Good point, Nod, I think it's advisable to differentiate between losses (4 in a row is not a big deal and should be expected for many good systems with a 30-40% win rate) and non-forced errors (i.e. you knew you shouldn't be doing X, but you do it anyway).