The ramifications of a variable reinforcement schedule for Nosework dogs
I noticed a strange phenomenon when I was judging recently that really stuck with me! In the event of more than one hide, I noticed that some handlers were choosing not to reward their dog until the last hide. Perhaps they were opting to use “Variable Reinforcement” with their dogs? In the end of course, we all make our own training choices, but this struck me as a little odd. After some reflection, I wanted to write about what I saw and what I believe could be the downstream ramifications.

I can only assume that in a sport where we are always allowed to reward our dogs for performance by either giving food or playing with a toy, that there must be a reason why someone might not want to reward at a hide when there are still more hides to find.
Some reasons why handlers may only feed at the last hide are…
(1) In this case, the handler is not feeding at the first hide in order to reduce their overall search time.
(2) The handler is using a form of “variable reinforcement” to keep the dog guessing, believing that variable reinforcement will strengthen the behavior.
(3) Whoops! The handler simply forgot to feed!
Ok… so if we eliminate #3, it’s still an interesting conundrum! Let’s explore the potential ramifications of not feeding at interim hides and how variable reinforcement fits into the sport of Nosework.
I ended up discussing this with my friend and colleague, Dr. Deb Jones, Ph.D. pretty extensively. It’s actually fascinating when you get into the science of it. (Deb is a psychologist, learning theory expert amongst other things. She has a GREAT blog at: https://k9infocus.com )
So, if we explore #1. The handler is trying to reduce their search time by not rewarding interim hides…
It’s interesting actually. Delivering a food reward usually takes between 3 and 7 seconds for a relatively fast reward delivery team.
Here’s an example of pretty fast reward delivery. This is me handling Judd during an NACSW Eilte level search. Because his odor obedience is excellent, and because he doesn’t actually have a “Final Response” per se, I’m pretty confident with feeding him away from source. This results in faster reward delivery. With dogs that DO have a Final Response or who are a bit greener, I would always feed at source!
I definitely understand the desire to shave seconds! In Nosework, mere seconds can separate placements!
In the case of rewarding however, I think it’s best to minimize the time it takes rather than eliminate it all together. It’s a matter of thinking long-term versus short-term. In the short-term, minimizing the rewards WILL shave seconds. However over time, the final result will be longer search times and potentially missed hides.
Without getting into variable reinforcement yet, what could be the downsides of only feeding at the last hide?
- Your dog could decide that the initial hides aren’t worth as much as the last hide resulting in “walking on odor”.
- Your dog may eventually become more easily distracted at the beginning of a search due to a lower rate of reinforcement.
- Without a reward, the dog may become “sloppy” in his precision, resulting in fringe alerts.
- You may actually confuse your dog, leading your dog to believe that you don’t actually want him to alert on a hide (since he’s not getting a reward!) which will lower his self confidence, cause missed hides, and ultimately slow the dog down.
All of these potential ramifications are pretty serious! They can result in a gradual drop in Q rate and ultimately may cause the dog to “top out” more quickly instead of continuing to move up the levels successfully.
If you remember my last blog, “My Dog Lies!“, we talked a lot about building trust between the dog and the handler. A big part of retaining that trust is keeping the lines of communication open. Your dog needs to understand when he is correct! That information will increase the likelihood of a correct find in the future.
Some handlers my believe that a Variable Reinforcement schedule will increase the dog’s speed to the hide
Between the desire to reduce search time and a somewhat dangerous basic knowledge of learning theory, a handler may believe that a Variable Reinforcement Schedule is more powerful than a Continuous Reinforcement Schedule in the sport of Nosework. It’s not a bad idea. The idea is just misapplied in this situation.
Part of the misconception of the reinforcement schedule for nosework, may come from thinking that always rewarding for a find will decrease the sense of urgency for finding the hide.
To make matters worse, consistent reinforcement often causes behavior to decrease in frequency and quality. Since the dog knows that he will always be rewarded … eventually, there’s no need to hurry and so, the dog eventually does it in his own way, in his own good time and his slow sloppy behavior gets rewarded. Spoiled dog syndrome.
– Dr. Ian Dunbar, “Reinforcement Schedules”
And… do you remember the tenant that you must ALWAYS reward after using a marker (either verbal or a clicker)? Remember that in Nosework competition, the dog is always verbally marked. Dogs are no fools! They understand that the word “Alert” is a secondary reinforcer and precedes the reward! No reward just means a disappointed dog.
A Variable Reinforcement Schedule is like a slot machine rather than a vending machine. If a dog is fluent in a behavior, a Variable Reinforcement Schedule can increase behavior.
But what about Variable Reinforcement for Nosework?
However, it would be a mistake to compare the complex finding and sourcing behavior chain to a simple behavior where Variable Reinforcement is easily applied (like a simple Sit or a Down). Liken doing a search to performing an entire chain of Obedience Behaviors. It’s difficult to go into the Obedience ring and not have a single miss! The lack of reinforcement between exercises is a contributor to missed exercises during Obedience competition. In fact, working a dog off of the reinforcement is a careful process and the dog needs to understand that reward is inevitable, if not delayed.
First of all, there really isn’t a reason why you would ever want or need to reduce reinforcement in a Nosework trial. Second, unless you very systematically work towards delayed reinforcement, you are just plain reducing reinforcement (and what dog likes that?).
Sourcing a hide is extremely complicate and requires a ton of effort. Dogs who are working odor, especially multiple scent cones, are in constant analysis, thought and discovery. Odor isn’t like a tractor beam. It doesn’t just pull the dog to the hide. The dog has to truly work for the find!
Also, consider this. Variable Reinforcement is already indirectly built into Nosework!
What??? Yes, that is correct! Variable Reinforcement is already built into the sport. Consider this… Your dog enters a large area with very little odor. He has to seek and hunt, clearing unproductive areas. Eventually, he will encounter a scent cone. THAT is variable reinforcement! This is true, provided that the sourcing of Birch is treated as a secondary reinforcer. Meaning, that it results in a reward.
Realize that the dog isn’t hunting a chipmunk. He’s hunting Birch. Birch itself has no intrinsic value. Yes, he thoroughly enjoys the hunt, but without the reward, Birch is actually pretty useless! If we break the association between the predictor, in this case Birch, and the reinforcement, we will eventually extinguish the value of the predictor. THAT my friends, is the start of the end of a Nosework career!
The punchline is…
Never miss rewarding your dog for a hide!
In order to maintain motivation for the search:
- ALWAYS reward for every find.
- Use high value rewards. If your dog has arousal issues with super high value, you can use a “lesser high value” reward at interim hides. (This works very well for “Toy Reward Dogs”)
- When the dog is green, ensure that you have little time between each find. You want a high rate of reinforcement. (I consider dogs to still be “green” until they are easily running Elite-type searches. Your AKC Advanced or Excellent dog is still VERY green!)
- Even if your dog is very experienced, it’s critical to not ask your dog to do long drawn out searches without finding and sourcing a hide.
- Jackpot your dog and have a party after every search!
Motivation and drive for odor will speed up your searches. Focus on that, rather than trying to shave seconds off by skipping the reward. Your dog will thank you!
Very well stated in the last 6 years of judging I have seen handlers not reward, some just forgot and others wanted to shave time off.
I completed my NACSW Championship with one major change in the last 2 years, using a walker as I have functional mobility disorder. My dog who in the very beginning was extremely environmentally sensitive was able to come to me when she found her hides. I believe she knew that no matter what she would get rewarded analyzed the situation and started coming to me,
I will never forget the way we completed our Championship. Interior large locker room with bathrooms, very large shower 2 hallways and at the very end a weight room. I knew I had to clear each area as in the weight room had to lift walker so I couldn’t return too hard. Ptaary was incredible finding hides returning to be rewarded. Come to find out we were first by a minute + and I rewarded at every single hide. So I firmly believe that the reward reinforcement has helped her become a very strong searcher never giving up.
Totally agree! I think if I reduced Judd’s reinforcement, I would see a significant drop in drive for odor…
Another reason I found a handler may only feed at the last hide is arousal for the reward. In this case it was a ball toy. Once it came out the dog was too over aroused to keep working. I saw the same as you did at your trial – the dog started to walk odor and lose value for odor. I didn’t accept this as a dead end reason and suggested using a lower arousing toy. They used the leash as a tug and rewarded at each hide. It was a great resolution! Also, at the higher levels you won’t know how many and when the last hide is! So this concept doesn’t really work for many reasons!
Ha! That’s too funny… yeah I had that as one of the reasons but took it out because it made the blog too long! I’ve done that in the past with students… lower arousal toy for interim hides, jackpot toy for Finish. It works great!!
I love the idea of odor being the secondary reinforcer!
Interesting and I wholeheartedly agree with you BUT we do have one dog that comes to me for training who gets very frustrated stopping for sweeties and is difficult to refocus on the remaining searches, we assumed that the search itself is so reinforcing she prefers to just carry on (the dog gets to decide what is reinforcement after all) any suggestions for this scenario?
This is actually a pretty common problem! However, doing what the dog wants, and forgoing reward in this instance will develop a dog into a dog who “catalogs” the hides before deciding to start indicating. What I would do is reduce the number of hides to only 2 or 3 and I would pair the hides with food for at least a couple of weeks. That is usually enough to fix the problem! Good luck!! 🙂
I remember my first advanced containers search. Newt quickly found the first hide. I called alert and since looked happy and he looked ready to go I moved on. Lol his ready to go look changed to confusion and he gave me a strong alert on a container a few containers away. Clearly I had confused him on what the game entailed. I am a firm believer of you finf it i reward you.