Placement of Rewards
If you want beautiful heeling, there's something you need to know.
It's more important than your reward structure. It's more important than verbal markers or clickers. It's more important than throwing corrections into your training.
It's where your dog receives the reward for their work, regardless of the quality of that work. Where that reward process starts and where it ends. It's perfectly fine to reward a dog that is out of position; just make sure that the reward placement makes them more likely to be correct the next time. And dont' mark the incorrect work; just reward in a manner that fixes the mistakes. Dog is lagging? Throw the toy straight ahead. It's ok if you throw when they're still lagging; they'll get better and better even if you do nothing else.
Many trainers believe it's not important -that as long as you mark a behavior with a "yes" or a clicker then you can give the reward anywhere you want.
They're wrong. It's as simple as that. You see, I used to believe that too; all that mattered was the marker - the "yes" or the "click". Then I battled a forging problem for a few years, until a pet dog trainer suggested that I change the placement of my reward. I wasn't inclined to listen to her ("just" a pet dog trainer), but to be polite I decided to give it a try.
It worked in a matter of days. She was right. I hate it when other people know things that I should know.
Markers are nice, but placement of reward is far more important. This is brought home to me every single time I teach a seminar or train a dog in a private lesson. If I could have either a marker OR excellent reward placement, I would take reward placement, hands down.
The rest of this blog is specifically for Gretchen and her wrapping, forging, crabbing Rottie. Gretchen, once we change the position of your reward in heeling, you will stop tripping over your dog in a matter of days.
The reward must be given in a position that inconveniences the dog if they are out of position. Forging? reward behind. Crabbing, reward to the dog's left. Wrapping? Behind and to the left. This technique works for all training, not just heeling, but today I'm going to demonstrate heeling.
Almost all of my own dogs will develop a tendency towards these sorts of problems - it's normal and typical of dogs that are trained in drive. They like the game. They want to watch your face. They want to see the toy you are holding. So they get closer and closer to it - I doubt they even know they are doing it. If I lose vigilance for any period of time then these problems will come back, so correct reward placement never ends.
Here's a video of correct reward placement for a dog prone to wrapping, crowding and forging. I also do specific moves in heeling which help Lyra learn to control her body, but this blog isn't about that today, though it is certainly demonstrated in the video. This blog is about rewarding her position. You'll notice Lyra's position is pretty good; not perfect but on the way to being both accurate and pretty.
The first minute or so demonstrates reward position when I use a tug - I have her spin away from me before she gets the toy. The second part demonstrates reward position when I throw the toy - to her left and slightly behind. If she were seriously crabbing, I'd change the position of the reward to be further back, so she'd have to turn even harder to the left to get the toy.
If I were using food the same principles would apply. Feed where you want the dog's head - high up and on the OUTSIDE of the head, if your dog is likely to wrap. Or throw the food. Behind, to the side, or both.
Not sure why people think you can mark a behavior and not give a primary infocement right away. I have trained exotics at Busch Gardens for 33 years and have use operate conditioning and a clicker as a machanical bridge since 1977. The bridge is to mark that moment in time that the desired behavior is offered followed up by your primary reinforcer. The clicker is a secondary reinforcer. Offering primary as soon as you click is very important until the animal is ready to start using a variable schedule of reinforcement.
There are so many people using clickers that really have not had enough training, and then end up training a different behavior. I have seen it many times when training people in the zoo field as well as search and rescue.
What happens when your timing is off and, or not being able to move beyond what you are curently doing to the next step soon enough. ( shaping the behavior ) is that you get sloopy behavior, ( Dogs breaking sit, heal, sitting on target scent, what ever. ) So don’t put down the clicker because it is still a useful tool as it was when B. F. Skinner invented it in 1937.
It like most things in life are only as good as the person using it. If used correctly you would be rewarding over the correct position.
Let me also add that Skinner did not invent the clicker, but did invent operate conditioning and the use of a mechanical bridge. For the purest out there.
Ditto, what Nancy said! Online heeling class would be awesome…or maybe do a heeling seminar in SoCal? :)
Early on in instructing agility, the dogs taught me the significance of where the handler rewards. I believe operant conditioning pioneer Bob Baily [think of training ocean swimming dolphins for the military]has a saying: “click for action, Reward for position”. U rock Denise!
Lynnda L in Minneapolis
@Heather: They do that too where I train except the decoy is holding a Kong in his fist (so no lure) and rewards the dog by throwing the Kong.