Back to the drawing board...

 My SAR team had evals about a month ago now (wow, am I behind...) and I went through it with Kenzi. Good news - she has made progress since last time!  Bad news - we didn't pass...

 I've thought about washing her from training, but she does continue to progress and she's doing really well in training scenarios. So I'm reevaluating how I'm training and tweaking it a bit.

 A couple things she had issues with were obedience related.  Like heeling off leash through a crowd. Kenzi loves people and is highly distracted by them. We did an okay on leash heel pattern. But I need to work harder on this. My plan is to go back a few steps and up the rate of reinforcement.

 She kind of flunked direct too.  For directional work, picture a baseball diamond with raised bases at 1st, 2nd, 3rd and pitchers.  These bases are 25 yards apart. The handler stands a home plate with their dog and sends it to different bases in a pattern given by the evaluator. For instance you could be instructed to send your dog to 3rd base, pitchers, 2nd base, pitchers, 3rd base then recall. The dog needs to stay on each base for at least 5 seconds and complete the given pattern in 3 minutes. Kenzi does great with the initial send out as well as going back and forth between pitchers 1st and 3rd. But she got highly confused when I asked her to go back to 2nd base from pitchers like she didn't understand the concept of looking back. So right now I'm just using pitchers and 2nd base, only placing them 20-30 feet apart and we're practicing short, highly rewarding sessions with a "go back".  This is working well and she seems to be getting the "go back" concept. I'll gradually increase the distance between bases and when she's consistently going back with the bases 25-30 yards apart then I'll add the 1st and 3rd bases back in.

I'm tweaking her reward system for searches too. I was having the victim reward all the time - her usual reward was playing/interacting with the victim and I rewarded only occassionally. But the handler has to reward during evaluations and I think this threw her off a bit too when all the rewards started coming from me rather than the victim. She was fine with the first few but then she start acting like "this is different - maybe I'm not doing something quite right"  Crazy, over-analyzing Border Collie.... So now when she goes in and alerts, the victim throws her a frisbee, then I follow up with playing a short frisbee game with her. I'll try this for a few weeks and see how it works.