Positive reinforcement as a training method is not *just* giving treats for good behavior. There is a precise methodology to it, developed by B.F. Skinner, refined by Marian and Keller Breland and generations of trainers since.
In order to get an animal working as hard and eagerly as possible for you, the rewards should be, surprisingly, small. Instantly digestible treats that the animal enjoys, but leave wanting more. They should come only when the desired behavior is performed; if the animal can get the treats in other ways, they won’t have as much incentive to do what you want them to.
Most importantly, the treats should not come regularly, but in what we call an “Intermittent Reinforcement Schedule” - in other words, in a random, unpredictable manner. If the treats come every time, the animal will feel free to slack off, confident that it can always get more treats later. If the treats aren’t constant, but are predicable, the animal will learn to not bother when treats are unlikely, and only try when they’re likely to come. So they must be moderately rare, and unpredicatable.
My iPhone is the best positive reinforcement trainer I’ve ever met.
Its treats are email messages, texts and Facebook updates. They’re all very small and instantly consumed (mentally). They only come if I check my phone. But they don’t come every time I check my phone.
What’s the result? My phone has trained to check in with it over and over throughout the day. I would actually hate for someone to actually do a count of how often I check it. And that’s with me knowing that I’m being trained, and how. When I look at many other people, reflexively checking their phones in the middle of conversations, conversations in which they are interested and engaged, I am in awe of the power of positive reinforcement as a training tool. Truly, we must use this power only for good. Unlike the gaming industry.