One interesting thing from the Ellis podcast is that he said he's now using food much more like he would use a toy. So he's no longer teaching the dog to sit quietly and take it gently, but teaching the dog to push for it and holding the food in a way that he won't get bitten. Basically building food drive into a reinforcer that works the same as a toy (energizing and drive building, vs quiet thoughtful work that food was traditionally used for). I've used some of his techniques and really liked them with Watson because I can get some of the key benefits of toys without having to use toys.One thing I did that I think helped with Soro is with toys I allow zero inhibition, except basic commands do apply in case I need to stop the game. Basically, whereas for food there is a very clearly established list of rules (no jumping up for it, have to wait, not allowed to just take it, be gentle, etc...) if he wants the toy in my hand and there is no other command given, he is allowed to do whatever he can to get it. He can jump on me, crash into my body mid jump, snatch it from my hand as I'm walking past, etc.