There are a lot of different techniques to solve problems in dog training. My intention in these posts, labeled "solution", is to present a solution that I found to solve a specific issue in Tess's training.
There are a ton of resources to help you solve your training issue - hopefully these posts generate more ideas to solve training issues for you and your dog.
Problem: Pulling on the leash
Solution: Discontinue "heel" ( defined as asking for a specific position by my leg on a loose leash) - instead, use "let's go", meaning - "whatever my leash length is, don't pull". If pulling occurs - immediate 180 degree turns.
AareneX has commented both here and on my other blog about her method of teaching a dog to walk nicely by her side, and if you are having a similar issue, I highly recommend you go back and read her comments on this post here.
What ultimately worked for Tess was a variation of what I read years ago in the "Monks of New Skete" books (linked in the resource page). To teach leash work they recommended working within a virtual box (ie - there is no box - but the pattern you are walking is a box shape!!!), making sharp turns to both keep the dog interested in your path, to "test" the dog's attention, AND to correct a dog. It's more of an "oopsy - I turned and you didn't! You must have not been paying attention....so sad."
I dropped the formal heel entirely - heel means stay at a certain position - something that in my current training I refer to as the Reinforcement zone (RZ). Teaching the RZ AND for Tess not to pull was too much. So I broke it down. Instead, during our informal walks and runs I used "Let's go" - which meant she could go anywhere she wanted at whatever speed - as long as she didn't pull. This was MUCH more fun for her - she felt like she had some control of the situation and choice - while still needing to respond to pressure and my cues.
Occasional bumps at the end of the leash were tolerated - after all, because I changed the leash lengths depending on the environment, she needed to know the breadth of her freedom. BUT - any sharp tug (darting forward for example, after a distraction) or pull that wasn't immediately rectified was IMMEDIATELY met with a 180 degree turn in the other direction. I continued at whatever speed we were moving at - whether that was a walk, run, or sprint.
The first 2 days I looked like a COMPLETE and utter idiot. Good motivation to get on campus EARLY and get my run over before classes started. I didn't travel more than 1 linear mile in 20 minutes because of my running back and forth. But now, 10 days later, I can get through an entire 30 minute run without doing a 180 correction except perhaps 2 or 3 times in very high level distraction areas. Tess is having FUN on our runs, my frustration level stays very low - and the most important thing - based on VAST amount of improvement in a short amount of time, I think Tess actually understands the concept, because I've made the criteria very black and white, and it offers her more of a choice in the activity.
I also like the concept that there isn't anything intrinsically "negative" with this approach. 90% of the pulling is because she gets fixated on something and darts towards it. By turning around, we are moving away from the thing she is interested in. It's only once she can walk/run towards and past it without pulling that she gets to go towards it - which she wants, and thus is a reward for being good!
I am told this exercise (turns, pivots) will come in handy once I start doing more with my Reinforcement Zone exercises, which will morph into my heel.
I'm beginning to think that if you don't look like an idiot while training your dog, you aren't doing it right.