Do the Houston Astros Have a Closer for the 2015 Season?


This offseason the Houston Astros brought in three arms which, on paper, look to improve the state of the bullpen. Luke Gregerson and Pat Neshek are both well-established commodities who enjoyed great success in 2014. Will Harris, a waiver claim, was up-and-down in parts of 3 seasons with the Diamondbacks.

Job of a Closer

Earlier this morning I stumbled upon an article that takes a stab at assessing a closers’ true ability. Alan Carpenter, a writer at Tomahawk Take, tries to create Closer Rating. The results yielded that Aroldis Chapman, Craig Kimbrel, and Eric Gagne had some of the best seasons since 1990 as a closer. On the other side of the spectrum, the season in which Chris Reitsma gave up 7 homes, had a strikeout-to-walk ratio of 13:8 albeit in just 28 innings ranked as the worst in his study.

Essentially his theory is to take a combination of WHIP and FIP data to create a weighted WHIP to add to a factor of blown save percentage.

"WWHIP (Weighted WHIP) = ((10 * HR) + (5 * H) + (3 * (BB+HBP)) – (2 * K)) / IPClR (Closer Rating) = WWHIP + (10 * BSv%)"

There is a clear bias here when dealing with guys who have little data in regards to saves/blown saves. I agree with this method. Closers are supposed to come into the game and slam the door. That is done by getting outs. Sometimes a guy can get by with allowing contact; that can get risky. Carpenter places a high penalty on the blown save – the closer failed to do their job!

Reviewing the formula there is a little risk here in double counting data. Every home run is a hit, but not every hit is a home run. But Carpenter has failed to include that in his calculation. There is also a point that could be argued. Should the closer be blamed for intentional walks? Though it seems counter-productive, some pitchers do have IBB on their record. Probably from non-save situations, of course.

When the formulas are applied to data from the 2014 season with six arms who are all but guaranteed a roster spot, the results seem to indicate Houston is not equipped with a closer. I do want to emphasize that, as Carpenter said in his article, these calculations are not polished as to what would be on fangraphs. These are rough sketches and may not paint the true picture. But that is a common risk associated with interpretation of statistics.

More from Astros News

2015 Contestants

Yes, Jeff Luhnow managed to upgrade the bullpen on paper. But Jason Burke, Climbing Tal’s Hill editor, has experience with watching Gregerson. Along with suggesting signing free agents Francisco Rodriguez or Rafael Soriana, Burke writes:

"“With Oakland last season, Gregerson went 5-5 with an astounding 2.11 ERA. Case closed, right? Well, not exactly. After Jim Johnson struggled closing out games at the outset of 2014, the A’s tried Gregerson in the role, and found he wasn’t much better, saving just three of eleven opportunities in 2014. For me, Gregerson is not the guy to close out games for the Houston Astros, even if he has the closer’s incentives laid out in his contract.”"

[table id=15 /]

Yikes. Carpenter’s formula yields horrendous closer ratings. An elite season in this aspect would be a value under 2.00 according to Carpenter. Billy Wagner‘s 2000 season is included with the data in the article linked to Tomahawk Take. A little relief is provided when realizing that only Will Harris posted a worse closer rating than Wagner’s 2000 season. To be fair Harris only had 1 save opportunity in 2014. Personally I believe that Harris is slated for pitching in the 6th/7th innings. Harris has a nice selection of pitchers and I am excited to see what he accomplishes working with Brent Strom this year.

Retaining leads…and comparing results

In an attempt to fight the bias due to a small sample, I took a look at holds. This started to give a clearer picture as to why Luhnow feels confident in these upgrades. Gregerson, Neshek, and Sipp all had a lot of experience setting up a save situation. They combined for 58 holds last season.

While people are trying to create new statistics, and perhaps this has been created, I tried to compile a retained lead percentage = (Saves + Holds) / (Saves + Holds + Blown Saves).

My concern with using league statistics is double-counting data. A save, blown save, and a hold are three distinct events. It is possible that there are multiple holds or blown saves in a game but it is impossible for there to be multiple saves in a game. With these situations outlined, I felt that it was okay to sum the number of saves and holds and then divide that number by the sum of saves, holds, and blown saves. That resulted in what I am calling a retained lead percentage of 86.61%.

More from Climbing Tal's Hill

With Fields and Harris at the bottom utilizing that statistic at 75%, my skepticism of the 2015 Astros increased. Overall these six players have a retained lead percentage of 81.06% which is below that league average I found a moment ago. For comparisons sake: 2014 Royals 92.5%, 2003 Astros 90.9%, 2003 Dodgers 94.5% and the 2008 Yankees 93.0%.

These data all seem to indicate that the Astros are still not prepared to be one of the better teams in the MLB. Of course they have improved on the whole. Overall, however, the bullpen is still going to be a source of heartbreak. I hope that I am wrong and that Gregerson, Neshek, or Fields are able to break through as a closing pitcher.