Happy Labor Day Weekend! Regularly Scheduled Articles Will Resume Tuesday, September 8
April 20, 2007
Lies, Damned Lies
PECOTA Takes On Right-handed Pitching Prospects
Excellent Prospects Player (Age) Upside xx. Daisuke Matsuzaka, Red Sox (26) 216.7 1. Tim Lincecum, Giants (23) 205.6 2. Philip Hughes, Yankees (21) 197.2 3. Yovani Gallardo, Brewers (21) 132.4 4. Kevin Slowey, Twins (23) 123.3
The conventional wisdom is that the title of best pitching prospect in baseball--non-foreign import category--is a toss-up between Philip Hughes and Homer Bailey. PECOTA begs to differ, suggesting that Tim Lincecum belongs in the discussion. Certainly, comparing Lincecum to Hughes or Bailey does not seem like an apples-to-apples exercise--Lincecum had only 31 2/3 professional innings under his belt entering this season, whereas Hughes and Bailey have a couple hundred apiece. Indeed, PECOTA was not originally designed with the idea of evaluating partial-season performances, and it may not be discounting the limited extent of Lincecum's performance enough.
Nevertheless, Lincecum's numbers were absolutely off the charts last season. Following are the top 10 strikeout rates for minor league starting pitchers in the PECOTA database, which runs back to 1997. These are translated (major league equivalent) numbers, framed as strikeout percentage rather than strikeouts per nine innings. A minimum of 100 batters faced is required.
Pitcher Year K% Tim Lincecum 2006 30.9% Scott Kazmir 2002 28.6% Ryan Anderson 2000 28.5% Brad Lidge 2001 27.4% Josh Beckett 2001 26.8% Jesse Foppert 2002 25.9% Juan Pena 1999 25.9% Rich Hill 2005 25.5% Francisco Liriano 2005 25.5% Paul Abbott 1997 25.3%
Yes, that's Paul Abbott, who was 29 and pitching in Triple-A Tacoma at the time, which goes to show that minor league strikeout rates cannot always be read at face value. Lincecum's stuff is for real, and he has the very best translated strikeout percentage in our database. Including his performance at Fresno this year, Lincecum has faced 191 batters as a professional and struck out 86 of them. That simply doesn't happen very often; his strikeout rate is roughly six standard deviations above the average for minor league pitchers. In fact, only a handful of major league starting pitchers in our database had a better strikeout percentage than Lincecum last year. Remember, these are translated numbers:
Pedro Martinez 1999 39.2% Pedro Martinez 2000 38.0% Pedro Martinez 2001 36.3% Randy Johnson 2001 34.7% Randy Johnson 1997 34.6% Randy Johnson 2000 33.5% Randy Johnson 1999 32.8% Randy Johnson 1998 32.1% Pedro Martinez 2002 31.6% Kerry Wood 1998 31.5% Tim Lincecum 2006 30.9% Randy Johnson 2002 30.8% Johan Santana 2004 30.5% Pedro Martinez 1997 30.4%
Lincecum is probably best conceived of as a hybrid of Kerry Wood and Francisco Rodriguez, his top two PECOTA comparables, and he probably has better mechanics than either of them. He certainly isn't a sure thing, as the example of pitchers like Foppert and Ryan Anderson attest, but the upside here is very, very special.
With that said, I'd still probably take Philip Hughes ahead of Lincecum. Hughes, after all, is the Practically Perfect Pitching Prospect, rating well above average in the "big three" categories of strikeout rate, walk rate, and flyball/home run prevention, with bonus points for having a good pitcher's build and being very young for his levels. There isn't a lot else to say about him, other than that, following his 10-strikeout performance on Wednesday, it's time for him to get called up.
However, to go back to the exempted import, I'd concur with PECOTA and take Daisuke Matsuzaka above them both, at least if we're looking at the 5-7 year time horizon that the Upside ratings imply. But I don't think it's a slam dunk, and it's probably going to come down to who stays healthiest. Matsuzaka has had to bear the heaviest workload, Hughes has the most checkered injury history, and Lincecum has the most awkward build. We don't know enough about injuries to know which of those three things represents the greatest risk factor, and I'm not going to stray from my competencies by trying to speculate.
There's a pretty clear drop-off between those three prospects and the rest of this year's class. PECOTA goes for Yovani Gallardo and Kevin Slowey in the next two slots. Gallardo's presence on this list is understandable; his statistical profile was every bit as good as Homer Bailey's last year--in fact, his translations came out ever so slightly better in each of the big three categories:
It's easy to be a bit skeptical about Slowey's ranking, starting with his fastball topping out at about 90. Nevertheless, let's keep a couple of things in mind. Firstly, PECOTA is not mistaking Slowey for a power pitcher (as it might have done with someone like Yusmeiro Petit last year). It expects his strikeout rate to be only about league average in the majors. Secondly, it recognizes that finesse pitchers have less upside than power pitchers; Slowey's Beta (0.85) is noticeably less than that of pitchers like Gallardo (1.03) and Bailey (1.00). Furthermore, it recognizes that finesse pitchers have shorter careers than power pitchers. The graph below displays the Upside ratings for Gallardo and Slowey by year:
You'll note that there is about a 10-point swing between 2007 and 2011, with Slowey looking better for 2007, but falling well behind once we get a few years out. Presumably, this trend would continue to expand if we drew the graph out further. To repeat myself, the PECOTA prospect rankings are designed to evaluate a 5-7 year time horizon, and a 5-7 year time horizon specifically, because that's when a young player can be retained at below market price by his team. Guys like Slowey are the pitching equivalent of Dustin Pedroia, players who are unlikely to be remembered 40 years from now, but could produce a surprisingly high return on investment for their clubs in the meantime.
Very Good Prospects 5. Matt Garza, Twins (23) 98.2 6. Will Inman, Brewers (20) 95.4 7. Homer Bailey, Reds (21) 94.9 8. Jeremy Hellickson, Devil Rays (20) 88.0 9. Andrew Sonnanstine, Devil Rays (24) 69.8 10. Jason Windsor, A's (24) 66.2 11. Sean O'Sullivan, Angels (19) 64.1 12. Jimmy Shull, A's (23) 60.9 13. Sean Gallagher, Cubs (21) 60.0 14. Adam Miller, Indians (22) 58.9 15. Mike Pelfrey, Mets (23) 57.6 16. Nick Adenhart, Angels (20) 57.4 17. Robinson Tejeda, Rangers (25) 57.4 18. Jason Hirsh, Rockies (25) 55.9 19. Jeff Niemann, Devil Rays (24) 53.9 20. Deolis Guerra, Mets (18) 53.8 21. Mitch Talbot, Devil Rays (23) 52.9 22. Jose Garcia, Marlins (22) 52.8 23. Rick Vanden Hurk, Marlins (22) 52.3 24. Jae-Kuk Ryu, Devil Rays (24) 50.6 25. Micah Owings, Diamondbacks (24) 50.0
We've danced around the subject of Homer Bailey, so let's address him directly. Bailey was one of those players that served as one of my benchmarks as I was tweaking PECOTA this winter; it felt like the system was underrating him. Eventually, I found an empirical backing for a change in PECOTA's weighting mechanism, which has been redesigned to rate the most recent season more heavily for very young (and very old) pitchers. This helps Bailey because his numbers improved a lot from 2005 to 2006. It also helped a couple of other pitchers like Gallardo, while hurting the Yusmeiro Petit types of the universe.
Still, Bailey just barely fails to cross the 100-point Upside threshold. The reason, simply enough, is that Bailey's command has been below average. That's certainly something that can be worked around; not only does PECOTA expect Bailey's walk rate to get better, but it also identifies a series of favorable precedents in pitchers like Rich Harden, Scott Kazmir, and Matt Cain. Still, there is also the risk that he heads down the Edwin Jackson/Matt Riley career path, and while Bailey's strikeout rates have been good, they haven't been Lincecum-level good. I can certainly see elevating Bailey above pitchers like Gallardo on the basis of his makeup, but I think Hughes has an unambiguous enough statistical edge that Bailey's scouting reports cannot overcome it, and the same is probably true of Lincecum if you're willing to tolerate the smaller sample sizes.
There are three other players that make up what amounts to PECOTA's top tier of right-handed pitching prospects. One of these--the Twins' Matt Garza--is no surprise, while Will Inman and Jeremy Hellickson are regarded quite a bit more favorably by PECOTA than they are by scouts. Inman and Hellickson have similar profiles; both were dominant last year in spite of being very young for their leagues, but get merely decent grades for their stuff. They also both get compared to Roy Oswalt since they're a little bit undersized, though of course that is a best-case scenario. Baseball America damned Inman with faint praise, describing his "best case" outcome as Joel Pineiro, while Hellickson doesn't crack their Top 100 at all (nor did Hellickson make Kevin Goldstein's list). As is usually the case, PECOTA probably goes too far to the other extreme, and the truth lies somewhere in between.
From this point forward, the pitching prospects become harder to distinguish. The Indians' Adam Miller is almost certainly underrated, since his 2005 statistics were muddled by injury. I'm less sympathetic toward Nick Adenhart, whose scouting reports read similarly to Inman and Hellickson, but whose numbers weren't as good. You can still make a strong case for him based his excellent groundball rates and history of overcoming Tommy John surgery.
The other big names here are Jeff Niemann and Mike Pelfrey. Niemann impressed me a great deal when I watched him in the College World Series a couple of years ago, and probably belongs in the same category as Miller since he's also had to overcome injuries. On the other hand, Pelfrey would seem to have a bit more work left to do, needing to improve both his secondary offerings and his control. His control numbers deteriorated with every promotion last year, including some unimpressive work in the major leagues, and I think he's been rushed a bit. What scares me in particular is that Pelfrey had some very good walk rates in college, so the command problems are relatively new.
Take this with a grain of salt, because I have zero training as a scout, and I have no more credibility to assess body language than one of those faux experts on the O'Reilly program. But I watched Pelfrey's most recent start against the Nationals, and while his mechanics looked clean, his demeanor was very casual and he almost seemed to be rushing through his at-bats. I personally read this as nervousness rather than a lack of concentration, though of course nerves can impede concentration. Either way, I have a hunch that he's going to have some rough sledding if he stays in Shea this year.
I'll leave you to browse the PECOTA cards for assessments of the rest of these players. Props to the Marlins' Rick Vanden Hurk, who has an excellent chance to become the best Dutchman in the major leagues since Bert Blyleven.
Good Prospects 26. Tyler Clippard, Yankees (22) 49.3 27. Kyle Winters, Marlins (20) 47.5 28. Carlos Carrasco, Phillies (20) 45.8 29. Jamie Richmond, Braves (21) 45.7 30. Justin Germano, Phillies (24) 45.6 31. Brandon Erbe, Orioles (19) 43.0 32. Randy Wells, Cubs (24) 42.2 33. Jair Jurrjens, Tigers (21) 40.8 34. Chris Sampson, Astros (29) 39.2 35. Yusmeiro Petit, Diamondbacks (22) 39.0 36. Michael Bowden, Red Sox (20) 38.8 37. Lance Broadway, White Sox (23) 38.0 38. Humberto Sanchez, Yankees (24) 37.9 39. Johnny Cueto, Reds (21) 36.8 40. Jesse Litsch, Blue Jays (22) 36.5 41. J.D. Durbin, Phillies (25) 36.3 42. Jack Egbert, White Sox (24) 36.0 43. Ubaldo Jimenez, Rockies (23) 35.7 44. Taylor Buchholz, Rockies (25) 35.3 45. Steven Johnson, Dodgers (19) 35.2 46. Matt Albers, Astros (24) 34.7 47. Neftali Feliz, Braves (19) 34.5 48. Dustin Nippert, Diamondbacks (26) 34.4 49. Clay Buchholz, Red Sox (22) 34.3 50. Samuel Deduno, Rockies (23) 34.1 51. Eduardo Morlan, Twins (21) 33.4 52. Darrell Rasner, Yankees (26) 33.0 53. Hayden Penn, Orioles (22) 32.5 54. Charlie Haeger, White Sox (23) 32.3 55. Thomas Diamond, Rangers (24) 30.4 56. Jeff Karstens, Yankees (24) 29.0 57. Shane Lindsay, Rockies (22) 28.3 58. Gabriel Hernandez, Marlins (21) 28.0 59. Josh Banks, Blue Jays (24) 27.9 60. Cesar Carillo, Padres (23) 27.6 61. Eric Hurley, Rangers (21) 27.6 62. Christopher Tillman, Mariners (19) 27.0 63. Juan Gutierrez, Astros (23) 27.0 64. Billy Buckner, Royals (23) 26.5 65. Thomas Mendoza, Dodgers (19) 26.5 66. Michael Elkstrom, Padres (23) 25.0
The two that PECOTA is most notably down on are Ubaldo Jimenez and Humbero Sanchez, each of whom have had a lot of problems with their command. What's notable about these two pitchers is that, while their statistics superficially aren't that far removed from the Bailey/Gallardo class, they're also a couple of years older, and that makes some of the higher-upside names on their comparables lists disappear. Jimenez, for example, doesn't really get any interesting comparables until you get to Freddy Garcia at #18, and A.J. Burnett at #19. If a pitching prospect is polished, age might not be terribly important, but if he has some work left to do--and both Jimenez and Sanchez fall into that category--each passing year reduces the chances that he's going to make a big leap forward.
Let me run one more graph, which is a plot of the Upside ratings of all 25-and-under players in the PECOTA database, plotted on a logarithmic scale:
This is meant to provide some visual evidence of the phenomenon I discussed last week. While elite pitching prospects generally rate a bit lower than elite hitting prospects, the flip side is that pitching prospects have a longer tail; mediocre pitching prospects are more likely to turn around and surprise us. This finding might seem like an artifact of the way that PECOTA is constructed, but I don't think it is, as it conforms relatively well with Rany Jazayerli's findings on the amateur draft. This has some interesting implications for draft strategy; in particular, there's an argument to be made for using the first 2-3 picks in the draft on hitting talent almost without exception, but then heavily scouting pitchers all over the country, and loading up on them in the back end of the draft.
Average and Marginal Prospects
Wade Davis, Devil Rays (21) 23.9 Philip Humber, Mets (24) 16.6
Wade Davis ranked at the very bottom of Kevin's positional top 20, and his numbers didn't translate well out of the pitcher-friendly Midwest League. Kevin further suggested that 2007 will be a make-or-break year for Davis; so far his results have been just average. I can't quite figure out why PECOTA is so down on Philip Humber; his minor league playing time has been limited because of injuries, and I think that's causing the system to mistake him for a journeyman 'tweener.
Honorable mention goes to the Yankees' Dellin Betances, who didn't pitch quite enough to qualify for our list, but whose top four comparables are Matt Cain, Francisco Liriano, Joel Zumaya, and Francisco Rodriguez. A Bronx native who is just 19 and stands at 6'9", he's the sort of pitcher that everyone could be talking about a year or two from now.
The Big Picture: Rankings Combined With Non-Rookies 25 Years Old Or Younger 1. Jeremy Bonderman, Tigers (24) 269.2 2. Felix Hernandez, Mariners (21) 250.5 3. Jered Weaver, Angels (24) 220.3 4. Tim Lincecum, Giants (23) 205.6 5. Philip Hughes, Yankees (21) 197.2 6. Josh Johnson, Marlins (23) 148.2 7. Rich Harden, A's (25) 143.5 8. Justin Verlander, Tigers (24) 134.3 9. Yovani Gallardo, Brewers (21) 132.4 10. Kevin Slowey, Twins (23) 123.3
Forty-eight hours ago, I would have gotten hate mail for rating Bonderman ahead of Hernandez, but all of the sudden that's looking surprisingly prescient. What's interesting about Hernandez is that, without knowing anything about his questionable mechanics, PECOTA assigned him a remarkably high attrition rate this year (28%). What's also interesting is that PECOTA has his Upside score climbing a lot by the time we get to 2009 and 2010. Both the Mariners and their fans may just need to be patient.
We'll wrap our prospects series up next week by running more lists than you can shake a stick at.