Revisiting Baseball America’s Top-10 Prospects from 2006

Flashback to February of 2006. Lindsay Lohan has never been to jail, Steve Irwin’s alive and well, and we still have three years of Bieber-free living to look forward to. More importantly, though, Baseball America has just put out its annual top 100 prospect list.

Now that a decade’s passed, and we know how each of these players turned out, let’s look back at some of these players’ prospect years. Perhaps there are lessons to be learned — statistically or otherwise — from their case studies that might be useful in evaluating today’s prospects. Keep in mind that these players are all just anecdotes. They represent merely a few data points among thousands, and therefore shouldn’t be used to draw any sweeping conclusions. But still: real-world examples are always fun, and names and faces are a great way to bring macro-level trends to life.

Below each name, you’ll see three WAR figures. The first is that player’s historical KATOH forecast — that is, what KATOH would have projected given the relevant prospect’s minor-league numbers. The second was calculated using the formula derived by Jeff Zimmerman in this year’s Hardball Times Annual applied to each player’s 2006 BA ranking. The third is that player’s actual (positive) WAR total throughout the period of time forecasted by KATOH. (So for players 22 and younger, that’s thru age-28. For players 23 and over, it’s the first six years.)

Essentially, you have KATOH’s projection, BA’s projection, and what actually happened. I tried to glean some sort of lesson from each of these case studies, even if the lesson wasn’t particularly ground-breaking.

1. Delmon Young, OF (Profile)

BA: 20.3 WAR
Actual: 3.6 WAR

Just like everyone else, KATOH was all in in Delmon. It was kind of hard not to be. Young hit .322/.388/.538 as an 18-year-old in A-Ball, and then slashed .315/.354/.527 between Double-A and Triple-A. And he stole bases, too. He had it all, or at least it looked like he did.

His one flaw was his plate discipline. Both his strikeout and walk numbers left a little to be desired, but dwelling on those numbers feels like nitpicking. His 7% walk rate was certainly acceptable, and his 19% strikeout rate wasn’t terribly concerning (even considering that 19% then might be equivalent to 22% now). When a player is so good at everything else, you can let things like that slide. You kind of have to, or else you’d be down on nearly every prospect.

As you know, Delmon never panned out. The plate discipline was the biggest culprit — although his weight problems certainly didn’t help, either. As of this writing, Young is a 30-year-old who’s been without a team for the better part of the last year. He was a replacement-level player or worse for most of the last decade.

I guess the takeaway is to be wary of hitters who have great tools and minor-league numbers, but also possess an unrefined approach at the plate. Although, I’d be lying if I said I could have seen this coming. The strikeout rate was a warning sign, but the truth is that 19-year-old Delmon Young gave us little warning that we were in for a decade of replacement-level play. More often than not, players as talented as Young make the necessary adjustments. Sometimes they don’t.

2. Justin Upton, SS (Profile)

BA: 14.8 WAR
Actual: 25.9 WAR

Upton was the clear best player in 2005’s amateur draft, but didn’t debut in the minors until 2006. So at the time of this list, he had zero experience above the high school level. Upton had a very successful year and a half in the minors, debuted with the Diamondbacks in 2007 and was an All-Star-caliber player by 2009. He’s been on of the top dozen or so players in the league ever since.

Consensus number-one overall picks have a habit of becoming really good players, even before they’ve played in the minors. But you already knew that, right? I hope you did.

3. Brandon Wood, SS (Profile)

BA: 12.4 WAR
Actual: 0.0 WAR

Wood’s 2005 season was very asthetically pleasing. As a 20-year-old shortstop, he slashed .321/.383/.672 in High-A with 43 homers… as a shortstop! Sure, some of that had to do with his home ballpark in Rancho Cucamonga, but a lot of it didn’t. Wood also hit loads of dingers at every other stop.

It’s easy to see why Wood failed: he struck out too much. He whiffed 22% of the time in both of his A-Ball stops, and upwards of 25% of the time at the higher levels. He pushed 30% in the big leagues with minimal power. As with Young, Wood should have had enough pluses to outweigh his one minus, but it didn’t work out that way.

High-strikeout prospects are risky. A high strikeout rate shows that a hitter is getting fooled by minor league pitching. There’s still some development that needs to happen with a player like this, and sometimes that development doesn’t happen. Cases like Wood’s illustrate the downside of the high-strikeout profile, and explain why KATOH’s so hard on prospects with strikeout rates well north of 20%.

4. Jeremy Hermida, OF (Profile)

BA: 10.9 WAR
Actual: 3.6 WAR

KATOH was very down on Hermida after 2004. Sure, he was fresh off of a .297/.377/.441 as 20-year-old in High-A, but his 19% strikeout rate looked a little suspicious. Then 2005 came along, and KATOH fell in love. Hermida slashed .293/.457/.518 in Double-A that year. Walks and extra-base hits galore! And steals, too! All while shaving a point off of his strikeout rate.

Unfortunately, Hermida never really performed at the big-league level. He had a decent year in 2007, but otherwise hovered around replacement level. He predictably struck out a good amount in the show, but he also never developed the sort of power most thought he would. His terrible outfield defense also didn’t help matters. He was bad all over.

You know the drill by now. High-power, high-strikeout hitters are volatile. Most won’t fail nearly as badly as Young, Wood and Hermida did, especially when they excel so much in other areas, but it’s always a possibility.

5. Stephen Drew, SS (Profile)

BA: 9.8 WAR
Actual: 11.4 WAR

Drew was drafted 15th overall in 2004, and after a stint in Indy Ball, he split 2005 between High-A and Double-A. He hit extremely well in A-Ball (.389/.486/.738), but struggled in Double-A (.218/.301/.386). The latter performance, along with Drew’s age (22), led to the poor KATOH projection. Stats aside, Drew was viewed as a plus defensive shortstop who could also hit and hit for power.

Drew went on to have a nice, yet unspectacular, career. He got off to a slow start, but ultimately wound up spending a few seasons as an everyday player. He even had an All-Star-caliber season in 2010. He wasn’t a star, but was much better than most others on this list.

Don’t dwell on small-sample stats from a player’s debut year, as they’re prone to understate his true potential. This is especially true of KATOH forecasts, which regress small-sample stats to the mean. This makes it hard to earn a great projection with only 200 or so plate appearances.

6. Francisco Liriano, LHP (Profile)

BA: 9.0 WAR
Actual: 13.5 WAR

Liriano was an excellent minor leaguer and his performance improved at every stop, particularly in the strikeout department. He split 2004 between High-A and Double-A, and split 2005 between Double-A and Triple-A. Although he was consistently young for his levels, he struck out roughly 30% of opposing batters as a starter. Also keep in mind that 30% was even more impressive then than it is now.

Liriano burst onto the scene in 2006, but sat out all of 2007 following Tommy John surgery. He returned to action in 2008, and although he was wildly inconsistent, he showed flashes of brilliance over the next few years. Those flashes were more than enough for him to blow past his KATOH forecast.

Pitching prospects carry a lot of risk. Sometimes they succumb to injury, inconsistency or both. Even so, for pitchers as dominant as Liriano, it only takes a couple of excellent seasons to push the WAR total into double-digits.

7. Chad Billingsley, RHP (Profile)

BA: 8.4 WAR
Actual: 19.5 WAR

Statistically speaking, Billingsley’s minor-league track record wasn’t outstanding. He spent 2005 in Double-A, where he pitched to a 3.51 ERA on the strength of a 27% strikeout rate. He was certainly good, but so were John Stephens, Travis Blackley and Renyel Pinto. Let’s just say there’s a reason you’re unfamiliar with those names.

Billingsley’s stuff, however, was very good. His fastball-curveball combination enabled him to carve up major-league hitters for the better part of the last decade. It feels like Billingsley’s been borderline irrelevant for ages now, but here are his WARs from 2007-2012: 2.1, 4.4, 3.0, 4.2, 2.2, 2.8. Consistently solid.

Don’t take the projections too literally for top-ranked pitchers. Factors like stuff and velocity are very important, but are ignored by KATOH. Consensus elite pitchers are liable to blow past their projections, even when those projections are really good.

8. Justin Verlander, RHP (Profile)

BA: 7.9 WAR
Actual: 29.8 WAR

Verlander dominated in his brief stint in the minors. The 2004 second-overall pick opened his career at the High-A level in 2005, where he rattled off a 1.67 ERA in 13 starts. His next stop was Double-A, where he posted a 0.28 (!) ERA in seven starts. Yet, despite his dominance, KATOH wasn’t completely sold. ERA aside, Verlander’s Double-A performance wasn’t actually all that great. His strikeout rate dipped from 31% to 28% as he moved from High-A to Double-A. But more importantly, he essentially pitched like a reliever. The Tigers were understandably monitoring Verlander’s workload, and limited him to four or five innings per game. In KATOH’s eyes, Verlander spent a chunk of his season as a very good 22-year-old Double-A reliever.

Verlander had an excellent rookie campaign in 2006 and never looked back. He began his career with a 2.8 WAR season in 2006 and hasn’t recorded a lower WAR since. He was also arguably the best pitcher in baseball from 2009 to 2012.

It’s important to consider why a pitcher might be pitching in relief and/or only lasting a few innings. Since KATOH’s a dumb computer, it reasoned that Verlander’s short outings signified a flaw with Verlander — that he was converting to the bullpen or lacked stamina. However, that certainly wasn’t the case.

9. Lastings Milledge, OF (Profile)

BA: 7.5 WAR
Actual: 0.7 WAR

I’m proud to say that KATOH would have nailed this one. In fact, it would have come in a tad high. Like the hitters above, Milledge had some contact issues as a prospect. Like the hitters above, those contact issues manifested themselves in the majors. But unlike the hitters above, he wasn’t excellent in other areas. His numbers were good, but not phenomenal, while his strikeout rates hovered around 20%. Despite a .318/.388/.449 batting line in 2005, Milledge was mostly speed and BABIP with just a touch of power.

Milledge failed hard. He hit for minimal power, never learned to walk and was somehow a terrible defender in the outfield despite his excellent speed. Milledge was one of those near-mythical five-tool players, but he never learned plate discipline or how to take proper routes in the outfield. These factors explain most of his downfall.

Toolsy players with good minor-league numbers are risky bets if they struggle to make contact. Players like this don’t always parlay their tools into on-field success.

10. Matt Cain, RHP (Profile)

BA: 7.2 WAR
Actual: 29.3 WAR

Cain’s 2005 campaign was about as promising a minor-league season as you’ll see from a pitcher. At just 20 years old, he struck out a remarkable 28% of opposing batters faced in a full season of starts at Triple-A. Keep in mind that strikeouts weren’t as common a decade ago.

Cain immediately started producing in 2006, with a 3.8 WAR campaign. Cain never recorded more than five wins in any one season, but he recorded at least 3.4 WAR for seven consecutive years. Injuries have since derailed his career, but he was consistently very good from age-21 to age-28. Both KATOH and BA had lofty expectations for Cain, and he exceeded them.

KATOH rarely projects pitchers for more than a handful of WAR. When it does, there’s a strong chance that pitcher will be really, really good.

Chris works in economic development by day, but spends most of his nights thinking about baseball. He writes for Pinstripe Pundits, FanGraphs and The Hardball Times. He's also on the twitter machine: @_chris_mitchell None of the views expressed in his articles reflect those of his daytime employer.

Newest Most Voted
Inline Feedbacks
View all comments
Joeys Bat Flipmember
8 years ago

Interesting read, Chris, thanks. I wonder what a similar sort of analysis would look like if you examined the correlation between the two predictors here and a much larger sample of players. My first reaction to this is: woah, these are all over the map.

I’m sure that there’s a stronger correlation between them when looked at the aggregate, and I’m sure that as you have built KATOH you’ve analyzed the predictions to evaluate for accuracy. Have you ever written something up about this? Would you?

8 years ago
Reply to  Chris Mitchell

Maybe create a KATOH+ model, as well?

Similar to what Nate Silver and FiveThirtyEight have done with Polls-only forecasts and then Polls-plus to include endorsements and such.