ZiPS Has Good News for the Cardinals

A few weeks ago, I went on a mini-rant about the 2021 version of Fangraphs’ Steamer projections. Steamer looked at the 2021 version of the Cardinals’ pitching staff and concluded that it would be one of the worst staffs the club has fielded in years.

Depth Charts believes the Cardinals pitchers will regress by a half run overall to a 4.43 team ERA. As I said at the time, that would require every pitcher on the staff to lose .5 ERA from their 2020 performance or half the arms to lose 1.00 ERA with no one improving.

The cumulative fWAR for the entire staff was just 11.2. That’s close to the final totals of the forgettable 2007 Cardinals that won just 78 games. As the pitching goes, so go the Cardinals. Fangraphs projects the Cardinals to win 78 games next season.

The Steamer projections simply aren’t very realistic. While some of the individual projections can hold up to scrutiny, the cumulative effect is so reductive that it’s tempting to throw the whole model out.

ZiPS – Dan Szymborki’s proprietary projection model that Fangraphs also publishes – is significantly more optimistic, while still maintaining its “Dan hates your team and all its players” objective approach.

As stlcardsfan4 wrote when the projections came out, Szymborski’s computers believe the club is pretty much the same as it’s always been. It has them landing in the 85-90 win range. That’s at least 7 wins higher than Steamer’s estimate.

That’s good news for the front office, which is hog-tied by the ownership groups’ budgetary demands. Regardless of what the team’s final budget report says – if they were in the black or the red, as they claim – the team desires to recover the revenue they lost from no ticket sales in 2020.

Payroll is already approximately $40M below last year’s 162-game Opening Day totals and there is little indication that the team plans to spend much this offseason. If the club wants to remain in contention – which they claim they do – they will have to rely on their existing players.

That means that team will need to be closer to Szymborski’s ZiPS projections than Steamer. How does ZiPS make up the 7-12 win gap between the two models? Let’s take a look at some of the critical projections for the offense and pitching and see where the gains are coming from.

Here are the ZiPS and Steamer projections that I’ll be working from.

STEAMER VS. ZiPS: OFFENSIVE PROJECTIONS

In the chart below, I’ve included the top 8 offensive players for the Cardinals, representing the starters at each position sorted by production. That means, for example, that Tyler O’Neill is included and Dexter Fowler is not because O’Neill is the more valuable player (though only slightly).

I’ve also included projections for both Andrew Knizner and Yadi Molina merely for convenience. We’ll only consider Knizner for now.

Below this list are a dozen options for the bench and spot-start roles, but none of the players come in much above replacement level. On both models – Steamer and ZiPS – their value added to the club is negligible.

The first thing you might note is that the projections are very similar down the line. There are no huge discrepancies of which to take note. Still, a dozen small differences add up to notable production. That’s the way baseball works over 162 games. A bunch of small moments can mean the difference between a winning season and a playoff bid or a sub-.500 record.

That’s what’s happening here. Consider the WAR totals for the starters (not counting Molina) for both systems:

STEAMER WAR: 14.2
ZiPS WAR: 15.8

That 1.6 WAR difference is probably within the margin of error for both systems. Still, ZiPS is more optimistic across the board. Every player except Molina has a higher ZiPS projection than Steamer.

It’s also not hard to see the areas where the club could overachieve. Carlson’s projections are comparable and very reasonable considering his youth and lack of MLB playing time. Still, neither projection reflects the offensive talent the young slugger has shown during his pro career. His 1.8 ZiPS WAR could become 3 WAR or higher if he reaches his in-his-prime ceiling now. Don’t count on it but don’t count it out.

O’Neill’s offensive projections have some space for growth, too. He could make up 1-1.5 WAR with his glove and baserunning alone, if he has enough outfield innings. Overall the offense probably has slightly more room to grow than it does to fall, assuming health and that the team doesn’t make silly decisions in playing time allotment.

STEAMER VS. ZiPS: PITCHING PROJECTIONS

The pitching chart below follows the same philosophy as the offense. I’ve picked the top 5 starters (as of today – no Wainwright, though he only improves team projections) and the top four right-handed relievers. I stuck with the righties to keep the size of the chart down and because their projections better illustrate the type of variance that exists between ZiPS and Steamer when it comes to the Cardinals’ pitching.

Simply put, the Cardinals’ season hinges more on how Gallegos, Reyes, Hicks and Helsley perform rather than Trevor Miller and Tyler Webb.

I also had to cut out some names that will be vital parts of the clubs because their innings totals make an honest comparison difficult. John Gant, for example, has a 70-80 inning difference in their projections, which has significant implications on WAR totals. Cabrera and Ponce de Leon also fill these swing-starter roles, but Helsley and Reyes provide good “representative models” for this type of player in ZiPS vs. Steamer without having to do innings conversions.

Maybe you can see why I was so frustrated with Steamer’s pitching projections in my previous article! The ERA (and therefore overall runs scored vs. runs allowed) regression for the whole squad was just too extreme. It dumbed every measurable skill that the Cardinals’ arms and defense have had down to a baseline that rendered them average or below.

ZiPS doesn’t do that. Take Jake Flaherty as an example. Despite a nice 10.04 K/9 and 3.31 BB/9 projection, Steamer still regressed the Cardinals ace to an ERA/FIP right around 4, throwing him below 3.0 fWAR. Why? That doesn’t fit Flaherty’s history.

ZiPS, though, bumps his K rate – as it should – cuts his BB rate – as it should – and concludes that Flaherty is one of the better starters in the league – as it should.

That kind of difference continues down the line. Of the rest of the starters, Kim, Mikolas, and Gomber all have better ZiPS projections compared to Steamer. Martinez’s ZiPS WAR projection is only lower than Steamer’s because of innings. Flaherty, Kim, Martinez, and Gomber also have fewer innings pitched in ZiPS than in Steamer. The production gap between ZiPS and Steamer for the rotation would be even greater than it currently is if the innings were neutralized to Steamer’s Depth Chart projections (which we’ll do below).

The same thing happens in the bullpen. There our representative samples all have ZiPS projections that are significantly better than Steamer. Gallegos is a good example. His Steamer projection of 4.03 FIP doesn’t fit at all with his major league performance to this point. ZiPS, though, things the dominant righty will have a 3.21 FIP. I can’t argue with that! The difference is .8 FIP.

So it goes throughout the staff. ZiPS’ correction of Steamer’s regression makes it easy to find the win difference between ZiPS (87 wins) and Steamer (78 wins).

Here’s the difference between Steamer and ZiPS WAR for just the names listed in my representative sample:

STEAMER WAR: 10.8
ZiPS WAR: 15.1

When considering the whole pitching staff, that gap jumps even more:

STEAMER DEPTH CHARTS WAR: 11.2
ZiPS DEPTH CHARTS WAR: 18.6

There’s the win difference that we’re looking for!

ZiPS is slightly more optimistic about the offense – around 1-2 WAR added.
ZiPS is significantly more optimistic about the pitching – around 7-8 WAR added.

Total difference: 8-10 WAR or 87 wins compared to 78.

What does that mean? It means the Cardinals better hope that Dan Szymborski’s computers are right. My gut – and my analysis of this Cards club – would line up behind him.

The Cardinals are not as bad as we might have thought. They might even be the best team in the NL Central today, though I’ll have to look at other projections as the offseason progresses to determine the veracity of that claim.

If the club adds depth with Wainwright and Molina and secures 1-2 other pieces, particularly on the offensive side, they could be pushing 90 wins again with their 50^th percentile projections.

How’s that for some good news this Christmas weekend?

2020 is almost over. It’s time for us to get optimistic about 2021. It could be a better year than we think.