There is a trend in dog agility involving toy play that goes like this: your dog needs to play tug; he needs to play tug anytime, anywhere; if you give him the tug he must return it to you immediately; if you offer a tug to him, he isn’t allowed to say no thank you; and if he bites you that is totally 100% his bad.
As you might have guessed I have a few problems with this piece of agility culture. Let’s start at the top.
You were sleeping? TUG NOW
How many agility training videos have you seen in which the trainer is speaking to the camera, has just yanked her dog out of a relaxed state, and is expecting the dog to tug on a toy with everything he has? My answer to this question is too many. Not allowing our dogs to warm up let alone wake up is a terrible disregard for their experience. Expecting them to come out of the box “hot” has shaped our training and our breeding practices; and not for the better. There is a reason I am traveling the country this year talking to handlers of dogs who are worked up to point of being outside their own control. Hint: it’s not because some dogs are just difficult.
Rules and Play
Play is defined several ways, but the definition that refers to this act of “play” between two beings looks like this: to engage in activity for enjoyment and recreation rather than a serious or practical purpose. Of course if we use play as reinforcement in training there IS a practical purpose at work but the dog needn’t know about it. All the dog needs to know is that playing with his human is FUN. Should we have rules? Sure, most games do. But perhaps we shouldn’t be the only ones following them? If we are going to use play to our best advantage (and dare I say it, enjoy this act of reinforcing behavior!) we shouldn’t only think of ourselves when devising the rules. Sure, spitting the toy out when asked, grabbing the toy when invited, and avoiding teeth on human skin are helpful rules; but what about some rules for humans to follow? How about I will be consistent in my use of marker signals (telling you when you are invited to bite the toy), I will deliver the toy in such a way that makes it easy for you to avoid my skin, and I will let you win the toy as often as you feel necessary? Sounds like a more fair system to me.
Mechanics
A lot of agility trainers have very specific requirements in how their dogs tug. They want the dog to hang on tight, pull hard, not let go until asked, and return the tug immediately if the handler lets go. They want the dog to pop out of a sleepy or relaxed state and latch on like an alligator. I’m all for shaping specific behavior, but again I ask, what about the human half? Tugging our dogs side to side (rather than up and down) with only as much force as they are putting toward us is a good place to start. That means handicapping ourselves much of the time but not others; a tiny sheltie and a malinois require different types of human mechanics. We should also consider offering our dogs a horizontal bite surface whenever possible to avoid misplaced teeth. Adding consistent marker signals that inform our dogs of the location and type of toy we are delivering will help us even further. It is more fun to play if you both know some rules; and less fun if one side is the dictator to whom the rules do not apply.
Choices
Have you ever been made to play a game? Tennis in high school? A board game you didn’t care for with your family? Surely there was a time when you were coerced into a game, and surely you recall that the coercion made the game cease to be fun. When we “play” with our dogs but insist that they play; in fact giving them no choice in the matter, we remove any fun from the equation. We change play from a game to a chore. From an activity that is desired to one that is required. In essence, we wreck a damn good thing.
Reinforcement is Defined by its Function
Dogs play tug with us because that behavior has been reinforced. The behavior itself could be reinforcing. If it is, just the act of tugging will build the behavior into one that is robust and reliable. We should never have to insist the dog tug, we should never need to remind the dog that he does not have a choice (because just like he has a choice to eat the cheese we hand him, he should have a choice to play with us or not). Yet in agility I see so many dogs that are tugging due to reinforcement, yes, but negative reinforcement specifically. Defined as the building of behavior by means of removing a stimulus, the negative reinforcement procedures I see look like this:
- Maniacally batting a tug on the ground or waving it around until the dog latches on
- Putting pressure on the dog to engage in play again by using conditioned aversives (like saying “excuse me?” to the dog when he lets go)
- Hanging onto the dog’s collar until he chooses to engage the toy again
In each instance a stimulus was removed when tugging began and therefore tugging increased. Pay attention. You’ll see it everywhere. You might even be doing it.
Knowing Better and Doing Better
I used to teach toy play like this. It was all about “you have to play with me even if there is food in your face.” It was all about “get mad! Growl! HOLD ON!” The rules were tight and strict; no biting, not grabbing out of turn, and absolutely no delay in retrieve. I paid little attention to my mechanics. Idgie works great for toys but it is truly no wonder that she always preferred food when I look back on myself slapping her sides, growling, and pushing on her; insisting that she tug even if she’d really rather not for whatever reason. I am amazed she ever humored me at all. I do things differently now, as I want to be certain of two things: the toy play acts as a positive reinforcer in and of itself, and I can gather useful information about my dog’s emotional state through that toy play. If I remove choices, if I dictate tug, if I build this behavior through the skilled removal of aversives, I lose both of these.

Photo by H. Christenson