Arguably the most exciting part of the current Royals’ roster is the bullpen. Aaron Crow, Jeremy Jeffress and Tim Collins represent the tip of The Process iceberg. One doesn’t have to squint very hard to see those three plus Joakim Soria locking down wins for a contending Kansas City team: maybe not in 2011, but not too far into the future.
Some of the luster surrounding the rookie hurlers has worn off after a string of rough outings over the past week, but we all know that relievers – rookie relievers especially – are never perfect. The question that crossed my mind after seeing Tim Collins implode the other night was how often should we expect a good reliever to, well, not be good?
Searching back over the past five seasons, I sorted pitchers by number of American League games in which they appeared in relief. Starting with those with 200 or more appearances in that time frame, I removed the closers. I do so because we hold closers to a different standard of perfection and they are used in what has by and large become a very controlled and similar situation in most of their appearances. After that, I sorted their overall performance by xFIP, as ERA for relievers is a pretty poor way to judge them.
The above process left seven non-closers with 200 or more appearances over the past five years and an xFIP under 3.90. Why 3.90? Well, Shawn Camp – THAT Shawn Camp – was the next pitcher to come up if I went any farther down the list. The appearance of that name followed in short order by Kyle Farnsworth screamed out ‘stop here!’.
From this point, very simply, I counted the number of appearances by each of the seven relievers in which they allowed a run. The results were as follows:
- Matt Thornton – 78 out of 342 appearances (23%)
- Scott Downs – 56 of 323 (17%)
- Rafael Perez – 65 of 266 (24%)
- Grant Balfour – 56 of 213 (26%)
- Darren Oliver – 78 of 294 (27%)
- Joaquin Benoit – 58 of 241 (24%)
- Jason Frasor – 70 of 290 (24%)
I don’t intend to get into a debate over whether we expect more out of Collins, Crow and Jeffress than the guys on the above list. Suffice it to say that these seven pitchers have been effective enough middle relievers and set-up men to pitch in a large number of games over a five year period.
During that time, these seven gave up a run somewhere between once every four or five outings. For the sake of boiling this into real life and not statistical decimal point dreamland, I think we could roughly say that a good non-closing reliever allows a run in two of every nine appearances. The Royals’ Tim Collins, by the way, has appeared in 10 games this season and allowed runs to score in two of them.
Without question, this is a pretty crude way to study the subject. There is a big difference between being asked to get one or two batters out and being asked to pitch three innings and the data above makes no adjustment for an appearance where Matt Thornton was asked to retire one hitter with a runner and did so versus an appearance when he pitched two and two-thirds innings and allowed a solo run with his team up four.
Defining who is a truly effective reliever is a much deeper study and the point of this quick analysis was simply to find out – in casual fan terms – how often one can expect even your best relievers to get dinged for a run. The expectation among all of us when a reliever enters the game is for that pitcher to be lights out. It is not realistic to expect that every time and we all know it, but we still expect it and agonize when it does not happen. When the Royals are up 3-2 with a runner on first and one out in the seventh inning, we really don’t care what Aaron Crow’s WHIP is when the inning is over, we only care that no runs scored.
Of course, there is the second part of the scenario: it’s not just runs charged to a reliever, but inherited runners he allows to score. That data, noticeable to all, is not included in the above study (I’m not sure study is the right word for the small amount of research, but here at Royals Authority HQ we like to think so).
Going back to our list, we find the following numbers on inherited runners and inherited runners that ended up scoring on our seven pitchers:
- Thornton – 93 of 296 (31%)
- Downs – 54 of 163 (33%)
- Perez – 43 of 174 (25%)
- Balfour – 42 of 154 (27%)
- Oliver – 57 of 171 (33%)
- Benoit – 24 of 105 (23%)
- Frasor – 48 of 156 (31%)
Using this group of relievers, it seems that allowing somewhere between one of every three and one of every four inherited runners to score is the norm. While the sample size is so small as to be irrelevant, Tim Collins has allowed three of six inherited runners to score – two of those coming in last night’s seventh inning. Jeremy Jeffress has inherited two runners and neither have crossed the plate while Aaron Crow has inherited SEVEN runners and has yet to allow one to score.
There is much to like about the Royals’ young bullpen this season. Ignoring the Crow should be a starter argument for now, I truly can see this group being a ferocious bridge between what we hope will be a powerful young rotation and a back-to-normal Joakim Soria for years to come. As good as they might be or become, however, the above shows that perfection simply does not happen.