That is, with a small enough sample, if it just so happened the nicest cubs had (some traits) in common, you can easily end up exaggerating those traits in later generations.
The Russian study is heavily inbred. They only ever had a maximum of 2000 foxes - now bred for a few dozen generations.