Saturday SOC: Fun ZiPS Notes from Dan Szymborski

Over the holiday break, Fangraphs Senior Writer and creator of the ZiPS projection system Dan Szymborski took time to join Ben Godar and Ben Humphries on the fantastic Cardinals Off Day podcast.

Ben and Ben are former VEB contributors (the second Ben by alphabetical order is a former Site Manager around here). The podcast, which publishes during the days off during the season and sporadically during the offseason, has a strong VEB vibe to it. The quality of the conversation is high. So is the analysis. And the humor.

You can find their Substack here to read up and follow the podcast.

Or catch up on the podcast directly from Apple here. I’m sure it’s also available through Google and any number of other sources. Check your podcast app of choice for details.

You can also follow them for updates and witty commentary on their Twitter account:

I’ve been wanting to write about the podcast for a while since they’ve given a few of my articles some attention and compliments. I’m socially obligated to repay the favor and happy to do so.

With Szymborski joining in last week, this seemed like the perfect time. The ZiPS projections for the Cardinals were published just before Christmas and they were very encouraging. You can find the report and write-up here at Fangraphs. There have already been several posts addressing them around here but it’s pretty awesome to be able to hear directly from Dan himself about the process and some of the results.

What follows are some pretty random highlights of the conversation:

1. Age 43 is middle-aged. Apparently.

This has nothing to do with anything and I shouldn’t lead off with it, but Szymborski noted that he’s now 43 and has been doing ZiPS projections for half his life. That means he is, by his measure, “middle-aged”. I also will be turning 43 this year. I had never thought of myself as “middle-aged” before but if Szymborski says it, then there’s at least a 50% chance it’s true, just based on probability modeling. The computers, after all, don’t lie.

I also have been talking baseball on the internet for half my life. I wasn’t cool like Dan and Dave Cameron, talking sabermetrics as teenagers in obscure corners of the early internet. Instead, I filled the old Post-Dispatch Cards Talk forum with deep-level analytic takes. Like, “if Larry Bigbie can just get better at hitting and fielding and walking, he’ll be a pretty good baseball player for the Cardinals.” Or “Juan Encarnacion could really take a step forward at age 29 by playing alongside such veteran talent in StL.” How’s that for a projection system? I had to reach “middle-age” before I had a take that a few people actually wanted to read.

2. The Cardinals are the Toyota Camry of Baseball.

This year, again, ZiPS projects the Cardinals to win between 85 and 90 games. They’ve done this with a high floor of quality production but a low ceiling of superstar production. Just like the Camry, it’s a quality car that can retain its value on the market. But it’s not going to win any style points.

From a projection standpoint, that seems accurate enough, even though this team has some superstar-caliber names. Nolan Arenado is a superstar in this league, even if his production is starting to tick down. Paul Goldschmidt has been one of the most productive players of his generation. Yadi Molina. Adam Wainwright. Superstars at the very end of their career.

Nolan Arenado has the highest projection on the Cardinals with a 4.2 zWAR. (zWAR is just ZiPS WAR). Very good? Yes! As a point of comparison, superstar Fernando Tatis Jr. clocks in at 7.6 zWAR. Goldy is down a bit at 3.4. Molina is at 1.1. Wainwright is at 2.0. The best shots at future superstar production are probably Tyler O’Neill (3.9) and Jack Flaherty (2.4), who have both flashed elite production for spurts but don’t have the consistency to earn those projections from Szymborski’s computer.

The Cardinals project to be a good team again. A reliable contender. But they’ll need some players to take a step forward and beat projections to have a chance to challenge the loaded teams at the top of the NL.

3. ZiPS loves Tyler O’Neill and he has a crazy 10% top projection.

ZiPS incorporates batted ball data into its evaluations, which is great news for O’Neill’s future projections. Often we would expect statistical oddities to progress back toward the mean, especially where sample sizes are limited. Apply that logic to O’Neill’s BABIP. Last year he produced an incredible 5.4 fWAR but it came with a .366 BABIP. Regress that to the .300’ish mean and suddenly everything about TON’s performance drops dramatically, from his .286 batting average to his .274 ISO (isolated power).

ZiPS, though, doesn’t just regress outliers on principle if there is data to support it. That’s where O’Neill’s crazy ability to generate incredibly hard (and relatively consistent) contact comes into play. Batters with the ability to hit the ball hard consistently are more likely to sustain a higher-than-average BABIP.

So, yes, ZiPS has a 50^th percentile projection for TON that cuts back his slash line, but it’s bullish on that power. O’Neill had a .274 ISO last year. ZiPS thinks he’ll have a .270 ISO next season. That’s a 4 zWAR season in 554 projected plate appearances. Considering TON’s up and down career, that’s a nice projection!

The key there is 50^th percentile. That means that O’Neill has a 50 percent chance of producing at that level. It also means, according to Szymborski, that O’Neill has a 10 percent chance of doing something ridiculous – like hitting .311/.384/.667 with 44 home runs. That’s a 1.051 OPS. For those counting at home, that’s better than Bryce Harper’s league-leading OPS last season. (I won’t give you the 10% numbers on the downside. It’s Saturday.)

4. ZiPS likes Tommy Edman more than you do.

The Ben’s have been universally down on Tommy Edman for a while. Edman gets a ton of plate appearances batting at the top of the Cardinals lineup. There, he doesn’t do anything all that well. He limits K’s by putting a lot of balls in play but can’t generate a high batting average. He doesn’t walk. He doesn’t have much power, despite his high doubles totals (which are tied more to his speed and high number of plate appearances than his power). He does play very good defense; he earned the Gold Glove that he won.

Strangely, ZiPS has Edman beating his slash line from both 2020 and 2021. Some of that might be the influence of his 2019 rookie season. Part of it, though, is that ZiPS just expects some improvement from the Cards’ keystone leadoff hitter. It goes back to what Szymborski has said about the Cardinals: a bunch of low floor, high ceiling players.

What I love, though, about Tommy Edman’s projection is the Tony Fernandez comp. Fernandez was one of my favorite players growing up – mostly based on his baseball cards. He was a great fielder. He usually hit for a good average. He could steal bases. He rarely struck out. Since I didn’t pay attention to walk totals back then, I would have thought he was a better hitter than his 102 career wRC+ shows me now. Still, Fernandez could rack up the production. He’s solidly in the “Hall of Very Good” with 43.5 fWAR.

That’s Tommy Edman. I mean, Edman probably won’t generate half that career WAR but they sure make sense from a comp perspective. You absolutely wanted Tommy Edman/Tony Fernandez on your team in 1987. You’ll probably underrate him in 2022.

5. Nolan Gorman can play defense at second base.

Just a quick note here about Nolan Gorman’s projection. Gabe noticed in his original ZiPS article that Gorman’s defensive projection at second base was inexplicably high. There was some discussion about it and we all kind of wrote it off as “we’ll see”.

Well, Szymborski explained that ZiPS tracks where balls are hit and develops a “probabilistic measure” to translate minor league defense. So, that +7 defense from Gorman isn’t just speculative. It’s based on actual data.

It’s also just 1 behind Tommy Edman. If Gorman’s defense is that much better than anticipated, then it’s going to be hard to keep him off the field. His bat already projects at or near Edman’s levels, and that assumes ZiPS is correct about Edman’s offensive improvement.

6. Steven Matz is Dr. Thunder. Or one of many other choices.

We’ll end with this contribution from Dan on Twitter in response to Ben and Ben’s description of Steven Matz as “Dr. Thunder” to Marcus Stroman’s Dr. Pepper.

I have had at least 6 of these. Can anyone beat my number of generic Dr. Peppers? I’m sure some of you can! My rarest one would probably be the old style of Dr. Topper.

Anyone ever have a Mr. Sipp?

Enjoy your Saturday! Don’t worry. Next week, I’ll be going back to analysis articles. Time to start some season breakdowns and previews.