clock menu more-arrow no yes

Filed under:

Visualizing Percentile PECOTA Outcomes for the Cardinals Offense

New, 27 comments

A follow-up to last week’s article

Kansas City Royals v Oakland A’s Photo by Owen C. Shaw/Getty Images

Last week, I put together a point-counterpoint for several aspects of the Cardinals entering 2020. As each point and counterpoint was made, it became (seemingly) apparent that there’s a lot of volatility to the upcoming season. Fluid situations in left field, the rotation, the bullpen, potential bouncebacks for key returners, and overall team speed and defense are going to dictate how far the Cardinals advance this year.

One of the easiest ways to capture this type of volatility is to use the PECOTA projections from Baseball Prospectus. They produce a 50th percentile outcome as the most likely, but they also provide outcomes in intervals of 10 percentile (e.g. 10th, 20th, 30th, etc.), along with 1, 5, 95, and 99th percentile. Speed and defense stays the same in their projections along the percentiles, so those questions will go unaddressed today. Pitching is a little different if only because Carlos Martinez is projected for the bullpen by PECOTA, Kwang-Hyun Kim is projected as a starter, and other smaller/relief options have roles that may drastically change. As such, I’ll shelve the pitching for today. However, that still leaves offense.

Here’s how PECOTA projects OPS for the Cardinals. You’ll find their 10th and 90th percentile projections, along with their 50th percentile projections for both PECOTA and ZiPS.

That sure looks like a wide range of outcomes. Last week, I identified Matt Carpenter, Harrison Bader, and Paul Goldschmidt as key bounceback candidates. That Carpenter and Goldschmidt each have 90th percentile projections over a .900 OPS is a good sign- in the very least, they have a path in these simulations where they post All-Star quality numbers. For that matter, Bader is just barely under a .900 OPS. If you’d rather something more modest- perhaps a 70th percentile projection is more your speed, even though it’s not in the graphic- Goldschmidt is still over a .900 OPS, Carpenter is at .822, and Bader still cracks .800.

The left field scenario is where it gets wacky, with a caveat. You’ll see a broad range for Lane Thomas, with a 10th percentile projection near .600 OPS and a 90th percentile just under .900. They project most players in buckets of plate appearances. For instance, most everday players like Edman and Carpenter are in the 500s, while someone like Goldschmidt- almost always healthy and surely an everday player if ever there was one- breaks 600. A player with a platoon split like Bader is in the 400s. For Thomas, it was just 154, meaning his projections are going to be noisier. O’Neill has a similarly broad range (.632 10th percentile, .900 90th) with more plate appearances projected (308). Even Austin Dean has a lot of outcomes where he does a good job of filling in for José Martinez.

Dylan Carlson is where it gets especially intense. With 287 plate appearances projected, his 10th percentile falls to .627 while his 90th percentile is a whopping .926. His gap between 10th and 90th percentile OPS is .299, the 40th largest gap among 565 players.*

*NOTE: There are huge swaths of players with 251 projected plate appearances, and almost all of them are for-sure minor leaguers. I’ve removed all of them.

It may seem like there’s a lot of volatility here. Well if that is what you think, I’ve got something to tell you. Something that may shock and discredit you. And that thing is as follows: I’m not wearing a tie at all. Er... wait, I’m not Lionel Hutz. The shock is that the Cardinals have a fairly normal- even smaller- amount of variance compared to most teams. We’ll take the gap between 10th and 90th percentile projections for each player, then weight it based on BP’s projected playing time. The Cardinals have the 28th lowest amount of variance in OPS, and 25th lowest if you prefer dRC+. In other words, the gap between the Cardinals floor and ceiling on offense is one of the lowest in baseball.

The average (weighted) gap, league-wide, is .248 points of OPS. Here is the same batch of players from the previous graph with the gap between their 10th and 90th percentile OPS projection, and a dotted line for the league average gap:

There’s not a lot of variance compared to the league for Yadier Molina, Kolten Wong, Paul DeJong, Dexter Fowler, Matt Carpenter, Paul Goldschmidt, and Tommy Edman. In Molina’s case, he has the 18th smallest gap. Those seven are going to eat up a lot of playing time, and it means the team’s offensive floor and ceiling is mostly limited. If they’re going to make a huge move, it’s going to have to come from the aforementioned Four-Headed Hydra of Lake Leftfield in the Argolis (née the Four Horseman of the Leftfieldpocalypse). Carlson, Thomas, O’Neill, and Justin Williams have the wildest variance on the team, and Harrison Bader isn’t too far behind. Admittedly, some of that is the limited playing time projected by PECOTA. Less playing time means more variance.

That’s a lot of fancy graphs to end up where we began. Sometimes, the answer really is the most obvious version possible. The Cardinals offense is going to be heavily reliant on left field and how well Goldschmidt, Carpenter, and Bader rebound.