Good morning, Viva El Birdos!
I’m writing this article on Monday afternoon. The sun is out. It’s warm. We’re getting a little hint of what’s coming.
Spring! And with it Spring Training! Yes, friends, baseball is almost back. The trucks are leaving Busch and heading south. That gives me just a few last chances to cover the off-season topics I wanted to cover before we have balls hitting mitts, players in the best shape of their lives, and the eternal optimism of Florida baseball.
Today, I want to return to a throw-away point I made way back in the heart of winter and never explained. It was a podcast episode. We were talking about our expectations for the rotation, and I said something like this: “Kyle Gibson is the key to making the rotation work.” That elicited one of Gabe’s patented smirks of skepticism and I promised I would detail what I meant later.
Well, it’s later.
I’ll admit that this is an odd statement and it might not resonate with all of you. After all, the Cardinals rotation features Sonny Gray. He’s their best starter. It’s not a debatable point. For this rotation to “work”, Gray has to be healthy and productive. That’s a given.
The team, though, can’t just rely on Gray being Gray. Someone else has to be able to provide a stable floor of production, a secure pivot point around which everything else can fluctuate.
A pitching staff is volatile. Starters are volatile. Injuries. Changes in performance. Luck with batted balls. You never really know what’s going to happen to the top five guys in any team’s rotation.
That’s why the more stabilizing factors a team has, the more locked-in, sure things – even if those sure things are only average-to-good players – the better.
That’s what I think Gibson provides. That’s why he’s the key to making the rotation work.
When I evaluate a player there are three factors that I try to look at and hold in balance. Those three factors are historic production, projected performance, and actual ability. Here’s what I mean by each of these terms:
Historic Production: the level of production that a player has provided in previous seasons, measured through a variety of statistics.
Projected Performance: the level of performance that we – ourselves, the club, or even projection systems – expect a player to provide in the upcoming season.
Actual Ability: our assessment of a player’s tools and skills using either measurable stats, scouting reports, or the eye test.
When we evaluate a player, we tend to take all of these elements, throw them into the pot, turn up the analytical heat, and see what kind of stew comes out the other end.
That, naturally, brings subjectivity into the equation. While we have uncaring, unthinking stats to measure all of those categories above, it’s up to us, our minds, our thoughts, our tendencies, our biases, to put them together into consumable content: written words, spoken opinions, or detailed predictions.
Depending on the player, we tend to weigh the above categories differently, emphasizing categories we believe in at the expense of others for individual players. I think we’ve seen some of this in our offseason analysis of the Cardinals’ rotation.
I’ll give you an example: Will Lance Lynn be good or bad in 2024? You might give a lot of credence to his historic performance and be pretty optimistic. Or you could be overly swayed by the decline in his stuff last season. We have a wide range of opinions on the site. Why? Because there’s a wide range of variance in the categories we use to evaluate a player like Lynn.
For us, this variance doesn’t matter much. We’re either right or wrong on the internet.
For the Cardinals? It’s everything. Variance in expectations vs. performance is the difference between games lost and games won. It’s the difference between a player meeting expectations and not.
There are lots of ways I could explain or illustrate this, but I think a Venn diagram will work pretty well. My goal, after all, is not to dig heavily into the statistical analysis for Gibson or the other starters, but to describe the way that balancing of a pitcher’s performance characteristics can sway our evaluation of players and a rotation as a whole.
Here’s how I would put the above categories into a Venn diagram with, roughly, equal weight:
That area in the center of the circle is the “sweet spot” for player evaluation. That’s the place where projections, past performance, and actual ability align. The more overlap that we have in those three areas, the more likely that a player’s performance is going to land there. The less overlap we have, the more likely we are allowing one category to sway us over the others.
Sonny Gray illustrates this nicely. If you were going to evaluate Gray as a pitcher, how would you weigh the above categories?
Historic Production: Gray has a strong history of performance as a quality MLB starter. Over the last five years, Gray has produced 16.3 fWAR in 670 innings (including COVID-shortened 2020). That’s an average of 134 innings per year and 3.3 fWAR produced. He has a high of 5.3 fWAR in 184 innings.
Projected Performance: Projections for Gray somewhat reflect a 50th percentile version of his historical stats. ZiPS believes he’ll provide 2.4 fWAR in 154 innings. That’s quite a bit lower than his annual averages and is likely due to ZiPS applying an age curve. ZiPS gives Gray his lowest K/9 since 2016 and his highest FIP since 2019. The computer thinks he’ll have one of the highest HR rates of his career. Still, considering his stable performance history, this seems a bit too regressive for me, even for a 50th-percentile projection. My expectations are a bit higher than ZiPS for Gray.
Actual Ability: If you head over to Gray’s Baseball Savant page, you’ll see quite a bit of “red ink” throughout his Statcast metrics. Red ink indicates a pitcher is average or better in any given performance metric. The redder the better. Gray benefits from just not being bad at anything. He’s pretty much at or above average in every conceivable pitching metric and is well above average at ground ball rate and walk rate. He has very good pitch ability.
That’s our evaluation. How would we reflect that on a Venn diagram? Maybe something like this:
This is where subjectivity comes in. I would give very strong weight to Gray’s historic performance and I have a high level of confidence in his actual ability. There’s a strong level of overlap between those two categories, as reflected in the diagram above. His projections though? I’m not sold on them. They don’t overlap as well with the other two categories. I pulled it away from the center a smidge.
That’s variance. That might be insightful analysis. Or subjective bias. We don’t know yet. Either way, it throws a little bit of unknown into Gray’s analytical equation.
That’s par for the course when it comes to starters. And the result is still a Venn diagram that still has a very high level of overlap between our three categories. We can say with a high level of confidence that Gray is a good bet to provide really good production for the Cardinals.
What about the rest of the rotation? Let’s look at more extreme ways that this kind of variance enters into our evaluations of players.
For the last few years, Mikolas has served as the club’s defacto #2 starter behind, at various times, Flaherty, Wainwright, and Montgomery. With Gray in StL, that designation remains true. Instead of acquiring another starter who is ahead of him, the club is relying on Mikolas to bounce back to his previous level of production. Is that wise? I’ve argued elsewhere that it isn’t. How would I weight each of his categories?
Historic Production: Mikolas has a fWAR per 200 of 3.1 in his Cardinals’ career. He matched that perfectly last season. The fluctuations in his performance annually are based on his HR rate and HR/FB%. Those can vary pretty naturally even when his stuff is even. Last season, his production was built on an unsustainable low HR/FB rate. That pushed his FIP well below his ERA and xFIP, which were both below average. His relatively low FIP helped him reach his fWAR averages, but it’s a tenuous situation at best.
Projected Performance: Projections for Mikolas strongly reflect this tenuousness. They can’t ignore Mikolas’ significant injury history, giving him just 165 innings pitched. And ZiPS sees the same problem with his HR rates. The computers drop him to just 1.7 fWAR. That’s a massive projected decline.
Actual Ability: I detailed this fairly well in the article linked above. Mikolas’ stuff tanked last season. There’s blue ink – some dark blue – throughout his Baseball Savant page. He was well below average in pretty much every significant pitching category other than BB rate, where he still excels.
That leaves us with a Venn diagram with a significant overlap between Mikolas’ projected performance and his actual ability, but less overlap between those two categories and what Mikolas has done historically. That creates a pretty strong likelihood for variance and, considering the strong correlation between his projections and stats, most of that variance would push his performance down relative to his history. The club is counting on a specific level of production from him based on what he’s done when healthy in the past. They need that production from him. They aren’t very likely to get it.
Matz signed a longer-term deal for a relatively low AAV back before 2022. He’s struggled with a combination of injuries and ineffectiveness since, but expectations have never been that high for him. That continues to be true.
Historic Production: Matz has a career-high of 2.7 fWAR in 150 innings. He produced 2.0 fWAR in just 105 innings last year. He’s a good bet to provide less than 150 innings pitched and 2.0 fWAR or lower every season.
Projected Performance: ZiPS falls along with his historic performance, giving him a 1.7 fWAR projection in 103 innings. That reflects who he has been and what he has done. He’s been consistently inconsistent and unavailable.
Actual Ability: Still, his projection shows us something about Matz’s true ability. At 200 innings, that’s a 3.0 fWAR projection. Matz should be a good pitcher! That ability shows up in Matz’s rate stats on Baseball Savant. He sits between 35-70th percentile in pretty much all key pitching metrics. He does everything well enough to excel. He just hasn’t.
Matz’s Venn shows a pretty strong alignment between his projections and historical performance. What’s out of balance is his actual ability, which exceeds both. That fits well with how the Cardinals view him. Matz is a #4 starter. He is, in some ways, the best kind of #4 starter. Back-end rotation arms are by nature volatile. Well, Matz is volatile. Expect injury. But when he’s on the mound, the Cards can be fairly certain that he’ll pitch relatively well. And there is always that chance that he’ll stay healthy one season and the club will collect 2-3 fWAR from him.
Based on salary and the bonus’ built into his contract, Lynn seems like he’s the Cardinals’ #5 starter. It’s really hard to have high expectations of such a pitcher. At the same time, they’ve made it clear they are counting on consistency and innings from him. Can they get it?
Historic Production: Lynn was one of the better pitchers in the game during his prime years. As recently as 2021 he produced 4.3 fWAR. That wasn’t that long ago and likely weighs heavily in the Cardinals’ decision to target him. It certainly wasn’t last year’s performance, where he produced just .5 fWAR with an ERA/FIP over 5.5.
Projected Performance: That kind of a dip in performance at that age is going to impact Lynn’s ZiPS projections. ZiPS has Lynn with a declining K rate, a slightly improved BB rate, and a terrible HR rate. The result is a 1.7 fWAR projection in 141 innings. That’s bad for Lynn historically but not that terrible for a #5 starter.
Actual Ability: Lynn’s stuff hasn’t declined at the same rate as his performance. He can still generate whiff’s and Ks, and control his BBs relatively well. He just gives up a ton of fly balls and fly balls leave ballparks. The ability is probably still there for him to be better than an average #5 starter.
The result is a Venn diagram with a good balance between Lynn’s projections and his actual ability. The outlier is his historic production. Here we see significant variance – and it could go either way. Lynn could have one more season of high-level production in him, like 2021. Are the Cardinals counting on that? I doubt it. He could also continue falling off the cliff, continuing his decline from ’23. Do the Cardinals expect that? Of course, they don’t. Which way will Lynn go? Your guess is as good as mine.
That brings me back to the thematic subject of my article. We’ve looked at four of the Cardinals’ starters and only one of them is what we might consider a constant for the Cardinals in terms of likely production. Sonny Gray has been good, should be good, and has the stuff to be good.
The other three have too much variance – maybes – built into them to provide the same kind of production security. Some of those maybes could move upward, exceeding the Cardinals’ expectations. Matz, for example, is more likely to trend up than collapse. Some of those could move downward. Mikolas seems almost certain to regress based on the club’s likely expectations.
With one solid at the top of the rotation and three variables, what the Cardinals need is one more stabilizing force. One more starter who can provide a secure production evaluation, a consistent balance between ability, history, and projections.
That’s exactly what Kyle Gibson does. He can’t provide the same overall quality as Gray. But his remarkable consistency and balance can function as a pivot point, a set value in the middle of the rotation to protect the club from the variance that surrounds him.
Historic Production: In 4 of the last 5 seasons (not counting 2020), Gibson has produced between 2.6 and 3.1 fWAR. In the one outlier, he produced 1.9 in just 167 innings, which is a 2.3 fWAR per 200 IPs. Even when Gibson has been a little off, he’s been close to WAR’s margin of error. During that span, his FIP high was 4.28. His low was 3.87. That’s an effective range of variance of .4 FIP over almost 1000 innings. You won’t find a more consistent pitcher.
Projected Performance: ZiPS takes all the consistency and then, for some reason, says, “let’s regress it!” ZiPS gives Gibson a head-scratching 1.6 fWAR projection in just 150 innings. Why? Beats me! For some reason, ZiPS thinks Gibson will have his highest FIP since 2017 and neutralizes his remarkably consistent K and BB rates. I would guess that aging curves have something to do with this. It’s very similar to Gray’s projection above.
Actual Ability: Gibson has average or below stuff. Still, he’s able to generate whiffs at about an average level and excels at controlling walks and generating ground balls. In terms of rates, he’s a poor man’s Sonny Gray. He’s a rich man’s Steven Matz. There is some evidence of decline in his chase rates, but largely I just see a relative consistency in his stuff that matches the stability in his performance.
If we plot that, we end up with a very strong overlap between Gibson’s historic performance and his actual ability. He just sort of is what he is. That’s not necessarily reflected perfectly in the ZiPS projections, so we have to kick out his projections a little. We shouldn’t ignore what ZiPS is doing. But how much do we weigh them? Like with Gray, I lean low.
Kyle Gibson is the Key to Making the Rotation Work
A baseball season never goes quite how we expect in the offseason. Having players who are more likely to reach their expected levels of production because of the consistency and balance in their history, projections, and ability should help mitigate the volatility that will arise along the way.
If we look back through our diagrams, we can rate the level of variance that the Cardinals have built into their rotation. How likely are these five starters to reach the level of production that the club expects, based on their histories, abilities, and projections? What kind of impact might they have on the rotation is they fail to reach expectations?
I’ll answer these two questions in the form of an equation plus a definitive statement.
Sonny Gray: High level of expectations + low level of largely neutral variance in ability, history, & projections = high chance of meeting the club’s high expectations.
If Sonny Gray collapses or has a significant injury, the Cardinals are in real trouble.
There is no replacing Gray. As Gray goes, so goes the Cardinals. But they have every reason to believe he’ll be just fine. He’s also not enough.
Kyle Gibson: Moderate level of expected production + low level of variance in ability, history, and projections = high chance of reaching the club’s moderate expectations.
If Kyle Gibson collapses or has a significant injury, the Cardinals are going to have to beat the odds elsewhere.
Simply put, no one else can provide the same kind of guaranteed stability in the middle of the rotation as Gibson. Someone has to. Someone has to provide 2.5 fWAR or better in production over 160-200 innings to be a complement to Gray. Gibson is the best bet to do so. That’s why he’s my key to making the rotation as a whole work.
If Gray are Gibson meet expectations, and they should, then it removes a lot of pressure on the rest of rotation, where we find more volatility.
Miles Mikolas: High level of expected production + high level of variance (mostly downward) in ability, history, and projections = low chance of reaching club expectations.
If Mikolas collapses or has a significant injury, the Cardinals are exactly where I expect them to be.
I already have quite a bit of decline built into my expectations for Mikolas. Simply put, either Gibson or Mikolas needs to provide #3 caliber production for this club. Both can do it. Both have done it. Ideally, both will do it. But I am way more confident in Gibson than Mikolas at this point in their careers.
That’s why, to me, Mikolas is a higher-level wild card in this rotation equation rather than the secure #2 starter he has occasionally been in the past. Mikolas can significantly improve the quality of this rotation. He can’t break this rotation if Gibson does what we expect.
Steven Matz: Low level of expected production + moderate level of upward variance in ability, history, and projections = high chance of reaching club expectations.
If Matz collapses or has a significant injury the Cardinals are fine.
The Cardinals can just replace Matz with an internal starter and likely not experience a significant change in the projected performance of their rotation. I would take Matz over Thompson or Liberatore but the difference is not significant. On the other hand, there’s always the chance that Matz stays healthy and excels this season. If he does, that’s pure bonus production! Matz can really help this rotation. It’s going to be hard for him to hurt it.
Lance Lynn: Low level of expected production + high level of variance (up or down) in ability, history, and projections = a total crap shoot.
If Lynn collapses or has a significant injury, the Cardinals don’t even blink.
There’s a pretty good chance that Zack Thompson and/or Matthew Liberatore will be just as good or better than what the Cardinals expect Lynn to be. Their projections are certainly similar. All Lynn can do is bump the Cardinals up. If by some miracle he can recover his ’21 or ’22 form, this rotation could shine.
There are a lot of things that could go right for the Cardinals rotation in 2024. There are a lot of things that can go wrong. Generally speaking, if Gray and Gibson do what we expect them to do and the team doesn’t have terrible injury luck with the other three arms, the rotation should be just fine. It would be more than good enough to support an excellent offense and contend in an NL Central that can’t be bothered to try to contend against the best teams in MLB.