New year’s Training Resolutions

I have been slack in completing my New Year’s Training Resolutions (what’s new? They are New Year’s resolutions) that I started on this blog in January. I completed the first two in that original post, and now included are 3-5, and so for completeness, I have all of them here in one place:

 

  1. Establish Goals
  2. Break Down Training into Steps
  3. Train from the result to the start (Back Chaining)
  4. Anticipate Outcomes and set up for success
  5. Use more Reward, and Compel Efficiently

Establish Goals

Too many police dog handlers fail to establish goals, whether for a training session they are about to begin, or for a more medium term goal such as the dog indicating on a high find deep in a building on a building search. As a handler you need to take control of your training program and understand why you are doing a particular exercise in a particular way. You should set up training to achieve specific, defined goals. Don’t just set out drug hides to run the dog on without regard for difficulty or objectives you may have in training. In every session you should be working on particular skills, such as the alert, or the searching behavior, or ignoring distractions you set up such as food or novel odors to proof the dog’s odor recognition. Training record forms set us up for lazy training: We fill in drug odors and amounts, and rate performance. Don’t confuse your records with training goals. Know what your dog needs to improve upon, and note it in the narratives, and then set up training to directly address those weaknesses. Set up training to reinforce strengths as well. You should be able to state your training goals with your dog for every skill set he possesses at any time. If you can’t do that, start thinking about it so that you can. Write down everything you want your dog to be capable of doing (within reason) and set those as long term goals. Then decide how to break each goal into a set of manageable training steps you can consistently train.

Break Down Training into Small Steps

 Let’s say a dog has an issue you are addressing in training. Say the dog is having trouble with the bark alert in the building search. If the alert is the problem, we need repetition of the alert to train and condition the response we want. So set up your session to address that specific issue. Having the dog search a giant building for one alert opportunity is inefficient and lacks focus, and is poor training planning because it fails to break down the training to focus on the issue your dog needs to have addressed. A simple search problem which will not tire the dog out and an easy find is what we want to create so we can concentrate on training the alert behavior. As the alert behavior becomes a habit, we can slowly make the search problem more complex, and if the alert maintains, we add further complexity. I hear so many times how handlers have been “working hard” on an issue and the dog isn’t responding, but working hard and working smart are two different things. You can dig a ditch with an ice pick, and yes, you will work hard at it and it will take a lot of time. Training must be goal oriented, and the session must be focused to achieve those goals, and this is accomplished by breaking the problems down into small manageable concepts and training in a progression, step by step.

 Train from the Result to the Start (Back Chaining)

One of the biggest novice mistakes in training is to train a sequence of behaviors in the same order that you view them in the final result. For example, when sending a dog to search for a suspect in a building search and perform a hold and bark on the hidden subject, the biggest mistake you can make is to not train from the result backwards to the start.

 Let’s take an even simpler version of this exercise, the hold and bark in a Schutzhund I. The dog searches 2 blinds one blank (first) then comes around and is sent to the hot blind where the decoy is passive, and the dog must bark at him for a fair bit if time before disengaging.

 I start with the dog on the back-tie developing the barking behavior – barking brings the decoy to the dog. The bark must be developed to the extent that the dog will bark at the decoy for the period of time defined in the rules, and this is developed with variable reward for the length of time the dog is barking with a grip reward.

 The decoy starts at a distance from the dog from behind or inside the blind. He moves on each bark in a quick prey-like step toward the dog for each, or every other bark, until the decoy is a step or two away, and on a good bark, give the grip as a reward. When the dog does this easily, every time, drawing the decoy our of the blind to him, and barking at him on a loose line when the decoy is within striking distance (this in itself is a process – getting the dog to settle in close), we can then transition to sends to the decoy. We start close to the decoy, first from only a few feet away (I like to do it on slick floors to control the dog easily without a lot of correction). Then the handler moves back from the decoy around the blind in the direction he wants the dog to run.

This is repeated systematically, using reward and correction (also varying fly-out bites with hold and barks of variable length followed by a reward grip), until the handler and dog are all the way to the start position called for in the trial rules – i.e., we have moved away from the decoy, around blind #5 (empty in the trial) back into the middle of the field.  The result, a hold and bark in the blind, is developed by starting at the result, and working backwards toward the start position.

 The building search for a police dog is developed with the same process, although there are some more variables that have to be accounted for.

Anticipate Outcomes and set up for Success

 A good trainer anticipates outcomes. Training is by definition setting up a process for the dog to complete where we, as closely as possible, set up the scenario such that we insure the outcome that we want, and we repeat this again and again to condition the response.

 For example, when training or maintaining a dog to out on command, I always have a line on the dog, ready for correction if the dog fails to comply, and I always have a reward (grip) available for rewarding the out when completed properly, no matter what scenario I am working on. Let’s say I am training a building search, and I send in the dog and he locates the bad guy, and we throw open the door for a grip, and then after a good fight we are ready to out the dog – but we have no way to enforce the command because there is no line on the dog – should you be surprised if he fails to out on command? The dog will learn in what context you are able to influence him, and in what context you are not.

 What I do is affix a short correction line on the dog once the dog is on the grip during the fight with the decoy so that either myself or my “back up officer” – just another trainer – on the apprehension can make a correction if the dog does not out, and if he does, the decoy will give him another grip as a reward for the release. No matter what I am working on with my dog in training, there must ALWAYS be a way to make the dog comply, and as well a way to reward compliance. I am ALWAYS anticipating the possibility of non-compliance, and setting up the training session to achieve the outcome I desire – clean outing in every context. I never rely on the dog to “be good” I always assume non-compliance, and am pleased to reward when that assumption is wrong! Imagine if your training as a cop was to assume suspects were unarmed, and rely on them to not take advantage of that assumption! Yet there are handlers who work their dogs with no way to ensure outcomes all the time in training!

 Use More Reward, and Compel Efficiently

 In 2009 I hope to get police dog handlers to use more reward in general. Forget about the old idea that praise is enough for the dog. This is outdated nonsense. When will you stop working for a paycheck and work for only praise from your superiors and the public? Tangible rewards (Ball, Jute roll, grips – or money for you) are far more effective in motivating behavior than praise. The idea that praise is enough is based on an ego-centric notion of the alpha – that your dog somehow lives to please you, the “great master.” Get over yourself right now, and give your dog something he really wants!  Not that praise isn’t nice, but to your high-drive K9, a jute roll and a vigorous game of tug is much more satisfying!

 And when you do have to correct, for a dog’s willful disobedience, do it efficiently. Don’t nag your corrections, do it once, get compliance, and then reward that compliance. Dogs will go down the easiest path. Make that path a clear one in terms of what to avoid doing (with correction) and what to perform (reward).

 Look for my article “The Power of Reward Part 1: Obedience” in the current issue of Police K9 Magazine which discusses in detail how to use reward in your obedience training to powerful effect!