Occam's Razor and Agility

Okay, so you're probably thinking this is going to be a long, rambling and BO-RING blog post about scientific stuff.

Sorry to disappoint you. Well, some of you. The rest will be relieved to know it's fairly simple.

Occam's Razor (Or, Ockham, for the purists) states that, "If you have two theories, which, both explain the observed facts then you should use the simplest."

In other words, "When two options are presented and will have equal outcome, choose the simple one."

The term razor comes from the German "Ockam's Messer" which translates to "Occam's Knife" and refers to distinguishing between two theories either by shaving away unnecessary assumptions to get to the simple explanation.

This is my philosophy as an agility handler, if I'm honest. There's so much fuss in the agility community right now about all these fancy handling moves. Front cross, rear cross, blind cross, reverse flow pivot, the list can be endless.

I often joke my handling method is point and pray. I point at the obstacle and pray my dog does it. But there's more to it than that and a lot more factors. I'm disabled and unable to run very fast. So I had to add layering skills and rear cross confidence in my dogs' repertoire.

But the more I trial, the more I realize the handlers who do things with the most basics possible are the ones that excel. they scorn the "Style of the Week" handling fads and stick with what works for them. I'm not saying this is the best way, or the right one. But I think a lot of NQs and mistakes could be eliminated if we all went back to basics and stopped trying to do the nifty popular moves and stuck with the time tested and proven basics.

Why do a front cross if you can not do it and get the same result? Why do a reverse flow pivot if you can simply call the dog to you and push them away to the next obstacle?

Occam may have been talking science, but he could have been talking about any dog sport!

Until next time, folks.

It's Not Wrong, It's Just Not Right!

I've been doing some thinking about training my dogs and I had another AH HA! moment.

In clicker training, there is no wrong, there's only not right and try again. If my dog doesn't do what I'm asking, she's not wrong. She's just not right. So she willingly tries again - and again - and again and again to try and earn that C/T.

So. What does that mean?

It means the dog isn't wrong, she's just not right. Kind of like when you answer an essay question in school. If you make a mistake, it's not an automatic fail. You get partial credit. So you weren't wrong. You just weren't right.

Clicker training is exactly like that. We reward everything that's even the smallest bit right, so the dog continues to try. There is no wrong when the dog does not compete the appropriate behavior. It's just not right, so no treat. There's no punishment, which implies wrong. A lack of c/t does not imply wrong, it implies not right. And the dog has the opportunity to continue trying to get it right.

Once we take "wrong" out of the training equation, we're in much better shape to reward what is right, even partially right. Even if it's not right, it's never wrong.

Now if we could apply that to ourselves, our kids, our parents and spouses and coworkers and even strangers, life might be better.

Imagine a world where you weren't wrong, just various shades of not right. Wouldn't you be more willing to keep trying?

The "Ah Ha!" Moment

"Management before modification." - Steve White

"Any time you and your dog are together, someone is being trained." -Jean Donaldson

These two concepts, as simple as they are, simply floored me. It was an "AH HA!" moment.

Those of us with issue-y dogs often jump right in and try to modify behavior. We try to modify fear reactions, modify behavior, all without a thought to managing it first. Or we manage for a few days, a few weeks, a few months, then move on to modifying without benefit of managing anymore.

While this may not have disastrous results with shy dogs, undersocialized dogs, it can have devastating effects on aggressive or reactive dogs.

Lesson number one: Treat every issue-y dog as if it is reactive/aggressive and manage all social interactions accordingly. This gets you into the habit of always managing your dog's area, your dog's interaction, judging his/her comfort level. This teaches you to watch your dog, be aware of your dog, and not push your dog.

Why do I say this?

Well, if my dog is shy and not aggressive, I may be tempted to let "one more" person pet her, one more dog sniff her, walk one more block. If I manage that shy dog the way I would manage an aggressive dog, I would never, ever risk over facing her or going outside her comfort level.

Lesson number 2: Create an emergency "recall" word. This means you'll be conditioning your dog to hear a certain word as a certain awesome item is fed. (For example: I'm going to say "boo" to Brandy each time I feed her chicken liver. Pretty soon, when I say "Boo" she's going to run to me in search of that chicken liver. For her, chicken liver is the best thing on earth.) This creates a huge, strong, positive association with that word and it has been shown to help stop aggressive or shy tendencies when used in potentially damaging situations.

Lesson 3: Our dogs are always learning. Be an active participant each time you interact with them. Whenever you're together, someone is being trained and it may not always be your dog. This is yet another place consistency gets pulled out. If you want X behavior, insist on it. You can never be too tired, too stressed, too anything to insist on X. Or the dog learns X is not always required.

One time is enough to solidify something in a dog's mind if the reward is big enough. For those of us with counter surfers, we know that if the dog is rewarded 1 time out of 100, they will persist. Well, it goes to follow that if we insist a dog do X 99 times out of 100, that one time we don't is enough to reward him/her and keep them from being totally reliable.

Lesson 4: Dogs are dogs, they are not machines. While we all pray for 100% reliability, it's not going to happen. Be satisfied with 80% reliability. And before you test a new behavior, remember you only get 50% of that behavior in a new situation. So if you want to be able to nail a novice course at a trial, be sure you can do 28 obstacles in practice. This is true of police dogs, this is true of almost every dog. When the situation changes, so do the results of trained behaviors.

The Ramblings of a Dog Lover

Hello. My name is Deb, and I am a dog dork. Now before you screw up your face and go "Bwah?", I don't mean it in a pejorative sense. I love dogs, I love learning about them, meeting them, discussing them and learning how they learn.

I want to know, need to know, how their brains work, how they learn, why they learn, what drives them and motivates them. I want to teach them to excel, I want to show their owners how to reach them, how to teach them, how to create their very own Lassie.

But more than all that, I want to understand. What drives my dogs? What makes them happy, makes them sad, what turns them on, what turns them off?

This is a blog about all of that and none of that. Of our training and lessons, what I learn, where I succeed, and more importantly, where I fail.

Every failure is a lesson learned, another opportunity to succeed. Failure is always an option, it is only by failing we learn.

So. Onwards, dear readers. Whoever you are.