So, we have splits now. We’ve talked about how to use them and how not to use them. Today, I want to talk about why our splits matter. After all, Baseball-Reference has a great splits section, and has for a while. But, B-R is focused on providing factual data (which they do very well – this is no knock on the site that Sean has built), and as we’ve talked about before, sometimes what actually happened is not the best predictor of what is going to happen.
Let’s use an example of why having split data for metrics like xFIP is important. If you go to Jered Weaver‘s page at Baseball Reference, you’ll notice his career L/R split is pretty large. Righties have hit just .232/.283/.365 against him, but lefties have managed a significantly better .267/.327/.449 mark. There’s a 126 point gap between his OPS against RHB and LHB.
Now, look at Weaver’s splits page here.
Vs Left: 7.17 K/9, 3.01 BB/9, 36% GB%, 4.51 xFIP
Vs Right: 7.47 K/9, 2.30 BB/9, 28% GB%, 4.42 xFIP
That’s a pretty different story. In terms of the numbers that measure skills, Weaver’s not actually all that much better against RHBs than LHBs. The things that drive Weaver’s large career platoon split? HR/FB and BABIP, two of the least predictive metrics out there.
For his career, lefties have turned 10.6% of their flyballs off of Weaver into home runs. Right-handers have just managed 5.6%, causing a big gap in home run rate. Also, lefties have tagged him for a .307 BABIP, compared to just .282 for righties.
Maybe more research into these issues will reveal that handedness effects these two metrics more than we currently know, but right now, Weaver’s profile is not that of a guy who will continue to post big splits going forward. He’s no Vicente Padilla. He’s a little bit better against right-handed batters, but not much more than the norm.
You’d have a really hard time knowing that from his results. This is one of the great things that we’ll be able to flush out with split data here on the site, and one of the reasons we’re so excited to have them now.