Shifting

When people ask me what I do, the first, answer that I always give is that science is about separating the things that we know from the things that we think that we know, and then figuring out how to move things that are in the second category into the first category. (If they press harder than that, then I start talking about mRNA and their eyes roll back up into their heads.) This is how I've always thought about the Pirates' defensive shifts (and defensive shifts in general), since they started using them a couple of years ago. Why do the seven position players stand where they stand, other than because it's where they've been standing since the days of Harry and George Wright? Why would it be wrong for them to stand somewhere else?

In that vein, James Santelli has an awesome look at the shifts and Dan Fox, the man behind them, at Pirates Prospects this week. I'm going to stop you here and say that if you haven't read James's piece, you should go read it before you finish what I'm writing here, because it's just a commentary on the points discussed in his article.

I love this quote from Fox in particular: 

“You want [the players] to be thinking ‘This is our normal positioning. This is where I’m supposed to be.I’m not out of position. I am perfectly in position because that’s where the ball is going,"

It is generally well-accepted that balls put in play turn into hits somewhere around 30% of the time no matter what a pitcher does; this is the basis for defense-independent pitching research and the hypothesis that gives us stats like FIP and SIERA. There are two ways that you can approach this; one is to just accept that pitchers will eventually regress back to a BABIP of about .300 and move on, and the other is that  to try and find a way to prevent that regression from happening. The Pirates have obviously been interested in this for years; when Neal Huntington first started cobbling pitching staffs together and rebuliding busted pitching prospects, they focused heavily on groundballs, because groundball pitchers have more of a chance to out-perform their FIPS. 

The shifts are a product of that goal. What if the reason that all teams and pitchers tend to regress back to the same BABIP is because all of the teams have their fielders standing in approximately the same place? And if that's the case, then why do all of the fielders on all of the teams stand in approximately the same place? The players have always stood there and it's certainly not the worst possible defensive alignment to space your seven position players evenly out on the field, but the reality is that we know much, much more about batter tendencies and where balls land than we ever have in the past. Why shouldn't we use that data?

It's probably too soon to really make any definitive judgments on whether or not shifts can cause a real change to defensive efficiency/BABIP, but there are some promising signs beyond the Pirates' success this year. The Rays were one of the earliest shift teams and they're annually among baseball's biggest shifters. They're also always near the top of the league in defensive efficiency. You have to have the right players to execute the shift, you have to interpret the data correctly to employ the right shifts, and so on. Still, the Pirates are trying something different this year and seeing good results. There's certainly some regression coming for the pitching staff, but if the shifts remain as effective as they have been early in the season it might not be nearly as much as some expect.