Here is good article to remind all of us that we need to always be fair to our dogs, especially in training. Striving to be the best teacher for our dogs starts with recognizing our own human shortcomings-
trainer skills
One trainer’s back to school review…
While I don’t always use a clicker I do always use “force free, mark and reward” as the basis for all of my training. My dogs know 2 markers from me: “Yes” or a clicker. Here are basic reminders that I tell myself at every single training session to keep me on my toes:
Reminder-“You get what you click for” Consistency and timing of your mark is everything. Once your dog understands your mark (word, click or other sound), they will know R+ is coming for the behavior that was marked. The most effective reinforcement is delivered within 3 seconds of the click or mark, and every mark or click must be consistently followed with reinforcement, even if you are wrong. So what if….
If you mark the wrong behavior by accident you must still back it up with R+. Trainers make errors, dogs don’t. It’s not the dog’s fault that you clicked/marked in error, and they will expect their reward. If this happens, just use a less valuable reinforcement with no fuss. To not R+ at all will confuse your dog and devalue your mark. Just move on and improve your timing to avoid those errors as much as possible (but it will happen to all of us on occasion!).
If you are slow on your delivery of R+, praise your dog immediately after the mark/click to keep them engaged until you deliver the treat or toy. This bridges your praise with the primary reinforcement (food or toy) and helps close that 3+ second gap to the goodies. (We all at times get our hand stuck in a pocket!)
If you are too soon with R+ and have it in view before the mark, then it becomes a distraction from the mark/click, essentially a lure and a bribe which takes the focus away from the task and distracts the dog from learning and thinking. R+ should be out of sight (pocket, bait bag, etc.) until after the click.
It’s not about the food and toys: I train with relationship being the priority and I want my dogs to enjoy training as much as anything else we do in life. I always invest myself along with any food, toys, etc. by praising, petting, or anything else that I know is reinforcing to my dog.
So in a nutshell:
- Be consistent with the mark and know what you want.
- Timing should be precise.
- If your timing is off, still R+ to be fair to your dog.
- If your R+ delivery is slow, add lots of praise in between your mark and delivery.
- If your R+ is kept in view, it can make your dog lure-dependent and distract from learning. Keep R+ out of sight until after the mark/click. (pocket, hidden bait bag, etc.)
- Remember that a “mark” is NOT the same as praise. Like food, praise is R+, and comes after the mark.
All training sessions should end on a positive note, and only you can make that happen! Happy training and train happy-or don’t train at all!
Char
When it’s time to get the food out of your hand-using food rewards for task focus and distance work
If you read my previous post on food and relationship then you already know that I believe in the power of food not only when teaching and working our dogs but also to build and enhance relationship. As a trainer my priority is always relationship. But having said that, many people fall into the trap of building dependency on food during the learning phases of training and then get stuck there. It then becomes, “oh no, my dog won’t perform without food!” If clean technique was used and task focus developed early on, this should not happen. And even when good technique is used there are still specific phases required in training for performance, and if any are skipped the ring performance will deteriorate. What I say here applies to toy rewards as well. And while we use extrinsic aids in the teaching process, I firmly believe that we must personally interact during the process as well. Since relationship is a huge topic all on it’s own-and there are trainers much better than I to cover that-the purpose of my article here is simply to address some ideas for effective mechanics in using extrinsic rewards in teaching our dogs. But keep in mind that all the while I am assuming a foundation based on relationship-and that relationship can and should include extrinsic rewards, but not be dependent on them.
So having established relationship as the priority and keeping that as a constant, we get back to the mechanics of using extrinsic rewards. My training focus in recent years has been more on distance work, independent thinking and problem solving by the dog. This can be very different from training a nice heeling performance or recall, where the rewards are often on your body; in a pocket, your hand, even your mouth (from whence you “spit” the treat out for the dog to catch in a perfect front position.) But even in training those behaviors, at some point you still must get the food off of your body in order for the dog to perform in competition.
So my intent here is to emphasize task focus, so that we can train our dogs to work from close up or at a distance without relying on the immediate presence of food or toys. A systematic approach and long term plan must be implemented if you expect to get a reliable performance from your dog in the absence of food in your hand or anywhere on your body. And please note that this step is to be used only after a behavior has become fluent by using a high rate of reinforcement followed by variable and random R+.
In my previous post on food and relationship I talked about hand feeding. Most of us know the “Zen game”- your dog maintains eye contact with you until you release them to take the food from your hand. That’s a great start! Over time then, for performance training we need to take this further with a systematic approach of raising criteria in two main areas: the level of difficulty of behavior required and the distance from the food. And finally, randomizing food R+ at a distance and close in. By doing this, your dog will learn to trust that the food reward will eventually come from anywhere at anytime if they just focus on the task. Using the Zen game as an example, the beginning behavior is: “make eye contact with me while food is immediately present”.
The key then, over time, is to ask for more difficult behaviors and chains of duration while increasing distance from the food. Once the behavior is completed, the dog is released to go and get the reward wherever it is. This can vary from a stationary food target, to tossing or feeding directly from your hand, or even multiple food targets and sending to a different target for each reward. Many of us do a version of this at trials, as in “let’s run to the crate for your jackpot” after the ring performance is complete. The difference in training this systematically by appropriately raising criteria over time, is that it builds confidence and focus on the task at hand. The food becomes tertiary to the game rather than the game being a means to the end (food). In order for this to be effective you will need to train a reliable “release to reward” on cue immediately following your mark or click. AND, you should praise and interact with your dog after the mark and before releasing to food. I use “get it” after the mark and point to the target wherever it is, and then I continue praise while my dog is being rewarded with food; bridging praise with food increases the value of praise alone as reinforcement. Obviously this is a very general description of the process, and the specifics will vary depending upon your chosen sport and the required behaviors. But just as food can be a bridge to a relationship that stands on its own, it should also be a bridge to building a reliable and independent performance. In any case, the food should be faded as you interact and connect during progressively longer behavior chains in the ring. How you personally interact and connect depends on many variables which you will have to determine based upon the sport, your dog, and you.
As always, the fundamentals and methodology are key to learning and teaching:
- Take baby steps and raise criteria appropriately
- Be accurate with your timing of marking and R+.
- Always accentuate the positive and ignore the mistakes.
- Praise after the mark to enhance the value of praise as reinforcement with the food.
- Don’t be a robot! Be sincere and engaging when training.
Train with a smile or don’t train at all. And above all do no harm.
Char
Training with food and relationship, (My grandmother’s chicken and dumplings!)
Yes, I will be guilty of anthropomorphizing here. But my grandmother’s chicken and dumplings were the inspiration many years ago when I had the light bulb moment and became a cross over trainer. As early as I can remember, my grandmother (we called her Nannie) always made chicken and dumplings when we came to visit. It became my most favorite dish on earth, and still is to this day. I am nearly 60 now, and Nannie is long gone from this earth, but to this day I never eat chicken and dumplings (I have her recipe!) without feeling the love and remembering my Nannie who made that special dish for me. It was only an extrinsic reward, but it served as one of many reinforcements to our loving relationship. As Nannie grew older and could no longer cook for me, food of course did not matter. I loved spending time with her no matter what because we had the history of reinforced love between us. Our relationship was all that mattered.
So we recently had a discussion in a trainers group about reinforcers, and of course there are many different perspectives here regarding food dependency in training and performance. In this particular discussion, the self -labeled “balanced trainers” said that clicker/+R training is solely about food and nothing more-a bribe as such. They also claimed that many dogs are not food motivated anyway. They did however see the value in training new behaviors with positive reinforcement-but they also believe that once the dog “knows the behavior”, corrections are in order when not performing properly on “command” . So, the “balanced trainers” use food and +R in the learning stages, but then justify corrections later on when they have decided that the dog should be error-free. Of course that idea is all based upon the notion that as trainer/handler they are always perfect, so therefore the dog should also be perfect, or “be corrected for being wrong”. To quote Susan Garrett, “When you correct a dog, you are actually punishing him for your poor training.” Yes!
The other side of the food dependency issue is that it can undermine relationship as the priority. I agree that food can certainly become problematic when used in training with no strategy or emphasis on relationship in the process. But when properly used, by prioritizing personal engagement along with strategic marking and food reinforcement, I maintain that food can and should be an enhancement to learning and relationship-not instead of. Food is ever present and necessary in our lives, and as humans it is not limited to being a primary reinforcer, (i.e. simply eat to live). It is also an event, a pleasure and enhancement to life, and a reinforcer of relationships as such (the chicken and dumplings). It is a huge part of our everyday socialization and family relationships. We shop for and prepare meals together, enjoy the time spent eating together, eat out for fun and entertainment. The association of anything with food, including work and a relationship, can be very powerful-conscious or not.
I also want to emphasize here that most dogs are at least potentially food motivated. As Bob Bailey has said, “any animal that is not food motivated is dead”. As one of the most powerful motivators for all living beings, it is rare not to be motivated by food when it is used properly in the training process. And I am not referring to “nothing in life is free”-as in having the dog work for every morsel of food they get. To quote Kathy Sdao, “plenty in life is free”! As it should be. My husband and I frequently share food with our dogs just for being cute, funny, or just because we are eating something that they want too (as long as it is not bad for them of course, and they know to wait politely). We also just talk to them and tell them how wonderful they are, while they sit on our laps. We play with toys and take them for walks. We comfort them during thunderstorms-and yes we use food then too! Our dogs don’t hear harsh words-they don’t need to. Some might say that our dogs are spoiled. But they are loved, happy and willing workers and enjoy our company. And food is only a part of that, as it is a part of everyday life. Personal interaction must be present to have a relationship. So while food may be ever present in +R training as such, that does not mean we cannot or do not have a personal relationship with our dogs. Not anymore than it does with the rest of our family with whom food is shared.
So here are just some things that I have learned (some quite incidentally) over the years with regard to incorporating food into my training and building the kind of relationship that I want with my dogs. Hopefully it carries over into performance as well. Nothing really new here, but maybe some good reminders for those who might read this:
- Build and maintain the positive association with the food and you into everyday life: I always hand feed my dogs and we often train for meals. But even if not training, I always talk to them while they are eating from my hand. We have a little ritual for meals and in doing so, over time, meals became an event and socialization between us, however brief. I originally started this many years ago to keep my dogs from bolting their food-and the added benefits have been enormous! This takes me less than 10 minutes a day (about 5 minutes, twice a day). And it is not necessary to hand feed, but at least make the mealtime a special event by personally engaging and teaching your dog to enjoy your presence along with their meal.
In training:
- Always pair food reinforcement with personal engagement. Prioritize your relationship, not the food. After the mark (or click), verbally praise and pet (if your dog enjoys touching, some do not) while delivering the food reinforcement. Associating your interaction with the food adds value to “you” as a reinforcer.
- For food to have value, your dog must have an appetite. Do not overfeed your dog, and consider the food used during training as a part of the total daily ration.
- Variation– In actual training I sometimes just use part of the regular meal, and sometimes other treats. Variation is key-even higher value switched to lower value often doesn’t seem to matter, just being different can add value from the dog’s perspective. New or difficult behaviors might benefit from an extra special treat-but some dogs can become fixated on the food if too high in value, so observe carefully and choose wisely.
- Delivery technique: Keep the food out of your hand until after the mark or click. This is hard to do sometimes, but I do keep it in mind always. This keeps your dog task focused rather than food focused. (The exception here would be if you are shaping novel behaviors-where the presence of food actually drives the offering of behavior for dogs who know shaping. That’s another topic.)
- Randomization of food rewards: Once your dog is consistent at performing a behavior on cue with a history of high rate of reinforcement (food and praise), begin to randomize the food. But still always praise and engage as part of reinforcement. The randomization of food with consistent personal engagement builds the bridge to performance in the ring, and most of all a positive relationship.
Treibball and the competition obedience dog…no herding required!
Do you need competition obedience background to play treibball? No. The necessary skills can be trained within the context of the sport itself. BUT the point is that if you are a competition obedience trainer and want to want to try a new sport that is very compatible with those skills in fun and creative new ways, try Treibball! I think you will appreciate and enjoy how much crossover training is involved.
My dogs and I have done lots of competition obedience and Rally over the years, and in training for Utility I came to appreciate what “distance work” really meant. After tons of reinforcement for heeling at my side and coming to front, I now had to teach my dogs that going away from me and taking direction at a 50 foot distance was also a good thing! Now for some dogs this comes easier than others. Some herding and hunting breeds are especially predisposed to working away from the handler. Not so for my 2 mixed breeds, which are a whippet mix and a “mystery” terrier mix. Training these two for distance work took some time, and it was worth it in the end. Over time they learned the go-outs with directed jumping and signals, etc. very well. And the key for me was to reinforce, reinforce, reinforce-“out there”- away from me-where they needed to be to do the exercises. For me, distance work is one of the most gratifying skills to train. It builds confidence and shows real teamwork to maintain that kind of attention and be able to work from afar.
Enter Treibball. Now we really needed to work on distance skills. Many who come into Treibball have herding and/or agility background. My dogs and I have neither (save for 2 agility titles, just enough to say that we “did it”). But having those utility go-outs made training the “send-away” to behind the balls a fairly easy transition. It took me a little while to figure out how I wanted to train what I refer to as the “clockface directionals”. But I have done that by envisioning a clockface and cueing accordingly for 3 O’clock or 9 O’clock, from the handler’s position at 6 O’clock. And retrieving…well, Treibball is basically a retrieving game, except the balls are pushed instead of being carried. Any of the obedience retrieve exercises- open dumbbell and utility directed glove retrieve- plus the directed jumping exercise, all provide a good foundation for teaching a selected ball drive. And the utility signal exercise and the open drop on recall provide skills needed for cueing positions behind the balls. And the recall to front- the perfect foundation for orientation to handler with the ball.
Are we fast and flashy on the treibball field? No, that’s just not our style. But we do get the job done and have fun doing it! It is most gratifying to utilize skills from another sport in a brand new and creative way. So while herding and agility can be a great fit with treibball, I have found that competition obedience by itself is also an excellent fit with treibball. I would also say the reverse-that Treibball would be an excellent foundation sport for any of the other sports mentioned, and fun for any dog and handler! Here is my 13 year old Abby, who is just learning Treibball after a lifetime of Rally and Obedience. Definitely not a herding dog! http://youtu.be/FuZnC_ytpP0
Dogs On The Ball 