What is xERA? xERA is short for "Expected ERA" and it is a statistic that is used to determine what a pitcher's ERA should be by removing components of the traditional ERA that are outside of a pitcher's control (fielding errors, inherited runners, bad luck, etc.) and focusing on ones that, more or less, are (hits, walks, homers and strikeouts). Today we'll compare the standard ERA of ABL pitchers to their xERA, see where the big differences are, and try to determine where some of that difference might come from.
Methodology:
You can skip this part if you're not interested in hearing how I crunched a bunch of numbers.
First, some cutoffs. The ABL season is already a small sample size, but there are some pitchers (two-way players, especially) who I feel did not throw enough innings to draw any conclusions based on their stat lines. I made the minimum 20 IP. Also, extremely high and low ERA's and xERA's were not considered. On the one hand, there are players who put up regular ERA's to the order of 0.35 or something ridiculously good like that. If you're already that low, you're good...we get it. Plus you would likely have a negative xERA which is impossible. Really high ERA's, on the other hand, I left off because I figured it would take a lot to make a terrible ERA look better (or even worse). Is a 6.70 ERA that much better than a 7.60? They're both pretty lousy so why bother.
I used this formula for calculation: ((.575 * H/9 ) + (.94 * HR/9 ) + (.28 * BB/9 ) - (.01 * K/9 ) - Normalizing Factor). The "normalizing factor" I used was 2.66. This factor will vary league to league; when calculating MLB stats a lot of people use 2.77. I started there and then adjusted it (thank God for Excel and the Goal Seek feature) so that the "average" pitcher's ERA and xERA would be identical. 2.66 was the number that got me there.
Results:
Here's a graphic of some of the pitchers whose xERA and ERA showed a significant difference.
What we can see here is that Rath and Dew fared significantly better than they "should have" in the ERA category, considering the amount of runners they allowed to reach base. On the other hand, McCarthy, Sumner, Hulse and Kendall might have pitched better than their ERA's suggest.
Analysis:
Kevin Rath had an interesting season. He managed to post a 2.03 ERA while leading the Oilers' staff in hits allowed (45). His expected ERA was much higher than his actual ERA, but he pulled this off by being a damage control expert: with runners on, he held batters to a stingy .229 average, compared to .347 with the bases empty.
Batters did the reverse to Oilers' teammate Cody Kendall. He posted a respectable 3.52 ERA, vs. a stellar 2.28 expected. Why? With the bases empty, Kendall cruised, holding batters to a mere .175 average. But once they got on, he suddenly became hittable, to the tune of .288. Comparing those two statistics, it seems as though Kendall may have problems keeping his composure with runners on.
For Dew, McCarthy, Sumner or Hulse, there aren't any clear hypotheses that come to mind. I don't have enough information to feel comfortable with attributing their discrepancies to anything other than luck. If you've got an idea, by all means, let's hear it, but I don't want to go out on any flimsy limbs. It's interesting, nonetheless.
Kevin Rath had an interesting season. He managed to post a 2.03 ERA while leading the Oilers' staff in hits allowed (45). His expected ERA was much higher than his actual ERA, but he pulled this off by being a damage control expert: with runners on, he held batters to a stingy .229 average, compared to .347 with the bases empty.
Batters did the reverse to Oilers' teammate Cody Kendall. He posted a respectable 3.52 ERA, vs. a stellar 2.28 expected. Why? With the bases empty, Kendall cruised, holding batters to a mere .175 average. But once they got on, he suddenly became hittable, to the tune of .288. Comparing those two statistics, it seems as though Kendall may have problems keeping his composure with runners on.
For Dew, McCarthy, Sumner or Hulse, there aren't any clear hypotheses that come to mind. I don't have enough information to feel comfortable with attributing their discrepancies to anything other than luck. If you've got an idea, by all means, let's hear it, but I don't want to go out on any flimsy limbs. It's interesting, nonetheless.
Good analysis. However, I would suggest looking at FIP and xFIP if those numbers are available, as hits are taken out of the equation in both (hits allowed still relies heavily on one's defense), and with normalized homerun/flyball rates in xFIP. Are these stats available? Or would you have to compile them all yourself. Because that would suck.
ReplyDeleteWhen I did the RC and xERA i had to compile them myself. If it's something that can be determined using basic everyday stats, it's not a big deal because once I punch the formula into Excel all I have to do is transcribe the statistics, and adjust the league factor to fit the ABL.
ReplyDeleteFIP doesn't look like a bad one to calculate at all. Runs Created was more complicated, I think. The only problem is that the IBB stat doesn't look like it's available for all the clubs. Some don't have HBP available either. I could just rig it to only count BB but I don't know how that would affect accuracy.
What I'm more interested in is a defensive statistic. Someone recommended UZR, and I was stoked until I looked up what you need to do to calculate it. Now I'm thinking probably not. I would like to do another post like I did for RC and xERA but for a defensive stat, if I can find one that's manageable, except this time have readers suggest which players they want me to compare. So any suggestions for a stat I could handle, or a player to analyze, let me know.