You’ve probably heard the expression, Life Is Short: Play With A Dog. “Okay!” you think. “I’ll do it!” After all, dogs play together until they are gray in the face. Dogs also play with people, although that’s not always a given. Have you ever tried to play with a dog and it just doesn’t work? “The dog’s not playing right,” you may think. “This stinks.”

Don’t be so quick to blame the dog. Recent research suggests that it might be you who is not “playing right." I initially found the implications of this research rather benign, extending no further than a particular dog-human duo. But a study published in 2014 found that whether or not a dog plays with a person can be life-changing.

I’m getting ahead of myself. Let’s start at the beginning.

In 2001, animal welfare and behavior researcher Nicola Rooney — now at the University of Bristol — had a simple question about dog-human play. She and her colleagues wanted to know whether dogs respond to peoples’ play signals. In the study, volunteers played with their dogs for 5 minutes in the comfort of their homes, and the play sessions were videotaped. Owners were asked to play with their dog “as they usually did,” but here’s the key: they were not allowed to use objects or toys.

After the play sessions, the researchers watched the videos and noted which behaviors owners used to initiate or maintain play. They identified 35 common play signals, including patting the floor, clapping, shoving, hitting or tapping the dog, and of course, play bowing. People also blew at dogs, barked at them, and grabbed their paws. And who can forget my favorite behavior, ‘hand spider,’ where the “person moved their hand or fingers simulating movement of an insect or other creature.”

Identifying peoples’ play behaviors was just the first step. The real goal was to understand the effect of these signals: did peoples’ play signals lead dogs to play with their owners? And more specifically, did the commonly used signals elicit play more often than the rarely used signals?

Before spilling the beans, I want you to think for a moment how it feels when a dog plays with you. AWESOME is the obvious and only answer. It feels AWESOME when you are on the same page with a dog and engaging in mutual, reciprocal play. I remember a dog who was a bit wary of me when we first met. I’d walk into a room where she was sitting, and I’d receive a sideways glance as she slunk out. Ouch. Over time, ignoring her, letting her sniff me, and dropping treats on the floor paid off. The slinking and leaving decreased, and one day, we were sitting on the floor, and I got a play bow! A full-on play bow! IT WAS AWESOME! That is, until I felt a huge thunk of responsibility. What should I do next? What would be the best type of playful olive branch to extend to her? Many play behaviors involve touching, but surely we weren’t at the playful touching stage yet. Clearly I was overthinking this.

I think I responded with a play bow, defined as a “posture similar to a canine play bow, with palms and knees in contact with the ground, forequarters lowered and arms outstretched.” Or maybe an “abbreviated” play bow, with my hands on my thighs. I didn’t want to be too loud or too sudden. Knowing me, I definitely had on a play face. Much later, when we were in the ‘touching is cool’ phase, I remember responding to a play bow by head-butting her side. I remember thinking, "I don’t know if this will work, but it’s worth a shot."

Now that she and I are well into a stage of playing together, I have a better perspective on what our no-play days were like. Those days are marked by a haze of trepidation, the realization that we were figuring out what play would look like for us, and the awareness that what I threw out there would not necessarily be well-received. I wanted things to move forward, but even more importantly, I didn’t want things to move back.

I was right to be concerned. Of the 35 most common play signals, Rooney and colleagues found that a signal’s popularity “was not related to its success at initiating or sustaining play.” For example, patting the floor was used the most often, but play followed only 38% of the time. It appears patting the floor was not very successful at initiating or increasing play with a dog. Sad face. Other not-so-successful but commonly used play behaviors included scruffing the dog and clapping. Some things people did even elicited play 0% of the time! These included picking up and kissing the dog as well as stamping one’s feet. These “play” behaviors produced a net gain of zero.

All is not lost! A few behaviors were incredibly successful at eliciting play with a dog. Rooney and colleagues found that chase-and-running-away as well as lunging forward were associated with play 100% of the time! “Signaling up” (tapping one’s chest to entice the dog to jump up), grabbing or holding a dog’s paws, and of course, play bowing were also successful at eliciting play.

The researchers’ conclusion is somewhat somber: “We suggest that humans often use ineffective [play] signals.” Instead of blaming dogs for “not playing right,” people could look at their own behavior and evaluate its effect, acknowledging that certain play signals are more effective at eliciting play than others.

Bring me a playful dog!

Sometimes we hold dogs responsible when we don’t get the type of play we want, even subconsciously. One way to explore this phenomenon is to look at whether a shelter dog’s play behavior affects whether it will be adopted. Alexandra Protopopova and colleagues at the University of Florida Canine Cognition and Behavior Lab and Arizona State University Canine Science Collaboratory found that when a potential adopter takes out a shelter dog for a one-on-one interaction, only two behavioral variables predicted whether that dog was leaving the shelter: 1) lying in close proximity to the person, and 2) responding to the person’s play solicitation.

You heard me right: of all the things a dog might do when first meeting someone, only two things were associated with whether a dog was adopted: dogs lying close to the person were about 14 times more likely to be adopted, and a dog who ignored a person’s play initiation was unlikely to be adopted.

Taken together, these two studies paint a potentially scary picture for shelter dogs: people do not always use play signals that result in play, but people are unlikely to adopt a dog who does not respond to their play signals. Nobody wins.

When I think about dogs in the shelter going up for their one-on-one interviews, I hope potential adopters cut them some slack. The list of factors that could contribute to whether or not a dog will play with a new, strange human they just met are endless. On top of that, the shelter environment is often a weird, chaotic land (a complex topic for a future post). Go slowly. Keep your expectations in check. For shelter dogs, like speed dating, a lot is riding on the first encounter. Reflect on your play behaviors just as much as you think about theirs.


Keep up with the researchers whose work is mentioned above: Alexandra (Sasha) Protopopova (Twitter), Clive Wynne (Twitter) and John Bradshaw (Twitter).

References and Recommended Reading

Protopopova A. & Wynne C.D.L. (2014). Adopter-dog interactions at the shelter: Behavioral and contextual predictors of adoption, Applied Animal Behaviour Science, 157 109-116. DOI:

Rooney N.J., Bradshaw, J.W.S. & Robinson, I.H. (2001). Do dogs respond to play signals given by humans?, Animal Behaviour, 61 (4) 715-722. DOI:

Todd, Z. (2014). Adopting Shelter Dogs: Should Fido Lie Down or Play? Companion Animal Psychology Blog.