The decoy method seems brilliant in design, but they aren't executing it properly. If she would have been warmer in the building, she should have gone there.

The exercise seems to be a wasted effort. It would be interesting to do a statistical analysis on past decoy counts.