Holland is a pretty hilarious example of the effects of HR rate variance. In 2014, he allowed NO home runs in 37 IP and had a 1.46 ERA. In 2015, his HR rate was 17.5% and his ERA was a ghastly 4.91. Add those samples together though, and the HR rate was a very typical ~10%, and his ERA snuck in under 4, while his K and BB rates stayed fairly consistent.