Agility questions
So instead, I thought I'd subject you clever folks to my agility-related questions.. Inca does relatively well (even held her own in a class of 5 border collies last week - much to my delight and surprise!!) and our instructor is great, however we have a bit of a crutch. We train with lots of high value treats and a toy. She's not so fussed about the toy - but she seems to have adopted it as a guide to follow and without it things get a bit messy. The treats are absolutely essential when doing full practice runs (not so much when just training individual pieces) - but if she doesn't deem you to have anything worthy of her participation she would happily wander off and not even stay at the start line. So my question is, how can I wean her off the treats/toy without taking a step back in accuracy and speed? Any suggestions appreciated! |
|
|
| When do you give her treats?
E.g.: After each obstacle? After three (or whatever) obstacles? At the end of a sequence? After successfully performing something that's difficult for her? You use a lot of rewards during the training stages of obstacles in particular. For most dogs, over time, the obstacles tend to become self-reinforcing - some more than others, depending on what got her the most rewards  . What that means is that she has gotten so many cookies for, for instance, performing a nice teeter than she now has happy and fun thoughts about the teeter and wants to do the teeter every time she sees it on the offchance that a reward (cookie) may be forthcoming!
If she's wandering off and so on now is not the time to wean her off treats, but rather to use them, perhaps, more judiciously. Figure out when she wanders. If she's breaking a startline to go wander off (not to do some obstacles on her own but rather just to not do agility at all), instead of leaving her in a stay (which can in and of itself be demotivating to some dogs until you build some drive into the stay) go back and run with her more than leaving her in a stay and reward her after three obstacles (or one or two or) Keep it short. And then go on. Do 4 obstacles. Quick reward. Go on. Do 2. Cookie. Then five. Cookie. It sounds like you're asking her to do too much too soon. But that's OK. You can still run the whole sequence. Just break it up into logical section as you're walking it. Or grab the opportunity to quick reward her for accomplishing something that's been hard for her. Be fun. Be flexible. Be UNPREDICTABLE! Probably she's learned that now she has to do say 14 obstacles before she gets any reward, whereas, as noted, when teaching obtacles the reward rate is very high. That can be a huge jump. As time goes on, do longer sections before you break and reward until you are running full sequences, but then with a huge jack pot at the end. What happened from the sound of it is that you (correctly) are effectively weaning her off treats, but in too large a leap at a time. I went through this with my Mace to some extent. She didn't wander but was pokey till she knew we were getting towards the end of a sequence  So the above is what various instructors had me do and it works a treat (no pun intended )
And never compare Inca to the border collies. It's well known those critters are from outer space A couple of my instuctors are forever paying homage to the outer space critters in Sybil's class. Never mind that Sybil may not have completed the sequence in a frantic panic, but she was still faster and cleaner because she ran more efficient lines (less yardage) and actually kept the bars up and avoided the off courses 
Be proud to have what is clearly the superior breed
How old is Inca and how long have you been training? Kristine |
| What Kristine said.
In operant conditioning terms its called "fading the reinforcer" It works best if you're unpredictable in offering the rewards. So offer a reward after 3 obstacles one time and then 2 or 5 the next. Keep mixing it up and you maintain the motivation as Inca is never sure when she'll get a reward. Kristine, do you use jackpotting? I thought that it had been shown to increase speed and accuracy initially but then to cause demotivation in the longer term due to downward comparison ie now I want an even BIGGER reward. Or is jackpot just the term you chose to use and not in the operant conditioning use of the word? |
| Ahh that makes a lot of sense to how she behaves sometimes (a few weeks ago she would not stop going over the contact obstacles - every time we got near - had the whole class laughing but she gets a lot of praise/treats for just standing there doing nothing on the contact point.. so I guess she was going for the least effort/most reward approach to training!!)
Inca is 2.5 years - we've been training with this class for almost 6 months though had done a couple of months elsewhere when she just turned 1 year - but it wasn't so great and we mostly picked up bad habits (e.g. missing contacts). The way we tend to treat her is when working on just one obstacle repeatedly she gets a treat everytime - and we rarely focus on a sequence of obstacles (other than a sequence of jumps - where she gets treated at the end of the sequence) - then when running the whole course, she will get treated right at the end. She does run the whole course without stopping, and she doesn't expect to be treated along the way, its just she won't start or appear at all motivated unless she knows whats on offer  .
I have to say I was mortified when I saw a class full of border collies, having never wished to compare Inca's performance to them - it wasn't our normal class, we just joined it because our usual class wasn't running (since it would have been christmas eve) - but after Inca's first run I overheard one of the bc owners remark "she's pretty good!" which was probably a statement of astonishment that it could be possible - but I took it as a compliment all the same
Though, I should probably point out that I don't run the courses with Inca myself, my partner takes on that role, and he does a much better job than when I try it (I lack the skills of direction and/or coordination  ).. I do help out with the training of individual obstacles though (much more suited to my level of capability ). |
Inca's Human wrote: Though, I should probably point out that I don't run the courses with Inca myself, my partner takes on that role, and he does a much better job than when I try it (I lack the skills of direction and/or coordination
).. I do help out with the training of individual obstacles though (much more suited to my level of capability ).Interesting! The owners of the two most accomplished OES agility males in the history of the breed in the US have a similar arrangement: the husband ran the dogs - he's a superb handler - but the wife taught the dogs obstacle performance. It worked for them. Very well, in fact
Kristine |
Mim wrote: Kristine, do you use jackpotting? I thought that it had been shown to increase speed and accuracy initially but then to cause demotivation in the longer term due to downward comparison ie now I want an even BIGGER reward. Or is jackpot just the term you chose to use and not in the operant conditioning use of the word?
Both
Yes, I do use jack potting as you describe when training something new and difficult (or retraining something we've neen struggling with) to reinforce the notion that this is what I want. I use it unpredictably and fairly rarely so I haven't noticed a trend towards demotivation. What I describe as 'jack potting" at the end of a run is predicable and more to reward with "more than one stinkin' little cookie" One suspects certain dogs can count and figure out that if they're not getting a cookie here and there throughout the run, they should get all of those cookies at the end of the run. Mostly feeding them multiple little treats at the end gives me a bit longer to tell them how wonderful they are.
Now, sometimes I will still use the jackpot concept while running. I have something extra special lined up and if they do something correctly that they have struggled with I may cut the run short, run to the cookies and give them something extra, extra special, say some canned cat food or smoked salmon or...whatever the dog thinks is WOW and which they very rarely get. It makes an impression. Over time running with you becomes self-reinforcing and you probably don't need to feed them half a steak at the end of a run. And in fact I have dogs who are rewarded with tugging instead. But I'll still usually give them a cookie too. Kristine |
| ^^^
what Kristine said! we will very occassionally drop a "jackpot" particularlly when learning something new. But most times we run - even in training - the run seems to be Marley's reward. She also loves to run and jump on me at the end to "celebrate" a run  :0 |
Didn't find exactly what you're looking for? Search again here:
Custom Search
|
|
| |
|
|
|
|
|