No lefties, no difference for Fish

With Andrew Miller officially to start the season in Triple-A, the Marlins will head into Opening Day without any left-handed starters in their rotation. The team still has two lefties in the pen in Dan Meyer and Renyel Pinto, but of course, those guys can’t pitch (ideally) until the sixth inning. Won’t the team be in trouble having to face such talented lefties as Ryan Howard, Chase Utley, Brian McCann, and Raul Ibanez? Let’s dive into some numbers to find out.

In 2009, the Marlins faced lefties in 44% of their PA that did not end in an intentional walk. Since 2007, that figure is 39.4%. I did a weighted average (5/4/3 for 2009 on down to 2007) of the three percentages (weighted also by PA) and got a weighed average of 40.2%. That’s a nice and even number to use for an estimate of 2010.

But we only want to deal with starters, and some of those PA will end up in relievers’ hands. So let’s make another assumption: PA versus lefties and righties are distributed the same between starters and relievers. In other words, both starters and relievers will face lefties a projected 40.2% of the time. How many PA can we expect from our starters? Doing the same method as shown above, I get a projected percentage of 62.5% of PA going to our starters, a raw total of 3929 PA.

OK, let’s make one final assumption (I’m stretching things for the purpose of making this a quick study). Let’s assume that lefty and righty starters will receive the same distribution of lefty PA (something I know is not true, but probably is not too far off). With that, what would be the difference between having an all-righty staff and a staff populated by only one lefty (the Marlins likely would not have played both Miller and Sean West). Well, first we need to presume the number of PA that those two different pitchers, one righty like Clay Hensley or Rick VandenHurk and one lefty like Miller or West, would have taken up.

Last season, the Marlins got 350 innings from Sanchez, Miller, West and a tandem of the team’s worst starters/long relievers. Assuming a 1.2/1 ratio (just a wild guess) of innings to 4th and 5th starters respectively, that puts us at 160 innings going to 5th starter types. At 160 innings, you would probably expect something like 720 PA (around 4.5 PA/IP) given the quality of pitcher. Last season, those pitchers gave up 4.55 PA/IP, for comparison. In those 720 PA, we’d expect those starters to face 40.2% lefties, or 289 lefties. What’s the difference between a righty and lefty facing a lefty batter 289 times? Based on the R/PA I established in the relief splits article, the difference for one PA for an average pitcher would be 0.0157 runs per PA. I can’t hazard a guess on what that would mean for a poorer pitcher, but perhaps something like 0.018 runs per PA? Whatever sounds reasonable to you.

That difference over 720 PA is 11.3 runs. With the bump for poorer quality, that’s 13 runs. That’s around a win, which is decently significant, but not earth-shattering, I suppose. And of course, that considers that we have that righty in the entire year. We all know that that fifth spot does not stay static the whole season, and we would probably expect to see a guy like Miller or West show up at some point due to injury or ineffectiveness. If that’s cut by even a quarter of the amount, then we’re looking at a paltry eight or nine run difference. For a comparison, that’s a little more than the difference between having Dan Uggla’s offense and Cody Ross’ offense last season. Which is to say, probably not that much at all.