Harnessing the power of the lottery for the greater good!
The Jackpot Effect was first talked about in Karen Pryor’s book, “Don’t Shoot the Dog!” In it, she describes the Jackpot effect as how willing your subject will act after you provide them a large reinforcer, ie. a jackpot.
On the science side, however, Karen Pryor notes that things aren’t very definitive. The science behind the jackpot effect ranges from being fuzzy, to pointing out how the jackpot effect doesn’t in fact exist. Despite its controversial nature, the jackpot effect is still something worth talking about since a bunch of animal trainers, myself included, practically swear on it.
Let’s dig into these topics in a little more depth.
What is the Jackpot Effect?
To be more clear, the Jackpot Effect is specifically when you reward your subject with a larger than average reward for a wanted behavior. By larger, I mean extravagantly larger than average. I personally define it, in the world’s most unscientific way, as “the jackpot must be large enough to make my subject give a figurative ‘holy shit.’” In return, the subject usually shows an increase in activity and more compliance with later requests.
In comparison, Karen Pryor defines it, in a more science-friendly way, as:
“A jackpot is an unexpected, high-value reinforcer used sparingly and contingently to reinforce a significant breakthrough in training.”
What’s important to note here is that it’s used sparingly and mostly for breakthroughs in training. Ideally, this cements into your pet that you absolutely adored what they just did, and your pets in turn will want to do that difficult behavior again.
In my own experiences with dogs, you’ll also want to keep jackpots to a minimum since these reinforcers are a huge intake of calories and if overused, it’s very easy for your dog to become overweight.
In her post about the Jackpot Effect, Karen Pryor notes how this was also used in sea parks to keep dolphins and orcas motivated through high-value treats. When they performed difficult behaviors or if they had a breakthrough in the training, then the dolphins would get a high value treat like a mackerel. Due to the high fat content in those fish however, each dolphin was limited to 1 per day.
In that same sense, this can be applied back to dog training by keeping in mind which of our treats are considered high-value by our pets. For my past few foster dogs, I’ve used peanut-butter filled kongs, or peanut-butter snacks as a high value treat. Most of the time, however, it takes some testing with my dogs to figure out which treats they consider high-value and which ones they consider low-value.
What Does Science Say?
There’s been some discussion on whether the Jackpot Effect is real at all. Accounts that claim the jackpot effect is useful are usually from animal trainers who will swear that jackpots help keep their subjects motivated.
From my perspective, it’s difficult to measure out what a “jackpot” is. Some animals won’t care about the volume of the reward they’re getting, while others will care a whole lot – meaning some animals won’t really notice a difference between a regular reward and a jackpot, or they’ll receive a jackpot and not care. This makes scientific measurement extremely difficult because you have to figure out what counts as a jackpot to your subject.
One particular study caught my eye about what the jackpot effect could be doing. It points out that pigeons who were signaled about when a jackpot was active would be more willing to peck a button for food. This study even increased the cost of getting jackpots by increasing the number of pecks required on average and found that the pigeons were always willing to try and get the jackpot as long as there was even the slightest chance of it. The study then reversed the signalling, and instead signalled when the chances of getting a jackpot were much lower, and noticed that the pigeons would only peck for a jackpot when there was a chance.
This study intrigues me because it shows that most animals are aware of what a jackpot is, and, despite being unsure when they were going to receive the jackpot, would still continue to put in the effort as long as they eventually got the jackpot. In my mind, this study at least provides evidence that the potential for a jackpot is enough to elicit repeated behaviors from a subject.
For people who are hoping to use these concepts in their day-to-day training, this study shows that the Jackpot effect can work. I’d love to test and see if signalling that a jackpot could happen will keep our animals more interested in training for longer. In that sense, I would hope our pets would act like the pigeons in the study and give us more behaviors in exchange for a chance at their figurative jackpot of treats.
Disclaimer: Before I start this story, I do not recommend doing exactly what I did with this dog. I’m just sharing a story of what I did and why I think it worked. If you are trying to be friends with an abused dog, please consult someone who has either worked with the dog before or has more experience than you in training animals. My approach was supervised by a more seasoned volunteer and I had also consulted other trainers about the behavior of this specific animal before I interacted with her. Alright, now on to the story.
When I was working in the shelter, there was yet another Akita I loved to bits named Elkie. She was extremely shy. Most dogs, when you go into their kennel, are excited just to get out. Elkie, however, would cower into the opposite corner from you. I would find out later, this is because Elkie was rescued from an abusive home that would coop her up in a room and then beat her if she ever acted the way a dog normally would, like bark at things or jump on people.
My approach with Elkie was to open the door and just sit at the opposite side of her. This way, she could approach me whenever she wanted and, since I was sitting down and making myself look as small as possible, I was very little threat to her. I made completely sure not to ever look at her, since this could be taken the wrong way by nervous dogs.
Luckily, despite everything that had happened to her, Elkie’s previous owners couldn’t beat the curiosity out of her. After I sat in her kennel for almost an hour, Elkie got curious and would nervously approach me. From the gray patch of the floor I was staring at, I could at least hear the soft pitter-patter of Elkie’s nervous paws against the cement every time she took a step closer to me. Each time Elkie would approach me, I would toss a treat her way. After that, I would toss another treat whenever she got closer than her previous attempt.
When her nose finally touched my pouch of treats, I gave her a jackpot – her reward for reaching the strange human huddled in the corner of her temporary home. From that point on, she would zip straight to my pouch and touch it with her nose as many times as possible. While I think shaping and patience did most of the heavy lifting for this exercise, I also believe the jackpot helped ease some of the tension between Elkie and myself. Her renewed enthusiasm also allowed me to start getting her used to human hands again. I put a treat in my hand and just let her eat from it, and other times I would put my hand in front of her and let her just touch it. I believe this would have been possible without the jackpot, but the jackpot in this scenario made her more willing to let me try these exercises with her.
The jackpot effect occurs when you give your subject a huge reinforcer. While the exact size of a jackpot is hard to determine, I usually go with enough to shock my subject. Ideally, this encourages your subject to be more motivated with you as training progresses.
Keep in mind, the jackpot effect is not proven in science yet. In fact, some studies will even say that the jackpot effect does not exist. However, I personally think the additional encouragement through a jackpot or a high quality treat may be useful to know for your pet.
In the case of Elkie the Akita, I used the jackpot effect to help motivate her and help her feel more at ease around me.
Thank you for reading!
If you’re interested in more content, please feel free to subscribe to this blog. I usually create a post every wednesday or friday, but I may scale this back to once every two weeks.
Thank you again, and I hope to see you all next week!