Rafael Soriano Is Slicker Than Your Average; Another look at Rays Bullpen Splits

In most cases, a reliever is a specialist. Outside of the few elite relievers who can get batters out on both sides of the plate and in a variety of different ways, your average run of the mill relief pitcher is likely to be better at some things than others. Whether that is batter's handiness, or batted ball types, there is some kind of disparity. If not, then he wouldn't be a run of the mill relief pitcher.

Thanks to the new splits on Fangraphs.com, we can analyze players splits is ways we were unable to before (or at least not as easily). With that in mind, I decided to take a look at the six bullpen members that are most likely to make the opening day roster.

Looking at the top six candidates: Rafael Soriano, J.P. Howell, Dan Wheeler, Grant Balfour, Lance Cormier and Randy Choate, we know each pitcher has a different skillset. Choate is a LOOGY, Cormier is a groundballer and Soriano is a strikeout machine. Joe Maddon's job is to combine all these unique talents into one successful unit. Beyond the ground ball/fly ball and handiness differences, this group has  potential for platoon usage based on who is or isn't on base. Here are the career platoon splits, bases empty and runners on, for the six names above.

*As a reliever over the past two seasons, Cormier has posted a 3.9 with runners on and a 4.3 without.

**As a reliever over the past two seasons, Howell has posted a ~3.5 in both situations.

Please remember these are just numbers. They aren't absolutes and aren't magic tarot cards. Nonetheless, they don't lie and should give you a better indication on which split works best for each pitcher.

With the data for all six cases, Grant Balfour and J.P. Howell are best suited as inning starters or taking over mid-inning without a visitor waiting on-base. They are far from terrible with runners on, but the goal is to optimize their strengths especially in Balfour's case. Howell's numbers have been more balanced since his move to the bullpen. Rafael Soriano and Randy Choate would also fall into these categories, but I'll get to them later.

Moving along, we can see that Dan Wheeler and Lance Cormier (looking at the last two years) like Balfour and Howell aren't terrible either way, but are slightly better when there are already a few ducks on the pond. Randy Choate is clearly better with the bases clear, but his role is so defined that unfortunately he'll be brought in to face that left-handed batter regardless of the runner on situation. Soriano is also better coming in with the bases empty, but is pretty damn good either way.

If a starting pitcher is struggling and is facing imminent danger with runners already on, then depending on the handiness of the batter, and point of the game, it would seem that Cormier, Wheeler or Choate would be your best options. Wheeler is a ROOGY where Choate is a LOOGY; Cormier fits somewhere in between.

If your starter's day is done, and you are able to start off your bullpen's night with a fresh inning or without runners on, Balfour and Howell seem like the smart choices. Through various posts we concluded that far too often last season that a starting pitcher would begin an inning off by allowing a base runner and in turn Balfour was called in off the bench. According to the split data, this is not the best way to introduce him into the game.

As mentioned, these numbers look great on this site, but rarely does a game work out so perfectly that you can stick to a script. However, this should give us a better understanding of which reliever should be used when. Each game is its own beast and a variety of factors play into bullpen usage. Personally, I would never think to invite Wheeler into a game with men on base over Balfour or Howell, yet my eyes/perception seem to differ from the data.

In the case of the Rays, some relievers are better off starting their night with a moment of clarity. Other's are at their best facing 99 problems. In either case, it is best to take the most likely road to success before it's on to the next one.