When you describe someone as a 4.75-5.25 pitcher you are talking his true talent level, that is very different than a small sample of results. When you are looking at ERA the stat itself it doesn't become significant until something like 500 IP. In fact over a 200 IP sample you would be better off knowing absolutely nothing except K/9 than knowing nothing but his ERA over that sample, that is just how meaningless the stat is over a small sample. Randy Wolf is most likely a 4.00-4.50 ERA true talent pitcher, doesnt' mean he can't have an ERA of 5 or an ERA of 3.50 this year. This isn't some crazy concept made to defend Suppan, it is just using stats that make a little sense over smaller samples.