)

We had a lot of fun doing this, and can't wait to work on more! :)