I tried something new with the the dogwalk today thanks to a suggestion from a reader who actually trains with Silvia. Rather than use the treat gizmo I took 2 toys, one that's Strum's absolute favorite and another that he likes but not nearly as much. I've been reluctant, nay chicken, to use toys in the past because for one he slams to a stop when he chases toys and finally gets them and also because I was concerned that he'd extend his stride even more if he had a toy to chase vs the stationary treat dispenser. But I love to experiment and thinking about it Strum probably slams to a halt even worse at the treat gizmo so I decided to give it a try. The method is to throw the first, lower value toy while he's on the walk then if he gets a hit throw the second higher value toy. The results were pretty good overall, much better than I was expecting and my throwing/timing was also better than expected though could use improvement.
The biggest improvement was that he added the extra stride for every rep. and didn't have any of the big obvious misses that he's been having recently. But now we're back to the less obvious misses and sure enough I rewarded 2 misses. One of them I suspected was a miss right after I rewarded it. The other was so close I had to spot it on the slow motion video. Both were fairly close hits, hard to see, but I could sense that one was a leap right after the fact.
The first 2 reps were backchained so I could get my timing down with throwing the toys. The rest were all full dogwalks. A few times he didn't get on the dogwalk and went straight for his chicken and while I wasn't thrilled about that it's interesting to have footage of him running on the flat. I put a short bit of it at the end.
Stats:
Cool sunny morning, about 29-30 degrees
4/6 (67%) hits
2/4 (50%) hits were full dogwalks
2/6 (33%) misses were rewarded
Both misses were close
I'm happy enough with today's results. I did make it easier by leading out and not having so much motion on my part. I'll continue this set up tomorrow and if we can get 80% then I'll start adding motion back in. So glad I didn't enter Strum in any Standard runs at his next trial in a week, I want this all worked out before I ask him to do it in the ring. Next trial after that will be in May so hopefully I'll have everything near perfect by then, or at least back to where we were.
He looks much better running for the toy. He seems like he is closer to the board when running it.
ReplyDeleteI think he looks great now! It seems as if you're throwing the toys pretty close to the end but I had to the do the same thing with my Vito at one point in sessions or all we were getting were leaps. After a few days of having close throws I was able to extend it again and go back to it being after a jump.
ReplyDeleteI love the rep at 1:15. Beautiful :)
ReplyDeleteI didn't like how he took off with both back legs together at the end of the contact twice, but it might have something to do with toy landing too close. Throwing takes a bit of practice, I know, but you got better real quick, so no worries there. If you'll need a better precision (and can't use balls) something heavy on a piece of rope flies really well.
There is no rush to throw the red chicken early. The first toy keeps him driving ahead just fine at this point :)
I can throw the high value toy several seconds after Ruby clears the contact if I'm not sure about what I saw, and he still runs ahead.
Remember to add a jump about 15 feet after the contact and throw the first toy over it. In the end you want him to see the jump as the finish line, so that even if no toy comes flying, he will first clear the jump, then turn to you.
Once Strum understands that the yellow chicken is not the real reward he might be tempted to look back at you before reaching the yellow toy. If that happens, send him to touch or retrieve the yellow chicken before throwing the red one.
Oh, and this two-toy stuff is not a part of Silvia's method. I use it because I wasn't successful with her "throw the toy in advance" idea. She puts up with a lot of my deviations :)
It's a good idea though, maybe you can patent your own method : ) I'm a big fan of experimenting with methods if something's not working. Sometimes my ideas don't work but it's fun to try.
ReplyDelete