The Myth Of A Low ERA
Measuring a pitcher's performance through wins and losses is generally not considered a true and accurate measure. Similarly the use of a simplistic measure like ERA is affected by ballpark factors, the quality of the defense as well as quite a bit of luck. The xFIP metric is an estimation of true ERA based upon the variables (hits, walks, strikeouts, HBP, fly balls allowed) that a pitcher is responsible for to give an indication of how well a pitcher performed independent of how well the rest of the side performed. As an added benefit, the xFIP estimation has an advantage in that it is a very good predictor of future ERA for a pitcher.
Looking at the top 10 qualified pitchers in terms of ERA during the first half of the season, it really is an All Star cast with four of the ten being called up for the Midsummer Classic. However, each of the top ten pitchers have a xFIP over a run higher than their ERA and averaging out at a +1.60 difference. For the four All Stars, it was even worse with an average difference of +1.68. With xFIP being a good indicator of future ERA, it might be prudent to sell any stock you have in Tim Hudson and Clay Buchholz.
Moving over to the worst 10 qualified pitchers in terms of ERA, you certainly won't find a 2010 All Star among them. However, their pitching performances in the first half of the season records are nowhere near as bad as their ERA would suggest as they average an ERA 1.16 runs above their xFIP. The news is especially good for the Minnesota Twins who have two pitchers in the list and who need all the help they can get as they currently languish in third place in the AL Central despite being heavy pre-season favorites.
Interestingly, the first half average ERA for the top and bottom ten pitchers are 2.70 and 5.33 respectively but the average xFIP is 4.31 for the top pitchers and 4.17 for the bottom pitchers suggesting that in terms of making the All Star Game, it really is better to be lucky than good.
.
Reply With Quote
Originally Posted by fightingirish
