September 20, 2006
After a fairly mundane start to his major league career, Brian Roberts has turned into one of the few bright spots on the Orioles' roster. His 2005 season was worthy of MVP consideration, and while his follow-up hasn't been as good, it's still been a very strong season.
Roberts was drafted 50th overall by the Orioles in 1999; he was a sandwich pick between the first and second rounds, received as compensation for Cleveland's signing of Roberto Alomar. His college career was highly regarded, as Baseball America rated him the top defensive college player, and he was second team All-American in his final season at the University of South Carolina after hitting .353 with 12 home runs and a NCAA-leading 67 stolen bases. Roberts finished his college career ranked fifth all-time for steals.
Roberts started out as a shortstop in Single-A Delmarva, and posted a poor batting line: .240/.345/.323 was disappointing, but he did walk in 13.9 percent of all plate appearances, and managed to steal bases at a 77 percent success rate. The walks and steals added up to a .317 Secondary Average, but with an Isolated Power of only .083. His 2000 season at High-A Frederick would provide a more aesthetically pleasing batting line of .301/.403/.374, but his success rate stealing bases suffered--only 13 of 23, or 57 percent--and his power was still nonexistent.
Roberts did manage to cut his strikeouts down from 21.6 percent of all plate appearances to a much more manageable 12.1 percent, a figure he has never since strayed too far from subsequently. His season was shortened by an April surgery to remove bone chips from his elbow; this kept him out of High-A until mid-July. For his efforts, however little playing time was involved, Roberts was named the tenth-best prospect in the organization by Baseball America heading into 2001.
Roberts' 2001 season had stops at three levels, with trips to Double-A, Triple-A and finally Baltimore.
AB AVG OBP SLG SecA XBH% ISO 2B+3B SB BB% K% Bowie (AA) 81 .296 .366 .420 .358 33% .124 7 10 9.5% 12.6% Rochester (AAA) 161 .267 .376 .323 .354 14% .056 5 2 14.7% 11.6% Baltimore 273 .253 .284 .341 .168 25% .088 15 12 4.5% 12.3%
After a strong start at Bowie, Roberts was quickly promoted to Triple-A Rochester. His walk rate improved, an excellent sign of development, but the small amount of power he had picked up evaporated. After the previous off-year on the basepaths, he was very successful once again, swiping 33 of 36 bags (91.6 percent) successfully between levels in the minors. Apparently the recent spate of quick promotions from the likes of Adam Loewen, Daniel Cabrera, Nick Markakis, and Hayden Penn have been going on in Baltimore longer than I initially recalled, as Roberts was brought up quickly as a product of the same organizational philosophy.
As you can see from the numbers in the table, he didn't exactly light up the scoreboard in his major league debut. A sub-.300 on-base percentage from a player whose main strength at the time was his ability to get on base is somewhat disheartening; he did manage to remain productive when he reached base though, stealing 12 bases while only getting nabbed 3 times. The fact that his strikeout rate didn't spike upward was also a very positive sign, and his Isolated Power and extra-base hit rates actually improved from his time at Rochester. Considering he was 23 years old with a mere 600 professional at-bats heading into his major league debut, we can cut him some slack for his initial play.
Baseball Prospectus 2002 was somewhat harsh towards Roberts, but with good reason:
In the organizational tradition of David Lamb and Jesse Garcia, Roberts has been praised for his fielding, so his performance in 2001 was a disappointment. He was coming off of elbow surgery in 2000, and his throwing was erratic. Nine of his 14 major-league errors at shortstop were on throws-not that a stiff-backed David Segui helped much. Roberts hit well for his first two months, but once teams saw him again, his bat disappeared like a teenager's allowance. If his defense doesn't recover, he has no future beyond the International League.
When your shtick is defense and getting on-base, and you don't do either, it's a problem. As previously stated, other parts of his game were starting to come together nicely, but they were supposed to complement his on-base skills and defense, not be the central part of his production.
The 2002 season saw Roberts split time between Triple-A and the majors, with trips to the Orioles on three separate occasions. He had managed to improve on his previous season's numbers during his 78 games in Rochester, but continued to struggle in the majors:
AB AVG OBP SLG SecA XBH% ISO 2B+3B SB BB% K% Rochester (AAA) 313 .275 .361 .377 .288 22% .102 16 22 11.1% 12.8% Baltimore 128 .227 .308 .297 .242 24% .070 6 9 10.1% 14.1%
His walk rate returned to about ten percent, rather than the poor 4.5 he posted in his major league debut, and his strikeout rates stayed relatively the same. His 82 percent success rate on the basepaths for the O's showed that was would be a capable leadoff presence if he could just start to hit more; between the stolen bases and the walks, the Orioles were looking at a quality leadoff hitter on a team that really could use some offensive help. Most of his playing time came at second base rather than shortstop, and according to Rate, he was exactly league average there defensively. This was a positive sign after all of the throwing trouble his surgically repaired elbow had given him previously. Roberts would participate in the Puerto Rican Winter League, and had himself a fine time at the plate--he tied for first in hits and steals and posted a .417 on-base percentage.
PECOTA forecasted a .241/.315/.331 line for the second basemen, which was actually an improvement on his major league career to date. Roberts would start the 2003 season in Ottawa, Baltimore's new Triple-A affiliate, and would hit very well there, putting up a line of .315/.401/.399 with walks in 12.8 percent of all plate appearances, and strikeouts in only 5.7 percent. He was quickly brought to the majors and started in mid-May; for the first time, his progress in the minors translated to the majors. His line with the O's for 2003 was .270/.337/.367, which wasn't anything special, but was a far cry from the paltry production of previous campaigns. He managed to draw a walk nine percent of the time, and he kept up his low strikeout rate. He also managed to steal 23 bases while only getting caught six times, and according to The Fielding Bible played a fantastic defensive second base, fifth-best in the majors. Really all that was lacking from his game at this point was some semblance of power; even the ability to slug .400 in the majors would have been a boon.
Baseball Prospectus 2004 outlined the second base "problem" the Orioles were having, with three capable players for the position. Due to Roberts' defensive skills, there really should not have been any sort of discussion, but the organization was obsessed with Jerry Hairston Jr. Roberts was projected by PECOTA to finish at .262/.334/.361 but, after finally taking the second base job away from Hairston, he managed to hit for some power at the same time that he walked and stole bases. Roberts would hit 50 doubles on the season, a record for a switch hitter, but he only managed to collect six other extra-base hits. For the season, his line was .273/.344/.376, besting his PECOTA projection, and included a fairly high number of extra-base hits for a second baseman.
PECOTA anticipated a .268/.342/.370 2005 season with only 28 doubles, but an increase in his home run production. The basic premise was correct, as he would hit fewer doubles and increase his home run power, but the projection was way off from anything anyone conceived over the winter. Roberts slugged an impressive .314/.387/.515 with 18 home runs and 45 doubles to go along with 27 steals and 67 walks. He was ranked as the sixth-best defensive second baseman in the majors to boot, by the aforementioned Fielding Bible. After a horrific collision with Bubba Crosby ended his season, everyone had themselves another data point for the "Age-27 season" phenomenon.
This year, Roberts has had a fine season bouncing back from his elbow injury, although not quite at the caliber of his 2005 campaign. His season line currently stands at .293/.356/.417, the second best of his career. He's stolen 35 bases at a success rate of 83 percent, and has managed to keep a portion of the home run power he developed, with nine on the season to go along with his 31 doubles and three triples. What changed in Roberts' game that made him a better hitter out of nowhere?
Year P/PA FB% LINEDR% BA/BIP GB% POPUP% HR/F ------------------------------------------------------------------ 2004 4.0 39.6% 21.3% .316 39.1% N/A 1.8% 2005 3.7 31.6% 24.4% .344 36.6% 7.4% 11.7% 2006 3.8 29.9% 21.6% .318 44.0% 4.5% 6.4%
We can see that from 2004 to 2005, Roberts increased his BABIP significantly at the same time that he cut down on his grounders and flyballs, while increasing line drives. When he did hit flyballs, he managed to hit them very well, with home runs coming on almost 12 percent of all outfield flies, a ten percent jump from the previous year. Though his BABIP has regressed, and his HR/F has gone down, Roberts still hits a considerable number of line drives, and doesn't pop up weakly as often. The most significant change comes from the transition from the 2003 to 2004 season. Roberts only hit outfield flies 25 percent of the time, with line drives and grounders making up a considerable portion of his batted balls. With the increase in flyballs came more doubles, and the next season, more of those flyballs turned into homers. This year has been a balance between the 2004 and 2005 transition period. All in all, his power makes him a valuable commodity, even more so when his defense and skill on the basepaths is considered. Kudos to PECOTA for discovering time travel in order to cheat on the forecasts; his 2006 projection was .282/.356/.417, and his season line is currently .293/.356/.417. Not too shabby, considering the changes in performance from 2003 to 2005 that the system worked with.
Looking back, it's pretty amazing that Roberts ever had to contend with the likes of Jerry Hairston Jr., Luis Matos, Deivi Cruz, and Chris Richard for a job with the Orioles. No wonder Melvin Mora is upset.