When Training at Home Backfires

Updated: Sep 2

You've been practicing Scent Work with your dog. Maybe you've been following the K9 Nose Work® training method that we cover in our Foundation Sniffing Program. Perhaps you are using an operant training approach. Or an entirely different training approach altogether. Regardless, you may be noticing an issue: your dog finds the hides perfectly when you're practicing with them at-home...but not so much when you are at a trial or when someone else is setting the hides for you. But why?! Let's discuss some potential causes. 

Before we start diving into this, I want to give credit where credit is due. All of these topics were raised by Dr. Lauryn DeGreeff of U.S. Naval Research Laboratory during her Chemistry of Odor, Odor Basics for Detection and Training talk at the 2019 NACSW/CNCA Symposium and webinar offered through NACSW.

What Are You Really Training Your Dog?

Whenever we are discussing dog training, clarity is crucially important. Remember: we are working with a completely different species, one that interacts with the world entirely differently than we do. 

With that in mind, we always need to ask ourselves: what exactly am I training, teaching or asking of my dog in this moment? Is what I WANT to teach, train or ask of them the same as what I am ACTUALLY teaching, training or asking of them?

This may sound silly or pointless, but taking the time to really evaluate what you are doing and WHY you are doing it and HOW it may go terribly wrong can save you a great deal of stress and heartache later. 

For instance, one of the examples Dr. DeGreeff discussed during her presentation was a dog who had supposedly been trained to detect twins, in that the dog could detect the odor of a twin person in a controlled setting. If true, this would be extremely interesting and demand further study! Only there was one problem...the manner in which the dog had been trained. 

You see, a ball box was used to deliver a ball to the dog when they located the correct twin odor. Great, reward at source! Here's the rub - there was only ever ONE ball box in the search area...and it was ALWAYS above or near the correct twin odor so that the dog could be properly rewarded. 


Our dogs are smart and contextual little creatures. Therefore, the dog in this example figured out just go to the place where the ball box is and BANG, reward delivered! If you moved the ball box AWAY from the supposed trained target odor - twin odor in this case - suddenly the dog was no longer alerting on that supposed trained target odor. Instead, the dog was alerting on the ball box. 

In other words, the dog was going to wherever the ball box was and twin odor was never part of the game in the dog's mind.


Again, it is not that the ball box approach in and of itself is "bad". In theory, maybe it could work well for a ball-driven dog. However, to make sense to the dog, you would need to have ball boxes in ALL the potential target odor locations and only deliver the ball when the dog found the correct twin odor. Otherwise, you are not training the dog to detect the target odor, you are training the dog to detect the ball box. You would also want to eventually fade the ball box or transfer the concept of delivering a ball another way that was easier and more convenient.

The take away from this example is the importance of taking the time to evaluate what you are doing in your training sessions: 

Are you actually training your dog to detect the target odor you think you are, or are you training them to do something else entirely?