Tony Bucciferro and the 23-Year-Old Strike-Thrower Question

Quick: Who led minor league pitchers in FIP this year?* You might guess one of the top pitching prospects in baseball, such as Archie Bradley, Taijuan Walker, or Robert Stephenson. Or you might rack your brain thinking about 2013 breakout A-ball pitchers like C.J. Edwards, Tyler Glasnow, or Edwin Escobar. In either case, you’d be wrong, because the answer is little-known White Sox pitching prospect Tony Bucciferro.

*Minimum 80 IP. Giants relief prospect Derek Law had a lower FIP in 77 2/3, but I wanted to isolate starters.

Bucciferro’s 2013 numbers are certainly something to behold. Across three starts in the Rookie-Advanced Appalachian League and 13 outings (12 starts) in the Low-A South Atlantic League, he amassed 96 strikeouts, six walks, and a mere three homers allowed in 90 2/3 innings, good for a 1.74 FIP (2.48 ERA). Those numbers seem impossible to ignore, and yet Bucciferro registers barely a blip on the prospect radar. In this post, I will examine why he’s been successful, why he’s been ignored, and take a systematic look at how pitchers with this sort of statistical profile fare.

Anatomy of a 1.74 FIP

Here’s a look at Bucciferro in his final start of the 2013 regular season, on September 2:

What can we discern from this video? There are a number of aspects that jump out that help explain how Bucciferro has been so dominant against low-minors bats.

Pitching Arsenal

The first aspect I want to hone in on is simply Bucciferro’s arsenal of pitches. There’s nothing especially remarkable—he throws a four-seamer at 89-91 mph, a two-seamer at 88-90, a slider at 81-84, and a changeup at 78-82—but all four offerings are credible, functional pitches. He’s got enough velocity to not embarrass himself, the two-seamer has solid run and sink, the slider has solid velocity and break, and the changeup has velocity separation, sink, and fade.

There’s nothing in there that a scout would be tempted to throw a plus grade on with any consistency, but none of the four pitches grade out significantly below average, either. That consistency of pitch quality is a very advanced trait for an A-ball hurler. The fact that Bucciferro has two solid offspeed offerings at somewhat different velocities, in particular, was a big factor in his being able to strike out an excellent 23.6% of South Atlantic League batters.

Pitch and Location Sequencing

Perhaps the most striking facet of Bucciferro’s game is his extremely varied pitch sequencing. Many pitchers–even up through the big league level–display clear sequencing limitations. One common one is throwing fastball/breaking ball to same-handed hitters and fastball/change to opposite-side hitters; a second one is being unable to come inside on opposite-side hitters. Bucciferro, however, takes advantage of the solid-across-the-board quality of his arsenal to pattern his pitches in a way that almost seems random. He’s not afraid to come inside on lefthanders:

The running action on his two-seam fastball allows him to start the pitch off the inside corner to lefties and bring it back over for called strikes (or, if the batter chooses to swing, jam shots). But it’s not just the fastball that Bucciferro can throw inside to southpaws:

All three of these are offspeed pitches on the inner half–something you don’t see pitchers throw much to opposite-handed batters, let alone at this level of the minor leagues. But Bucciferro has confidence in all three of his pitches in all locations to batters from both sides of the plate, which leads to him being extremely unpredictable in his sequencing. Lefthanded batters can’t just sit on fastballs and changeups away, because while he’s more than capable of utilizing those, he also might drop a back-foot slider or a front-door two-seam or changeup. This inability for batters to know what’s coming allows all of Bucciferro’s pitches to play up.

Furthermore, Bucciferro’s ability to claim the inner half to lefties has a significant bearing on his microscopic walk rate. He walked just one of the 152 lefthanded batters he faced in 2013, and that’s because he could provide location variance inside the zone as opposed to merely outside it. After all, if one is a righthanded pitcher who can only work the outer half to lefthanded hitters, one has, essentially, three courses of action with regard to location:

1.) Attempt to pound the outside corner.
2.) Attempt to work the outside corner, but vary height (fastballs up, offspeeds down) to get the hitter off balance.
3.) Start by attempting the work the outside corner, then go off the outside corner for variance.

The first strategy is utterly predictable and could well be problematic multiple times through the order, as batters could just sit on one location. The third is the least predictable and thus wouldn’t lead to a barrage of hard contact, but it also could lead to lots of walks–a sort of “damage control” strategy. The second splits the difference. None are optimal, really. But in claiming the inner half of the plate to opposite-side batters and using all of his pitches to them, Bucciferro can avoid walks while still remaining very unpredictable. It’s a huge part of his success and speaks to his polish. He can also do things like front-door a slider to a righthander:

The slider to righthanders is probably the closest thing Bucciferro has to a true bat-missing weapon. He struck out 28.83% of righties this year, as opposed to 21.05% of lefties, thanks to pitches like this:

That’s pretty good tilt, and of course, Bucciferro’s zone-heavy approach makes borderline pitches like this one that much tougher to lay off–he’s not a pitcher batters can wait out, lest they want to be in quick 0-2 holes. And as with the slider, Bucciferro also has plenty of ability to throw the most classic pitch in baseball, the fastball away:

Add it all up, and you have a pitcher who works both sides of the zone with all of his pitches to hitters from both sides. That allows Bucciferro to keep hitters guessing without falling behind, and it also makes his stuff play up and draw more empty swings than one might expect.

Delivery

I’m going to talk a lot more about mechanics in my next post, so I don’t want to belabor the point here, but Bucciferro’s mechanics are a big part of his success. It’s difficult to consistently find the zone if one cannot consistently find a release point, and Bucciferro has no problems with that. The word that jumps to mind when watching his delivery is “compact.” There’s very little wasted motion, and he has excellent tempo–both in the motion itself, and between pitches. He uses his lower half well, with the only oddity of his motion being a fairly pronounced head jerk–one that doesn’t seem to throw him off much. As such, he’s able to hit his spots, and can come inside to lefties without worrying about the ball running back out over the plate.

Groundball Ability

The other component of FIP is, of course, home runs, and Bucciferro was stellar in homer prevention as well, allowing just three big flies all year. Of course, homer numbers can fluctuate a lot–if five more players had happened to get a hold of Bucciferro pitches this year, his homer rate would’ve been just average. How legitimate is his ability to keep the ball in the park?

According to StatCorner, Bucciferro had a 50.9% groundball rate* this season, supporting the idea that he’s above-average at keeping the ball out of the air (and thus, inside the fences). With the exception of the four-seam fastball (which he doesn’t use as much as the two-seamer), all of his pitches have sinking action on them, and if he does have one tendency with regard to location, it’s to favor the lower half of the zone over the upper half. A look through the video shows his ability to get a lot of weak groundball contact on all three of his pitches, most notably the two-seamer and slider.

*StatCorner doesn’t have Appalachian League data, so this rate only applies to Bucciferro’s 72 Low-A frames.

While the 50.9% groundball rate may not be as stellar as the 0.3 HR/9 that Bucciferro posted, it removes the only other statistical avenue to futility. Bucciferro walks almost nobody, strikes out a well-above-average amount of batters, and limits damage on contact. So how can he not be a prospect?

Anatomy of Ignoring a 1.74 FIP

Here are the three numbers that explain a lot of the lack of Bucciferro awareness and/or optimism in the scouting/prospecting community: 89.5, 14, and 23.

The first is (roughly) Bucciferro’s average fastball velocity. It’s not a death sentence to average 89.5 mph on one’s heater, but it’s not optimal either–of the 145 pitchers who threw 100 or more innings in the big leagues this year, only 32 (22%) averaged 89.5 or less, and just 17 of those were non-knuckleballing righthanders. Sure, Bucciferro’s ball moves, which helps make up for some of that, but Jake Westbrook throws an 89.5 mph moving heater, too, and he struck out all of 8.4% of batters he faced in 2013. Further, Bucciferro doesn’t make up for his lack of velocity with some sort of dynamite offspeed pitch, nor is he particularly deceptive. With a reasonably thick build, he doesn’t project to add much more velocity later on, either. There’s no bigtime plus stuff that one walks out of the ballpark impressed with.

The second number, 14, is the round Bucciferro was drafted in in 2012, meaning he obviously has to work extra hard to get himself noticed, and the final number, 23, was Bucciferro’s age throughout the 2013 season. Drafted as a 22-year-old college senior, he didn’t even make the Low-A Kannapolis squad out of spring training, being finally dispatched as an injury replacement in early June. He made two bad starts and allowed a run in a relief appearance, getting sent back to the Appy as a 23-year-old. Even then, he didn’t stand out–he made one bad start and one good one, and then was promoted back to Kannapolis for a spot start on July 7 in which he was hit around and demoted again. Just looking at his game log, one can see Bucciferro wasn’t really producing much of anything until he was sent back to Bristol on July 12. He’s 23 years old, hasn’t left A-ball, and has a grand total of seven weeks of good professional pitching.

What a seven weeks they were, though: hitters hitting .229/.245/.338 with a 27.64% K% and 1.51% BB%, giving Bucciferro a 1.14 ERA in ten starts (one with Bristol and nine with Kannapolis). And that leads me to a question: Is it safe to ignore a pitcher like this? We know, intuitively, that these statistics are fantastic. We also know, intuitively, that 23-year-old pitchers who are stuck in Low-A, even with good numbers, generally aren’t thought to project well.

We can vacillate between valuing one extreme or the other, but instead, why don’t we look at some data?

23-Year-Old Low-A Control Pitchers: An Examination

In order to get a sense of how Bucciferro and his ilk project, I decided to isolate a sample of comparable pitchers. Here are the characteristics of the sample I selected:

• Was in Low-A between 1998 and 2012
• Was 23 while in Low-A
• Walked 5.5% or fewer batters in Low-A, age-23 season
• Threw at least 50 innings in Low-A in age-23 season
• Over 50% of appearances in Low-A, age-23 season were starts

In 2013, the average walk rate for Low-A pitchers who threw at least 50 innings was 8.5%, with a standard devation of 3%, so this sample consists of (roughly) pitchers who were a standard deviation above the mean at throwing strikes–like Bucciferro, they derived a large portion of their value from control. I wanted to hone in on pitchers who were mostly starters, since relief skillsets are a rather different matter, and I wanted to stay somewhat in the modern era (if one goes back to 1980s leaderboards, I discovered, low walk rates are fewer and farther between, and obviously strikeout rates were lower). I found 90 pitchers that fit these constraints.

How Did They Fare?

So, how did the pitchers in this sample turn out? Of the 90, 23 (25.56%) threw a major league pitch. They are:

Mark Malaska
A.J. Griffin
Edwar Cabrera
Joe Martinez
Brian Duensing
Graham Taylor
Thad Weber
Josh Geer
Phil Irwin
Jairo Asencio
Brad Lincoln
Scott Cassidy
Cory Luebke
Dirk Hayhurst
Scott Dohmann
Cory Wade
Bobby LaFromboise
Greg Burke
Burke Badenhop
Zach Clark
Tony Cogan
Ryan Rupe
Mike Burns

Griffin and Luebke are the two that stand out as the upside here, having established themselves as quality big-league starters. Rupe, Burke, Badenhop, Cassidy, Wade, Martinez, and Lincoln all had some longevity in varying MLB roles, and there’s still hope for Cabrera, Irwin, and LaFromboise to establish themselves as well. So, this group of 23 is split almost evenly between players who just made it up briefly and those who were able to sustain some success and get above replacement level.

Of the remaining 67 pitchers who did not make it to the big leagues, 35 at least reached Double-A, meaning that 58 (64.44%) of the 90 reached the level. Five of the 35 who reached Double-A but not the majors (Jared Rogers, Josh Romanski, Casey Lawrence, Eliecer Navarro, and Wes Musick) remain active, while three of the 32 that have not escaped A-ball (Jesse Hernandez, Matt Tomshaw, and Kyle Hald) are also still pitching, so there is a chance for some slight increases in the success numbers down the line, though none of these players are thought of as significant prospects at the moment.

So, our breakdown likely looks something like this (these are all very rough numbers):

13%–above replacement level
17%–near replacement level (figuring a couple of the top pitchers who did not reach the majors were likely around this threshold)
30%–organizational player (while over 30% made it to the upper minors, a few of these players only made it to Double-A briefly)
40%–injury casualties, regressers, and/or non-prospects

So 23-year-old Low-A control pitchers aren’t a great bet, but there are success stories, and there’s a pretty reasonable chance they at least make it to the upper minors and pitch another few seasons. That sounds reasonable enough.

All 23-year-old Low-A control pitchers aren’t created equal, though. A large part of Bucciferro’s statistical appeal, for example, is that he combines the miniscule walk rate (which, mind you, is lower than that of every pitcher in this sample) with a good helping of strikeouts. Our ears perk up when seeing his numbers; they don’t perk up when looking at a 23-year-old Low-A with a sub-5.5% walk rate and, say, a 12% K-rate.

And that leads me to my next area of examination. I’ve identified that 23 of the 90 pitchers made it to the big leagues, 35 made it to Double-A but not the big leagues, and 32 did not make it to Double-A. Is there anything in the statistics that can help separate the prospects from the pretenders?

Comparing Groups

For each of the 90 players in the sample, I recorded twelve pieces of information:

1.) BB% in age-23 Low-A season
2.) K% in age-23 Low-A season
3.) K/BB in age-23 Low-A season
4.) BB% in first extended exposure to High-A*
5.) K% in first extended exposure to High-A
6.) K/BB in first extended exposure to High-A
7.) BB% in first extended exposure to AA
8.) K% in first extended exposure to AA
9.) K/BB in first extended exposure to AA
10.) Highest level reached in age-23 season
11.) Level assigned to at opening of age-24 season
12.) Highest level reached in age-24 season

*Some pitchers, obviously, were promoted to High-A in the middle of their age-23 seasons and then started the next year in Double-A, while others finished their age-23 season in Low-A and then opened the next year in High-A. Since these are all older pitchers who need to adjust to levels extremely quickly to retain relevance, and because there’s so many different promotion paths between the pitchers in the sample, I decided to use the first extended sample of High-A data (and likewise for Double-A) rather than just using “next year High-A data” or “same year High-A data.” In a few cases where a pitcher had a just a handful of innings at a level one year and then another handful the next year, I combined them. Overall, it’s not a perfect method, but it allows for the broadest comparisons.

With this data in tow, we can begin to make some comparisons between our three groups: the players that reached MLB, those that reached AA but not MLB, and those that did not reach AA. Here are the averages of the age-23 Low-A numbers across groups:

MLB: 4.1% BB%, 20.7% K%, 5.32 K/BB
AA: 4.41% BB%, 18.71% K%, 4.46 K/BB
A: 4.01% BB%, 17.38% K%, 4.48 K/BB

What jumps out here is that the strikeout rate has a trend but the walk rate does not. In a lot of ways, that makes sense–the walk rates of the pitchers in this sample are all tightly compressed (2% to 5.5%), and the underlying skill difference between, say, a 4% walk rate and a 5% walk rate seems a lot less significant than the difference between a 16% K-rate and a 20% K-rate, even though they have the same impact on a pitcher’s K/BB. Part of that is just the fact that the walk number is smaller, but part of it is also the fact that, by definition, the pitchers in this sample throw a lot of strikes. It stands to reason that what separates those that succeed from those that don’t isn’t having plus-plus command as opposed to plus, but rather the other elements of the pitcher’s game, one of the most important of which is the ability to miss bats. No pitcher who struck out under 17% of batters as a 23-year-old in Low-A became an above-replacement level contributor–the lowest was Greg Burke at 17.4% (Jairo Asencio, Josh Geer, and Zach Clark all made it up with lower rates but were not above replacement).  In contrast, ten of the top twenty strikeout rate pitchers reached the majors, and only two (Scott Mitchinson and Ryan Copeland) never pitched in the upper minors.

While walk rate did not translate as a whole, it is interesting that four of the nine pitchers with sub-3% walk rates (Burke, A.J. Griffin, Brad Lincoln, and Cory Wade) made a big-league impact, with a fifth (Graham Taylor) also making it to the majors, and only one (Jesse Darcy) never pitching in the upper minors. However, none of the fifteen pitchers with walk rates between 3.0% and 3.7% made a big-league appearance, though, with only three even reaching Double-A.

How about the High-A numbers?

MLB: 6.21% BB%, 19.21% K%, 3.35 K/BB
AA: 5.79% BB%, 17.75% K%, 3.48 K/BB*
A: 6.33% BB%, 15.37% K%, 2.57 K/BB*

*Nine pitchers did not have High-A data (five skipped or almost-skipped the level, whereas four never made it beyond Low-A) and were thus excluded from the calculations.

What’s interesting here is that, whereas the big difference in the Low-A data is between the MLB and AA groups, here the difference is between the AA and A groups, with the AA group actually grading out better than the MLB group in K/BB. Whereas the AA group had the highest walk rate in their age-23 Low-A seasons, they actually had the lowest walk rate when transitioning to High-A.

Though the K/BB gaps change considerably here, what is clear is that the walk rates all went up 1-2% and the strikeout rates all went down 1-2%. As such, while the differences between the K/BB ratios moved around a lot, the clear trend in strikeout rate remains. The top three pitchers in High-A strikeout rate (Cabrera, Griffin, and Cassidy) all made it to the majors, and Malaska, Asencio, and Martinez were also in the top ten.

Of course, the “watershed” moment that tends to separate the prospects from the non-prospects is Double-A. We can compare how the MLB group fared when they reached Double-A compared to those that reached Double-A but not the majors. It is, indeed, a stark contrast:

MLB: 6.77% BB%, 19.31% K%, 3.22 K/BB
Non-MLB: 7.76% BB%, 15.89% K%, 2.26 K/BB

Here we have an even bigger K% gap than earlier, as well as a full percent of walk rate. Fourteen of the top twenty K/BB pitchers in AA reached the majors, as opposed to nine of the other 38. Twelve of the twenty-best AA walk rates made the grade, as did twelve of the top twenty strikeout rates. Clearly, this is a much more definitive separation, though there still are players that succeed in the bigs with mediocre AA numbers (Duensing: 8.5% BB% and 14.2% K% and Luebke: 8.4% BB% and 17.9% K%) and those that never make it despite stellar AA numbers (Chris Cody: 4.3% BB and 20.8% K% and Duncan McAdoo: 5.8% BB% and 20.5% K%).

Finally, I wanted to examine if there was anything predictive about the player’s near-term level progression. Do players with these characteristics who end their age-23 season in High-A fare better than those who end that season in Low-A? What about starting their age-24 season in Double-A vs. High-A, or ending it in Double-A vs. High-A?

Here’s the breakdown of final age-23 season levels:

45 (50%) ended in Low-A; nine of these (20%) reached the majors; 25 (55.56%) at least reached AA
38 (42.22%) ended in High-A; eleven of these (28.94%) reached the majors; 26 (68.42%) at least reached AA
7 (7.78%) ended in AA; three of these reached the majors

Clearly, there’s an upward trend here. How about where these players broke camp in their age-24 seasons?

9 (10%) started in Low-A; one of these reached the majors; three at least reached AA
55 (61.11%) started in High-A; 14 of these (25.45%) reached the majors; 37 (67.27%) at least reached AA
16 (17.78%) started in AA; eight of these (50%) reached the majors
2 (2.22%) were injured and rehabbing in their age-24 seasons; both reached AA but not MLB
8 (8.89%) were released or retired before their age-24 seasons

Again, a fairly clear trend. How about where they ended their age-24 seasons?

1 (1.11%) ended in Low-A; he did not reach AA or MLB
43 (47.78%) ended in High-A; five of these (11.63%) reached the majors; 19 (44.19%) at least reached AA
25 (27.78%) ended in AA; ten of these (40%) reached the majors
8 (8.89%) ended in AA; five of these (62.5%) reached the majors
3 (3.33%) ended in MLB
As previously mentioned, 10 (11.11%) were released, retired, or injured in their age-24 seasons

These trends are all quite pronounced and clear, and they all indicate that how a player advances in his age-24 season is vital to his potential. Players like this need to catch up to the age curve as quickly as possible to get noticed and taken seriously. Sure, it’s possible to finish one’s age-23 season in Low-A and spend all of one’s age-24 season in High-A and become an above-replacement level pitcher–Joe Martinez pulled it off, albeit barely (0.2 WAR in 55 1/3 innings)–but the odds of that aren’t good. If Bucciferro or a similar pitcher is going to get to the big leagues and stay for any length of time, he’s best served by getting to Double-A in his age-24 season, where the odds of doing so dramatically increase.

Of course, it’s fair to question which side comes first–does being promoted faster legitimize an older control pitcher as a prospect, or do the older control pitchers who are legitimate simply get recognized as such by their organizations and thus get prioritized more in promotions? The answer is likely a mix of the two, since most of these pitchers see time (in either their age-23 or age-24 seasons) in High-A and can certainly increase or decrease their stock in that interval.

Tying It Together

The above study definitely incorporates some fairly messy math, but there are definitely some general conclusions it points toward.

First, 23-year-old Low-A pitchers who thrive on attacking the zone aren’t great bets to succeed, but they’re not a bad place to look for sleepers. Few of these guys showed up on prospect lists, and yet a quarter made the big leagues, and an eighth made a reasonable big league impact. It’s certainly possible for a player with this profile to overcome the age-relative-to-level deficit and have a career. However, their excellent control does not necessarily hold up forever–even for those who reached Double-A, the average walk rate of pitchers in this group nearly doubled in moving from Low-A to AA–and comes far from guaranteeing upper-minors success, let alone time on baseball’s biggest stage.

The Low-A numbers themselves don’t correlate particularly well with anything, except that higher strikeout rates are clearly better. This suggests that much of what separates the successes from the failures is either a) latent variables that don’t really show up in K/BB stats (groundball ability and stuff being the two most obvious contenders, though one would think stuff correlates with K%) or b) further refinement and development after the age-23 season.

Finally, the promotion timetable for players in the immediate aftermath of these dominant age-23 Low-A control performances seems to correlate strongly with where they end up. This makes a lot of sense–if a player starts out behind the age curve at 23 and isn’t moving fast before he’s 25, he’s probably not going much of anywhere regardless of his raw production.

Bucciferro will be an interesting test case. His walk rate is lower than everyone else’s in the sample, his K/BB is easily the best, and his strikeout rate is well above average, yet not only did he end his age-23 season in Low-A, he wasn’t even definitively assigned there until mid-July of that season. Even when compared to others who share some of his statistical positives and scouting drawbacks, he remains an extreme case. It will be fascinating to see how seriously his organization takes him, where they assign him throughout the 2014 season, and how he responds to those assignments.