Very true, I'd probably have ended it. But the concern is that with a small sample that one never knows if it will improve or continue to go downhill.

10% over 4 years is brilliant.