Safe or Stranded On Base: Part II
In Part 1 of my look into LOB%, the general conclusion was for every .005 of Batting Average Against added to the base of .265, the expected LOB% of a pitcher should go down .5%. To make this even easier, I've created a simple chart showing the rule:
Today we're going to look at LOB% from more of an individual standpoint. I'm going to take a look at pitchers' LOB percentages from 2005 to 2007 and more closely examine those pitchers whose LOB% either remained surprisingly consistent or varied greatly.
Starting with the most inconsistent LOB pitchers for the last three years, their names are: Bronson Arroyo, Jarrod Washburn, Jon Garland, Josh Fogg, and Tim Hudson. Some boring pitchers, some good pitchers, some despairingly average pitchers, this group has little consistency at all. I quickly checked these pitcher's ERA's and BAA's and found that they correlated almost exactly with the LOB%. How Boring! LOB% almost single-handedly explains why these pitchers had bad years.
Perusing through my data again, I found a prime example of someone to delve deeper into. Livan Hernandez would be his name, and his 2006 and 2007 seasons looked something like this:
2006: 4.83 ERA, 70.4 LOB%, .288 BAA
2007: 4.93 ERA, 76.0 LOB%, .308 BAA
Livan pitched overall poorly in 2006 and that led to a 4.83 ERA, although a deserved 4.83 ERA. In 2007 Livan pitched even worse; his WHIP, BAA, BB/9, HR/9, and LD% all went up, yet somehow he managed to keep his ERA under 5.00. Looking exactly 1 inch above where you are now, you'll find part of the answer, his LOB%. Despite having a higher BAA in 2007 than in 2006, Livan's LOB% skyrocketed up to 76.0%. Well how did that happen? The other part of the answer is not located somewhere above on the page, but rather in his performance with runners on base, or his "clutchiness". The first method to measure a pitcher's clutch performance is to check out his BAA with no runners on vs. either runners on or runners in scoring position and the second method is simply using Fangraph's Clutch score. In Livan's case, both methods would have worked. In 2007 Hernandez led all pitchers in the majors with a clutch score of 1.53 and a simple check to Yahoo's situational stats found on its player pages shows this:
Bases Empty: .333 BAA
Runners On: .274 BAA
RISP: .264 BAA
When a pitcher's LOB% and ERA do not match up, usually clutch performance is the reason behind the anomaly.
After doing a regression analysis, ERA and LOB% do correlate well with an R-Squared of .573, but as I pointed out with the Livan Hernandez example, anomalies exist and can be explained. Even though LOB% is a stat used to predict luck in a pitcher's ERA, LOB% itself can also be lucky. That's where Batting Average Against and Clutch Scores come in, to determine how lucky/unlucky a LOB% is. Of course, BAA and Clutch Scores have a certain amount of luck in them too, as does every stat and that is precisely why statistical research is limited in its reach. However it remains the best option available since stats are easily obtained and taught and video and mechanics are not so easily obtained and taught.
Obviously this article did not go exactly in the direction I planned it to go and I think that's more an example of the purity in my research - that I do not know what the conclusions will be when I begin and learn as you do. Just to give you the information you have not gotten, (although it really does not pertain to the overall conclusion) the pitchers with the most consistent LOB% over the last three years are Aaron Harang, Brandon Webb, Johan Santana, Tom Glavine, and John Smoltz. A pretty darn good bunch of pitchers, this group of five would make quite the rotation. Again, the most consistent and inconsistent LOB% pitchers turned out to be quite uninteresting and thus I changed the focus of the article to something hopefully more interesting and fruitful.
| BAA | .240 | .245 | .250 | .255 | .260 | .265 | .270 | .275 | .280 |
| xLOB% | 74.5% | 74% | 73.5% | 73% | 72.5% | 72% | 71.5% | 71% | 70.5% |
Today we're going to look at LOB% from more of an individual standpoint. I'm going to take a look at pitchers' LOB percentages from 2005 to 2007 and more closely examine those pitchers whose LOB% either remained surprisingly consistent or varied greatly.
Starting with the most inconsistent LOB pitchers for the last three years, their names are: Bronson Arroyo, Jarrod Washburn, Jon Garland, Josh Fogg, and Tim Hudson. Some boring pitchers, some good pitchers, some despairingly average pitchers, this group has little consistency at all. I quickly checked these pitcher's ERA's and BAA's and found that they correlated almost exactly with the LOB%. How Boring! LOB% almost single-handedly explains why these pitchers had bad years.
Perusing through my data again, I found a prime example of someone to delve deeper into. Livan Hernandez would be his name, and his 2006 and 2007 seasons looked something like this:
2006: 4.83 ERA, 70.4 LOB%, .288 BAA
2007: 4.93 ERA, 76.0 LOB%, .308 BAA
Livan pitched overall poorly in 2006 and that led to a 4.83 ERA, although a deserved 4.83 ERA. In 2007 Livan pitched even worse; his WHIP, BAA, BB/9, HR/9, and LD% all went up, yet somehow he managed to keep his ERA under 5.00. Looking exactly 1 inch above where you are now, you'll find part of the answer, his LOB%. Despite having a higher BAA in 2007 than in 2006, Livan's LOB% skyrocketed up to 76.0%. Well how did that happen? The other part of the answer is not located somewhere above on the page, but rather in his performance with runners on base, or his "clutchiness". The first method to measure a pitcher's clutch performance is to check out his BAA with no runners on vs. either runners on or runners in scoring position and the second method is simply using Fangraph's Clutch score. In Livan's case, both methods would have worked. In 2007 Hernandez led all pitchers in the majors with a clutch score of 1.53 and a simple check to Yahoo's situational stats found on its player pages shows this:
Bases Empty: .333 BAA
Runners On: .274 BAA
RISP: .264 BAA
When a pitcher's LOB% and ERA do not match up, usually clutch performance is the reason behind the anomaly.
After doing a regression analysis, ERA and LOB% do correlate well with an R-Squared of .573, but as I pointed out with the Livan Hernandez example, anomalies exist and can be explained. Even though LOB% is a stat used to predict luck in a pitcher's ERA, LOB% itself can also be lucky. That's where Batting Average Against and Clutch Scores come in, to determine how lucky/unlucky a LOB% is. Of course, BAA and Clutch Scores have a certain amount of luck in them too, as does every stat and that is precisely why statistical research is limited in its reach. However it remains the best option available since stats are easily obtained and taught and video and mechanics are not so easily obtained and taught.
Obviously this article did not go exactly in the direction I planned it to go and I think that's more an example of the purity in my research - that I do not know what the conclusions will be when I begin and learn as you do. Just to give you the information you have not gotten, (although it really does not pertain to the overall conclusion) the pitchers with the most consistent LOB% over the last three years are Aaron Harang, Brandon Webb, Johan Santana, Tom Glavine, and John Smoltz. A pretty darn good bunch of pitchers, this group of five would make quite the rotation. Again, the most consistent and inconsistent LOB% pitchers turned out to be quite uninteresting and thus I changed the focus of the article to something hopefully more interesting and fruitful.
