Double Down on Training: How to Make Your Dog a Treat Gambler

Teaching your dog to gamble for their treat can pay off with increased desirable behavior. Photo – AdobeStock

To treat or not to treat? That is the question. 

As a trainer, I’m a strong advocate for “when in doubt, treat.” Treats are the keystone for positive reinforcement training and the hallmark of operant conditioning for all animals. Animals are universally food motivated because they (and we) need it for survival. Ever since Edward Thorndike coined the phrase “law of effect” - which basically states that an animal will do more of what is reinforced and less of what is not - behaviorists have understood the importance of positive reinforcement in training behaviors we want to see.

This discovery is a huge gift to those of us who want to have a connected, communicative bond with our furry friends because it helps us understand how to establish and build this kind of relationship. In short, you give the pup what he wants for doing what you want (aka the Premack principle). Your dog will invest more focus, energy, and enthusiasm into training when yummy reinforcement is involved.

But what does this have to do with gambling? The truth is these factors - focus, energy, enthusiasm - will actually increase when there is an unknown variable in the reinforcement. Like humans at a slot machine, dogs will keep trying, not only because they want to win, but because they can win surprising amounts of good things, but they’re never sure when. Think about it, if slot machines never paid, we wouldn’t play them. If they paid every time, we’d be pretty bored (because they’d never pay us more than we put in or the house would go broke). No, it’s because they pay intermittently and with variability that we keep playing!

To make your dog a good gambler, start by treating your dog consistently for the desired behavior. Then, once a behavior is generalized (meaning you’re providing treats for the desired behavior in any/all contexts), then it’s time to move to intermittent reinforcement (where the desired behavior is rewarded some, but not all of the time). Keeping your dog guessing - like a gambler at a slot machine - will actually help maintain the desired behavior. 

Naturally, you don’t want to overfeed your pup while training so to ensure you’re not going overboard, portion some treats out of their regular meals. Also keep in mind that treats only need to be  ⅛ of an inch in size. All dogs need is a taste and that’s enough to get them in the game. 

Of course, you can also substitute other favorite things, such as attention and play, for food treats, or use a variety of things in one session: cheese, tennis ball, praise, and so on. But it’s this intermittent reinforcement schedule (in which behavior is reinforced after a random number of responses) that will keep your pup’s attention. ( Annika Bremhorst, Sarah Bütler, Hanno Würbel & Stefanie Riemer. nature.com/scientificreports )

Varying the type and quality of treats is important, too. When teaching recall, for example, I have one recall word and three types of treats: one low-quality, such as kibble; one medium-quality, such as freeze dried liver; and one super delicious high-value treat, such as cheese. When the pup is wandering and sniffing, but does something good (like a friendly greeting), I give kibble, the low-quality treat. When the pup checks in without me asking, I give dried liver, the medium-quality treat. But when I say the recall word and the pup comes, which is the behavior I want to train, I give the cheese, the highest-quality treat. 

The dog now knows there are different types of reinforcement of varying quality on hand. He becomes instantly more alert because he wants to figure out what to do to get that delicious cheese. “Not every behavior gets it. What does? Oh! It’s when I go to her when she says that one word. Well, I’ll do that every time now, and more often because I want to get that cheese!” 

This approach also increases self-motivated check-ins, because that freeze-dried stuff is pretty good, too, and because the pup is hoping he might accidentally get some cheese. The dog becomes more motivated to be near us, to be connected to us, and to be more focused and responsive overall. The bond of communication and trust has been set and is being strengthened. What a reinforcing thing that is for both of us!

The key is to treat consistently until the behavior is generalized, but once it is, you want to introduce the gambling aspect. Will she treat me or won’t she? Will it be kibble or, omg, cheese!? By maintaining the gambling element in your training, your pup is more likely to keep trying because it wants to win that grand prize.

Pulling that lever and winning the jackpot at the casino is mutually beneficial to the customer and the house (after all, who’d go to a casino where no one ever won?). So teach your pup to gamble through variable reinforcement. It’ll be a win/win for both of you.