St. Louis Cardinals: Opening Day Redbird Rants Mailbag

Evaluating the St. Louis Cardinals based on contact quality and plate discipline.

Maq (elmaquino): Does asking for a straight up dump of x stats for every St. Louis Cardinals starting player and pitcher count as a question?

Yes, it does. I’ll include the 2016 expected versus actual stats. Additionally, I’ll show my 2017 projections for the St. Louis Cardinals with a comparison to ZiPS and Steamer.

So here we go. First up, a little bit of an explanation. Feel free to skip down to the tables if you don’t care about the methodology.

I started working with my xSTATS models a little while back. Today, I have a model for starting pitchers, relievers, and position players for each of the last two seasons, and a projection for the 2017 season. Each model focuses on batted ball data sets to judge contact quality and plate discipline statistics. The goal is to isolate player performance independent of defense, both for pitchers and hitters.

I’ve found that exit velocity and launch angle for hitters are highly correlated year-to-year, which suggests they are the consequence of a repeatable skill. Exit velocity and launch angle allowed is less correlated for pitchers, but still has a strong enough relationship year-to-year. We also know using Pull%, Cent%, and Oppo% that hitters have some control over where they hit the ball horizontally.

Thus, I use exit velocity, launch angle, and spray angle to determine an “expected” number of hits for batters, and an “expected” number of hits allowed for pitchers. To get expected walks and strikeouts, I ran regressions to determine significant plate discipline variables. Sacrifices, HBPs, and IBBs are assigned at a league average rate. From there, I work toward expected numbers for wOBA for hitters and ERA and FIP for pitchers.

Below are the St. Louis Cardinals pitchers. I’ve included their actual 2016 statistics as well for comparison to my xSTATS.

Overall, this is a mixed bag. One noticeable standout is Michael Wacha, who my model shows was really unlucky last year on batted balls. While his pFIP was lower than his actual FIP, I had his xERA nearly a full run lower than his actual ERA.

Additionally, I had Lynn’s expected results from 2015 as significantly worse than his actuals. Being that he’s returning from Tommy John, I honestly don’t know if either his expected stats or actual stats apply to next year, but I figured it’d be interesting to include them nonetheless.

With regards to relievers, there’s not a whole lot that sticks out. Seung-Hwan Oh was fantastic at suppressing contact quality, but how he fares after a full offseason of scouting will be the key for his 2017.

Next up, the position players who I think will be the main contributors this year.

First, the bad news. Dexter Fowler and Aledmys Diaz probably overachieved last season, most notably in xwOBA vs. wOBA. Fowler’s expected line basically matches his career averages, though, which isn’t a bad sign. Diaz, on the other hand, has some issues in his batted ball profile that I’ll get to in another question.

The model had higher expected stats than actuals for Jhonny Peralta and Randal Grichuk. I’ve talked previously about Peralta’s 2016 season and why the narrative that he was impacted by the thumb injury while on the field needs to stop – he didn’t hit the ball poorly last year, he hit into bad luck. Based on that, there’s reason to expect both players to bounce back.

Next up, here are my projections for the St. Louis Cardinals in 2017. I’ll compare them to the average of ZiPS and Steamer projections. I’m not projecting Lance Lynn, because I don’t think his data and statistics from 2015 will reliably predict how he performs post-injury. I have some concerns that my projection model favors hitters, but not terribly so.

For the starting pitchers, I’m a little less optimistic about three of the four, excluding Lance Lynn. While I have the staff managing contact better than ZiPS and Steamer project, I’m predicting worse strikeout and walk rates.

For relievers, I’m right in line with the other projection systems for Brett Cecil, Kevin Siegrist, and Trevor Rosenthal. For Seung-Hwan Oh, my model is not regressing his contact management ability back toward the mean, so I’m quite a bit higher on him than ZiPS and Steamer.

Now, for the position players.

The story here is similar to that of 2016. I expect Jhonny Peralta to rebound in a big way from last year, and the model projects he’ll be the St. Louis Cardinals second best hitter. Additionally, the xSTATS model really buys into Matt Carpenter. I have him at 29 home runs, and he will have the chance to rack up 100 RBI hitting behind Fowler and Diaz. If he hits like I’m projecting, he’ll be square in the NL MVP conversation.