Digital Love: Why ZiPS Thinks Lucas Giolito is Top of the Charts
In the 2021 ZiPS projections that are live on this very website, Lucas Giolito is projected with the most WAR of any pitcher in baseball. If my social media is any indication, this projection is, so far at least, the source for the most joy and the most consternation. What is it about Chicago’s young ace that gives him such an aggressively optimistic projection?
One of the common complaints you see from readers about projections is that they don’t go out on a limb very often. To me, this makes perfect sense: when talking about the mean projections, massive performance changes should rarely be the player’s typical expectation. Think back on José Bautista back in 2010. At the time, Joey Bats was a player pushing 30 who had hit .238/.329/.400 in the majors in more than 2,000 plate appearances for five major league teams. As we now know, his destiny was to explode on the scene, slugging .617 for the Blue Jays, resulting in the first of his eventual six All-Star appearances. But that doesn’t mean that his baseline projection going into 2010 should have reflected that result; that unlikely things happen doesn’t mean that they weren’t unlikely.
Projections do go out on a limb, but in a conservative sort of way. One of the more notable examples in recent years is the 2019 ZiPS projection for Shane Bieber. With a 4.55 career ERA (but a 3.23 FIP!) in just 114 2/3 major league innings entering the season, ZiPS gave Bieber an optimistic projection as the 14th most valuable pitcher in baseball with a 3.71 ERA over 187 innings for 3.8 WAR. That was enough to put him just behind Clayton Kershaw and ahead of notables such as Zack Greinke, Noah Syndergaard, Stephen Strasburg, and Patrick Corbin. Not only did Bieber meet this projection, but he also exceeded it, finishing eighth in WAR among pitchers at 5.6.
But why did ZiPS love the Beebs so much? It wasn’t one factor. Instead, it was an accumulation of smaller positive factors that significantly outnumbered the negative ones. Bieber was only 24. He had a record in the minors and majors that suggested he could avoid lofty gopher ball totals. His BABIP in his rookie year was extremely high. From his quality-of-contact data, ZiPS thought that batters “should have” hit .252 and slugged .415 against him in 2018 when the actual numbers were .285 and .467. And so on.
It’s the same thing for Giolito entering the 2021 season. No, he isn’t in the same position as Bieber was entering 2019, given that he’s already received Cy Young votes in two seasons. But it does take something special to get ranked No. 1 with a bullet. His top-notch projection doesn’t mean that it’s going to happen, only that it’s more likely to occur than the downside scenarios. And the latter do exist: ZiPS projects a 15% chance that Giolito will be worse than a league-average starter and about a 7% chance he’ll have an ERA on the wrong side of five, both things that would have a very negative effect on Chicago’s 2021 fate.
So what are some of the reasons for this digital love?
Lucas Giolito is Still Relatively Young
Pitchers don’t have a typical age curve, but it’s still preferable to be in your twenties than in your thirties. Like Bieber, the height of Giolito’s ceiling remains an unknown. Both Gerrit Cole and Jacob deGrom are amazing pitchers, but after a few more years in the majors, there’s less uncertainty about their remaining upside. Superstars in their mid-20s, on the other hand, frequently have another gear or two remaining. Among Giolito’s top comparables were a multitude of youngish pitchers who were already stars and did have such performance bumps remaining: Greg Maddux, Jose Rijo, Dave Stieb, Brandon Webb, and so on.
Of the top 50 most comparable pitchers in Giolito’s cohort, 32 of them beat their baseline performance estimates over the following three seasons — an astounding rate of success given how attrition claims pitchers. For Cole, that number is only 23; for deGrom, 26; and for Max Scherzer, 19.
Lucas Giolito is Well-Suited for his Home Park
Home run rates for pitchers are volatile, but they’re not random. Giolito fares well in velocity and barrel-based numbers and was in the top-tier in most of these measures in 2020. Among the pitchers projected in the top 10 overall in ZiPS, only Luis Castillo consistently beat him. And this is especially important because of the characteristics of the park. Guaranteed Rate Field (that name still makes me cringe) is a bit of an unusual bird, a homer-friendly park that tends to be neutral overall. For a pitcher with an elite ability to avoid batters crushing his pitches, this provides an opportunity to squeeze out a little more value. In other words, while Guaranteed Rate is a neutral park for everyone, Giolito’s homer-avoiding tendency makes it a de facto pitchers’ park for him. This one of the reasons ZiPS liked Dallas Keuchel’s chances at a bounceback season in 2020 and continues to think he’ll be very productive for the Sox. Were this a neutral park, Giolito would lose just enough in his projection to drop him to third in the league in WAR.
Lucas Giolito Left Some Strikeouts on the Table
At a 33.7% strikeout rate, Giolito certainly wasn’t struggling to punch out batters. But from the across-the-board improvement in his contact numbers in 2020, ZiPS thinks that he should have seen a larger bump in his strikeout rate from 2019’s 32.3% rate.
As part of its model for calculating baseline expectations, ZiPS has a measurement that I’ve dubbed zSO. (The Z stands for ZiPS, as you may have guessed.) Using contact data, velocity numbers, and the like, ZiPS makes an estimate from how many strikeouts a pitcher “should” have ended up with. It’s not a number I pulled out of my hat but one used as part of the model because it has more predictive value than actual strikeouts. Going back to 2002, if all you knew about a pitcher was his strikeout rate and his zSO rate, you’d have predicted the following year’s strikeout rate most accurately with a mix that was 82% zSO and 18% actual.
| Pitcher | Year | Actual K% | zSO% | Difference | Following Season |
|---|---|---|---|---|---|
| Francisco Liriano | 2011 | 19.0% | 24.4% | 5.4% | 24.1% |
| Mike Pelfrey | 2016 | 10.4% | 15.6% | 5.2% | 14.5% |
| Martín Pérez | 2013 | 15.9% | 21.1% | 5.2% | 16.9% |
| Martín Pérez | 2017 | 13.1% | 18.2% | 5.1% | 18.3% |
| Luis Castillo | 2018 | 23.3% | 28.3% | 5.0% | 28.9% |
| Jeff Fassero | 2004 | 11.8% | 16.8% | 5.0% | 15.6% |
| CC Sabathia | 2008 | 24.5% | 29.5% | 5.0% | 21.0% |
| Jeremy Hellickson | 2011 | 15.1% | 20.0% | 4.9% | 16.7% |
| Jason Vargas | 2017 | 17.7% | 22.6% | 4.9% | 20.8% |
| Craig Stammen | 2010 | 15.1% | 20.0% | 4.9% | 31.6% |
| John Smoltz | 2007 | 23.1% | 27.7% | 4.6% | 30.8% |
| Kelvim Escobar | 2006 | 18.6% | 23.1% | 4.5% | 19.7% |
| Kevin Correia | 2008 | 12.8% | 17.2% | 4.4% | 17.1% |
| Jon Lieber | 2004 | 13.6% | 17.8% | 4.2% | 16.3% |
| Kyle Gibson | 2017 | 17.5% | 21.6% | 4.1% | 21.7% |
| Pitcher | Year | Actual K% | zSO% | Difference | Following Season |
|---|---|---|---|---|---|
| JA Happ | 2018 | 26.3% | 19.7% | -6.6% | 20.6% |
| Stephen Strasburg | 2016 | 30.6% | 24.0% | -6.6% | 29.1% |
| Erik Bedard | 2007 | 30.2% | 23.6% | -6.6% | 20.7% |
| Lance Lynn | 2013 | 23.1% | 16.7% | -6.4% | 20.9% |
| Zack Greinke | 2011 | 28.1% | 22.5% | -5.6% | 23.0% |
| Tanner Roark | 2019 | 21.9% | 16.3% | -5.6% | 18.6% |
| Mike Fiers | 2012 | 25.0% | 19.6% | -5.4% | 14.6% |
| Mike Mussina | 2003 | 22.8% | 17.5% | -5.3% | 18.9% |
| José Quintana | 2017 | 26.2% | 20.9% | -5.3% | 21.4% |
| Tim Lincecum | 2009 | 28.8% | 23.7% | -5.1% | 25.8% |
| Eduardo Rodriguez | 2018 | 26.4% | 21.5% | -4.9% | 24.8% |
| Gerrit Cole | 2019 | 39.9% | 35.0% | -4.9% | 32.6% |
| Rick Porcello | 2018 | 23.5% | 18.7% | -4.8% | 18.6% |
| Yovani Gallardo | 2012 | 23.7% | 19.0% | -4.7% | 18.6% |
| Jon Lester | 2019 | 21.6% | 16.9% | -4.7% | 15.8% |
Looking at the top 15, while zSO is far from infallible — all models are wrong, but some are useful — it had a solid record at identifying the strikeout outliers correctly. So what about the 2020 season? There’s a lower minimum batters faced here (200 batters faced) because of the short season, so you’ll see some larger-than-typical variations between actual strikeout rate and zSO.
| Pitcher | Actual K Rate | zSO Rate | Difference |
|---|---|---|---|
| Ryan Yarbrough | 18.8% | 25.6% | 6.8% |
| Dylan Cease | 17.3% | 23.8% | 6.5% |
| Tyler Anderson | 15.8% | 22.1% | 6.4% |
| Julio Urias | 20.1% | 24.8% | 4.7% |
| Alex Cobb | 16.8% | 21.4% | 4.6% |
| Alex Young | 19.1% | 23.0% | 3.9% |
| David Peterson | 19.5% | 23.4% | 3.9% |
| German Marquez | 21.2% | 24.9% | 3.6% |
| Anibal Sanchez | 17.6% | 21.1% | 3.5% |
| Tanner Roark | 18.6% | 21.9% | 3.3% |
| Zack Wheeler | 18.4% | 21.6% | 3.2% |
| Jesus Luzardo | 23.8% | 27.0% | 3.2% |
| Brett Anderson | 15.8% | 19.0% | 3.2% |
| Antonio Senzatela | 13.5% | 16.5% | 3.0% |
| Randy Dobnak | 13.5% | 16.5% | 3.0% |
| Pitcher | Actual K Rate | zSO Rate | Difference |
|---|---|---|---|
| Trevor Bauer | 36.0% | 26.1% | -9.9% |
| Shane Bieber | 41.1% | 31.9% | -9.2% |
| Tyler Glasnow | 38.2% | 30.4% | -7.8% |
| Cristian Javier | 25.2% | 17.5% | -7.7% |
| Rick Porcello | 20.7% | 13.1% | -7.6% |
| Zach Eflin | 28.6% | 21.8% | -6.8% |
| Corbin Burnes | 36.7% | 30.4% | -6.2% |
| Marco Gonzales | 23.1% | 17.6% | -5.5% |
| Taijuan Walker | 22.2% | 17.2% | -5.1% |
| Hyun-Jin Ryu | 26.2% | 21.3% | -4.8% |
| Aaron Nola | 33.2% | 28.4% | -4.8% |
| Kevin Gausman | 32.2% | 27.8% | -4.4% |
| Framber Valdez | 26.4% | 22.3% | -4.1% |
| Johnny Cueto | 20.2% | 16.1% | -4.1% |
| Sonny Gray | 30.6% | 26.5% | -4.1% |
No, Giolito doesn’t make the top 15 of underachievers, but he’s close. Compared to his 33.7% strikeout rate, ZiPS thought he “should have” been at 35.7. And that’s unusual, as leaders in anything in baseball are more likely to have overachieved than underachieved. Of the top 20 strikeout pitchers in 2020, ZiPS thinks that only four pitchers actually underachieved: deGrom (0.2%), Tyler Mahle (0.3%), Castillo (0.6%), and Giolito (2.0%).
In summation, ZiPS sees Giolito as a nearly perfect storm of awesomeness and one of the top Cy Young contenders in the American League. With Cleveland reeling, the White Sox have an excellent shot at taking the division and going deep into the playoffs. If the White Sox raise a world championship banner in 2021, the right arm of Giolito will likely be responsible for a great deal of the hoisting.
Dan Szymborski is a senior writer for FanGraphs and the developer of the ZiPS projection system. He was a writer for ESPN.com from 2010-2018, a regular guest on a number of radio shows and podcasts, and a voting BBWAA member. He also maintains a terrible Twitter account at @DSzymborski.
I would be fascinated to see some of the volatility in these top pitchers’ projections. Like, for established stars like Scherzer, deGrom, Cole, etc. who don’t have that “extra gear,” are their 10th and 90th percentile projections closer together than up-and-comers? For example, if Giolito has a 7% chance of an ERA of 5.00+, what are the chances of deGrom putting up a 5.00+ ERA?
4% for deGrom, 5% for Cole.
0. My final answer.
Excellent article
It really is
Really loved the zSO! Would be curious to see which relievers rank at the top/bottom of that scale (I think about someone like Graterol, who “seems” like he should K more batters, etc.)
True, but I recall hearing that Graterol has poor extension and thus his perceived velo is quite low compared to his measured velo. His fastball also has plenty of two-seam movement, which are generally worse than 4-seamers at getting whiffs.
Curious, when using zSO to predict the following year’s K%, do you regress the zSO metric as a whole, or do you regress each of the individual components (velocity, contact-rate, etc.)? Which do you think would work better?
My only concern with Gio, is how great so many pitchers looked in the centrals last year. 3 of top 4 sw strk in al central, and 11 of 20 in one of the centrals. I think we have to heavily regress what they did last year, but who knows.
the strikeout over/under achievers is an interesting topic. are there any pitchers who regularly beat their projections or vice versa?
The other part, about HR rate in the minors being predictive (or at least used by ZiPS), is also interesting; during his time in the high minors (AA & AAA), Giolito posted a 0.76 HR/9, while his CHW teammate Cease, also seen near the top of the K% underachievers, posted a 0.52 HR/9 in the high minors. Cease’s MLB HR/9 was 1.85 both 2019 and 2020, bottom-tier among starters. Is the minor league track record why ZiPS is the most bullish on Cease’s ability to suppress dingers for 2021 among the projection systems at FG?
I’m a believer in ground ball rate over HR/FB. If a guy gives up a ton of fly balls, I just don’t believe he can sustain a better than average HR/FB rate over a career.
Are there examples that did this? Flyball-heavy pitchers who didn’t give up a ton of homers?
Edit: Since 2000, only Matsuzaka, Verlander, Scherzer, and Jason Schmidt kept HR/9 under 1.00 while putting up a groundball rate under 38%. So a ton of strikeouts helps, but even still, it’s almost impossible to suppress home runs if you are giving up a ton of flyballs.
Some pitchers like that have excelled regardless. Granted, Fergie Jenkins and Robin Roberts didn’t have to contend in this HR-happy environment, but it is, or at least was, possible.
Giolito is such an interesting study in how a player is handled by an organization. I am certain that everyone knows that he was considered a generational talent when he was originally drafted by the Nationals. He started off great but at some moment the organization decided that what he brought with him wasn’t what they wanted and attempted to mold him into something he wasn’t. This is very different than just blowing a pick on a Tyler Kolek or being unlucky like with a Danny Hultzen and doesn’t compare to attempting to change something that hasn’t worked after giving it a few years because it worked for his first couple of years in the Nationals system. Why would the Nats do that? But the story has been told that they did. To quote from the BP Annual in 2018, there were “reported mechanical tweaks” and also ” He’s not going to be the ace he looked like he might be years ago, but if he were the Nationals would not have traded him in the first place”. Has the circle been completed and he is going to be the ace that so many predicted? He is at the threshold and I, for one, will be following closely. There is another reason i will be watching and that is because Tanner Houck was treated in a similar manner by the Red Sox. It appears, from what I have been able to gather, that the Red Sox, despite drafting him in the 1st round, did not give him any time at all to show whether his stuff was good enough but immediately altered his delivery. Why? I guess the data from whatever computer inputs that they had spit out information that suggested he would be better if he changed a few things. It certainly looks like whatever was done, to be kind, wasn’t successful and he disappeared into post-prospect purgatory, a dreaded place from where few escape. The stories from the alternate site this summer indicated that he had returned to his college form and his 3 starts were astonishingly good. Was this just a flash in the pan? It very well could be but it is very surprising to see every, and I mean every, form of baseball media treat these as if they never happened. Here at Fangraphs he wasn’t bumped up nor were his exploits mentioned. He still resides at #10 on the Red Sox 2021 prospect lists at MLB and BP behind guys that I will bet never pitch 3 games like he did or hit 30 HR’s in the major leagues. I watched every pitch and major league hitters looked helpless flailing at his pitches but post-prospect purgatory is what it is. There is a thread that ties this together. In my experience coaching golfers all the way to high levels of success at the professional level I can say that it is very tricky to change great natural talent and that the coach’s job is to fine tune that gift, lead the gifted athlete to find his own way, but never lay down a my way or the highway regimen or make d drastic alteration until all appears lost. To find that line of balance between maximizing the athletes talent and interfering with their development is mystical. The Indians seem to be able to do it with pitchers while the Orioles record is abysmal. Now let me watch how these two young men perform.
Fun baseball read, love the format for like a series on breakout candidates big and small. I mostly wrote Giolito off as a prospect because of the TJ, prejudiced in that way I guess against young RHP especially. Now yeah, I think he’s the kind of pitcher who might put up 40+ WAR in his career and even see him as a trendsetter with that revamped arm action likely saving his and at least a few other careers (Robbie Ray seems to have made similar changes)
For all the noise around the White Sox’s young talent, he and Anderson seem like the most stable and best players on that team?
Moncada has more career fWAR than Anderson and is 3 years younger. He is clearly the biggest star after Golito.
Maybe he’s referring to Moncada’s pattern (so far) of alternating very good years with mediocre ones when he calls Anderson “most stable.” You can argue whether two consecutive years is all that stable.
Or how stable the high-K/low walk, .400 BABIP formula is
Yeah, Anderson might be the least stable position player in all of baseball. He was on pace for something like a 5-win season last year, but virtually his whole profile is BABIP. When your whole profile revolves around the most unstable thing in baseball you’re going to be in for a wild ride.
Moncada spent all of 2020 battling the after-effects of COVID. It’s actually pretty remarkable that he still managed to be valuable, because his speed and power were completely sapped.
Very informative article. Perhaps you can explain why zips is so down on teoscar hernandez with the longest home distance in 2020, 42 home runs in his last 650 pa . above average steals
Digital Love? Sounds like the name of a late, late night movie on Showtime.
Agree on the name of Sox park. Some have referred to it as Second Rate Field.
Daft Punk song.
OT, but I can’t help noticing how bad the Giolito-Eaton trade turned out for Washington. It’s one thing to trade future performance for present, but by the time Eaton had a full season and the Nats won it all, Giolito was outperforming Eaton in the majors.
And the White Sox got Dane Dunning in that trade who was flipped for Lance Lynn. BUT…flags fly forever and Eaton did a good job at the top of the Nat lineup in 2019.
If you won the championship, then you don’t mess with the timeline.