Again, I'm posting out of order, but here's a video I took at our previous session before this morning's. I will post the video that helped us diagnose RG later.
I suppose I'd be looking to take heat out of Shadow's need to guard by very slowly building tolerance. So if the RG is only when other dogs approach you, and he's fine with you approaching other dogs - this might be something like Shadow on a boundary next to you; Katie (or other dog) on a boundary next to their human a way away; gradually over time, working on Katie and owner stepping towards you, then returning to the boundary. Monitoring Shadow, and slowing down if he's uncomfortable. Mixing it up with distance, duration and difficulty in terms of breaking the sight line (e.g. a fence), having a different physical orientation to the dog, being higher up, lower down, noise distractions etc. A challenge would be how subtle his indications are, or if he has a very short fuse, but knowing him as well as you do would help with these things. Alongside that would be confidence building games, improving working independently, thinking in arousal - all stuff you do already.
Now I'm ready to post our tactical approach, here are my thoughts on this. Shadow is a dog who learns what is expected of him very quickly, will repeat patterns and will become absorbed in work without it actually helping his emotional state. He will use work as avoidance, switching off from the world around him. So the issue we have with slowly building up the complexity, decreasing distances etc is that all that happens is he learns how to avoid the situation at X distance, and then applies the same at Y. It doesn't give him any coping mechanisms, and doesn't help him therapeutically. He won't look uncomfortable, because he's
quite comfortable, in his own little world. Think of it the same as a child who turns up their music on their headphones so they can't hear their parents arguing. The music masks the sound, but doesn't help the child resolve how he feels about it, and if he were to hear an argument without access to the headphones, he'd still feel the same as before.
Instead, my approach has been rather more confrontational, pushing him into situations where he has to deal with his feelings, but giving him a way to do so. So I've been teaching him that when he sees the thing (Luna interacting with another dog), he turns to me. As soon as he does this, he gets the throw of a ball away from the other dogs.
This accomplishes several things:
1. The chase of a ball is his biggest reinforcer, so it reinforces the operant behaviour of turning towards me when he feels uncomfortable.
2. Because he has such strong feelings about his ball, and the interaction will become a predictor of the throw, that will classically condition him to have a positive CER to Luna being with another dog.
3. It sets us up for the next repetition of the approach/behaviour so we can get in more reps.
4. He feels good about moving away from me (something else he RGs) in the presence of another dog.
5. It relieves the pressure of the proximity of the other dog.
6. Very often, the throw of the ball also attracts Luna from the other dog, so he gets the function of his original behaviour fulfilled - if she had a perfect recall and I had enough brainpower to manage everything, I would also recall her every time - but that's setting myself up for failure, so I didn't add that in
We started this without Luna, with Shadow in a field without muzzle or long line, and Katie (or one of the other dogs) the other side of a fence. We did parallel walks, head-on approaches and various other scenarios teaching him that coming to heel in that situation got him the ball. We also used me as the resource, with me interacting through the fence with the other dog. It was OK, but he wasn't uncomfortable enough! "What?!", I hear you cry... "don't you want to keep him well inside his comfort zone?". Well, yes and no. It depends on the dog. Shadow doesn't learn anything inside his comfort zone. He
wants to ignore other dogs, so is more than happy to do so. We also need to look at what my ultimate goal is here. I don't care one jot if he doesn't like other dogs. I have no need for him to become pally with them. If I thought that would help, I would continue on down that road, but after the sessions we had with Katie, I'm unconvinced that is the right course of action for him. Instead, I need to look at our environment and what I think is achievable. And that is, when we are walking out and about, if he sees another dog, he comes back to me. Whether or not the other dogs are interacting with it. That's all, no more no less. When he sees another dog, he is going to go into a state of heightened arousal, that's just a given. So I need to be able to get him to think when he's in that state, and to pull out an automatic behaviour that is associated with his emotional state. Feel icky = return to me. And that means, working on and around the threshold where he would react, so that he learns the replacement behaviour in that same emotional state. Yes, hopefully the work we're doing will help to raise that threshold, too, but that's not my main focus. My main focus is giving him an alternate behaviour to perform to give him a positive outcome.
So we moved on to a slightly different approach. Our next session (the one in the video), we went out into a big stubble field, with Squidge loose, Katie on lead, and Shadow on a long line, unmuzzled. And I worked on at first calling him to heel when he looked tense, but later just waiting for him to make that choice, and then throwing the ball. He made lots and lots of good choices. He did react a couple of times (trying to run towards Katie, vocalising etc), but that was simply useful information, we moved away and set him up to make a better choice next time. He made far more good choices than bad ones. Bravo, Shads!
We also set up some scenarios with the other dogs (with and without Luna being involved) where they would "appear" into the field without warning from the road below. That also worked pretty well. And, to finish, we had baby GSD Tazzy doing some direct head-on walking, which was difficult for him, but he took on the lessons from earlier and made good decisions.
I'm happy with this approach for now, and have more plans on how to progress it.
She's not my trainer, and she's not advising me - I'm running the sessions, she's just helping me 
