All of Maisy’s early training was done with lure-reward. As a novice owner, it was easy and intuitive, and Maisy learned quickly. I absolutely agree with Ian that lure-reward is the best method for teaching novice owners who just want the basics. However, lure-reward training is not without its critics, and I remember seeing people in my puppy class that struggled to get their dogs to work without food. Today, I’m going to share what Ian said about using lure-reward training correctly.
The first step is to phase out the lure as quickly as possible. The lure should be used briefly, as an instruction to do the dog, and you shouldn’t use a food lure more than six to twelve times before you move to luring with air-cookies. After another six to twelve reps, your air-cookies become a hand signal! Once you’ve reached this step, the treat should always come out of the bag/your pocket/wherever after the dog performs the behavior. Otherwise, you run the (very high) risk that the treat becomes a bribe, not a reward.
Next, you need to get the behavior under verbal command. Since lure-reward training so effective at getting behavior, Ian says that you can name it right away. However, he emphasizes standing absolutely still when you say the cue so that you don’t introduce any inadvertent body cues. After you’ve given the command, wait for half a second before you lure/use air cookies/give your hand signal. If you say the word and lure/give a hand signal at the same time, the dog will be focused on the food/your movement, not your voice, and as a result, is unlikely to pick up that what you said was a cue. When your dog responds to the word, and not your physical movements, have a treat party!
Now you need to motivate your dog to want to work for you. Treats are awesome, of course, but you need to phase them out, too. Ian does this in a number of ways. He asks for more than one behavior before giving the treat. He uses differential reinforcement to reward only the best responses (he recommends using a 1 to 3 ratio as ideal). And most importantly… use life rewards!
There are three main kinds of life rewards: playing with other dogs, walking and sniffing, and interactive games like fetch and tug. Ian said that if you have a dog that plays, you have a trained dog because it is so easy to use a toy as both a lure and a reward. (Incidentally, while he does phase out treats, he doesn’t phase out interactive games ever.) Any time your dog wants one of these life reward activities, ask for a behavior first. When he complies, release him to go play or sniff or chase the ball. Then have training interludes frequently (every 15 to 30 seconds of play, or every 25 yards or so when on a walk) to request a behavior before letting the dog return to the activity.
Finally, you need to figure out how to force a dog to do what you want… without coercion. This is where RRNR comes in, and since I’ve already written a whole post about it, I won’t belabor the point.
Following these suggestions will prevent lure-reward training from turning into bribery. My puppy classes with Maisy were very good about getting us to phase out both the lures and the treats, but didn’t emphasize verbal control- we used a lot of hand signals in class, and Maisy and I are still quite dependent on them to this day. We talked a little bit about using life rewards, but I wish we would have practiced the concept to help drill it home.
What about you guys? Did you use lure-reward to teach your dog the basics? How did you do at phasing out treats, getting verbal commands vs. hand signals, and motivating your dog to work? I’m curious to hear about your experiences!
15 comments:
For important commands (sit, down, stay, etc.) I did very much what Dunbar describes. I lured, then did "air cookies" (I love that term) and then turned that into a hand signal. I would slowly make the "air cookies" into a smaller and smaller signal so down first was my bringing my hand all the way to the ground, then not quite all the way, eventually working my way to a large pointing gesture and then a small one. I quickly fazed out using the treat in my hand signal hand, tough I admit that I'm SURE my dog can smell the treats on me! She knows they're there.
Once she knew the hand signals I worked on getting her solid with voice commands. I would say the command, wait a moment to see if she did it, then give her the hand signal. It didn't take too long before she was responding to the vocal commands alone.
My big difference (and maybe Dunbar also does this, I don't know) is that I never ever remove the hand signals entirely. I have whole sessions of "silent" training where I do it all with hand signals. And I do other ones with just voice and others with a mix. I never want my dog to rely on just voice or just hand signals for important cues. My main reason is old age. What if she goes deaf and cannot respond to vocal cues? What if she goes blind and cannot respond to hand signals? I really want to be able to use those important cues no matter what and so I consistently work those.
To add to that, some things, like loose leash walking and stopping on command I've worked 100% with life rewards (mostly continuing the walk or getting to run to me). She responds really well to games of tug and to getting to do something she likes. The "wait and come game" is one of our favourites. It's stop, stay, and recall all in one (I tell her to stop and then walk quite a distance away, then call her to me...she loves it!).
Now, this is all for the important things. For tricks I haven't bothered to fade out the lure. It's just not that important to me if she can sit pretty or stand up on her hind legs when there's no treat present.
A lure a lot of the basic tricks. But at the same time I'm doing a ton of shaping work to get the dog used to thinking. I haven't had a hard time getting rid of the lure because it does flow nicely into a hand signal.
I can't say I'm great at getting rid of rewards all together though. My dogs seem to get it as I don't ever have a problem with their basic manners without treats. I really can't pinpoint the stage or age when I stop treating and start expecting.
But with their comp obedience work I am horrible at fading rewards. I'm good with duration behaviors, but not with the other 90%! But knock on wood we haven't yet ran into a problem in the ring, at least not since I started adding in tricks in between exercises with Lance. I'm sure it will come to hurt me as I keep showing and eventually start working on that UDX, but so far I think all the rewards are transferring to making the exercises themselves highly reinforcing.
I'd love to see Dunbar and others work with some of the puppies/dogs in my classes....
There's a surprising number of dogs that just aren't responding well enough after 6-12 reps to be able to even do a fake-treat. And that's not counting the occasional dogs where it takes weeks to even get a sit or down!
Blaze learned with lures. It was hard to fade the lures and to get the behaviors on cue. Luna learned with capturing/shaping only and later I had to train her to lure. Griffin has food magnets (a specific type of lure). He's the best of my dogs at verbal cues, but he also has the benefit of a more experienced handler!
CRYSANIA- I'm not sure if Ian keeps the hand signals or not. I imagine he does, but I can't remember if he said one way or another. He did say you want cues on verbal only for times the dog isn't looking at you. But, I take your point re: hand signals. Both of my friend's dogs have gone deaf, so she relies on hand signals for them (although one can, apparently, read lips. It's fascinating).
LAURA- I'm horrible with duration stuff. And, honestly, I haven't worried too much about fading out treats. We can use them in APDT, after all! Plus, with a reactive dog, the odds are high that I'll have food with me at all times. That said, Maisy is quite willing to work for verbal praise, so I don't NEED food for her.
KRISTEN- I totally hear you on the down. I've got a friend whose dog could NOT be lured into a down. She had to capture it. As for the rest... how many reps have you seen it take before you can lose the lure for some/most dogs?
I did some luring this weekend with a friend's dog and your previous comments on it from Dunbar's seminar came to mind as I was doing it. I think I usually rely on the treat lure too long. Actually, I know I do. This time, I made sure to get rid of it ASAP. She is a super smart dog, so I'm pretty sure I was at the low end of the 6-12 before I started empty hand. She caught on to that quick. Then I started naming it and we only did a few reps before ending the session. We were working sit pretty (and she already had the muscles and coordination to do it).
I agree with Kristen though--while lure-reward might be easiest and most intuitive for new trainers, it does not work equally well for all dogs.
I'm also with Laura on getting rid of rewards altogether. My dogs' basic manners and recalls are pretty consistent without any treats ever, but tricks and more complex behaviors . . . if I don't have treats I'm not going to get much.
Oh, and one more thing. This quote . . .
<>
I think you'd probably agree based on your experiences with Maisy that it's not quite that easy!!
NINSO- Your quoted material didn't come through... I probably agree (it's been a difficult journey), but I'm not sure what I'm probably agreeing with!
Oh weird. I was referring to this:
Ian said that if you have a dog that plays, you have a trained dog because it is so easy to use a toy as both a lure and a reward.
I've found that html doesn't always work in comments, so if you tried to do the italics tag or something, that might be why.
Anyway... I think that, like so much, it depends on the dog (and the toy). But yeah- it's not so easy for Maisy and I, especially if it's a tennis ball!
I used primarily luring from the beginning of my training with Shiva, mostly because that's what the books told me to do and I didn't know anything different. Perhaps I am just really lucky and have a very motivated dog (I am and I do) but I didn't find fading out the lure very difficult at all. It didn't take very long before we were able to string a few commands together and only reward at the end. This may be harder with my next dog, I don't know. Or maybe it was more simple when I didn't know what I was doing and just went with it?
I try to use real-life rewards as much as possible. But it's hard because Shiva will only accept some rewards in certain places. For instance, in a class setting, a toy is not good enough and she will blow me off, but at home or outside she loves to tug. It's always complicated, isn't it? So much depends on the individual dog and nothing is set so that it works for everyone.
Anyway, thanks again for sharing all of this valuable information!
Kristine- you're quite welcome! I love the conversations in the comments. I really only know how one dog works, and while I can draw conclusions from that, I love hearing how things work for other people/dogs.
You know if you ever want to borrow one to try training a different dog, well, I have several unmannered heathens!
I haven't counted, and I don't know how many or how good the reps are at home. But sometimes it can be weeks before we're getting the behavior of the dog immediately following the lure into position. I typically can get it faster than the owners, so after I get it a few times I pass the dog back to the owner. But with some dogs we're feeding approximations for a really long time.
I like to see the dog immediately following the hand before getting rid of the lure. If the dog is hesitating or really slow I won't fade the lure yet.... because if we tried the dog most likely just wouldn't bother following the hand that doesn't have a treat in it (and with his slower motion he KNOWS it!).
DAWN, I love your dogs. Maybe Magic would like to come and stay for an extended board and train? I can already tell that he's going to need A LOT of training, lol.
KRISTEN, on average, how many of your students struggle to get their dog to follow the lure? Half? More or less?
I do think there's a skill to luring well. You need to get it in the right place, and that's something you need to teach new dog owners.
About a third of the dogs will not go down with luring alone. They need shaping-luring or a leg-chair type barrier.
3/4 of the students have a hard time with luring. They want to hold the treat too far away, making the dog frustrated, making the dog take the treat harder, making the person want to hold teh treat too far away. That's when I pull out something like cream cheese. Lick lick lickckkkk luring.
We teach it treat-magnet like in Agility Rigth from the Start, but people really really don't like dogs licking them!
I always find it strange that people don't like puppy kisses... I love them!
Thanks for the info. Interesting that so many people struggle with luring.
Post a Comment