Baseball Digest Daily
HomeAbout UsArticlesBlogPlayer TrackerMLB StatsBaseball ProspectusSeamheads

Profile - Jay Bruce

by Brian Cartwright

The Cincinnati Reds selected high school outfielder Jay Bruce with the 12th pick of the 1st round of the 2005 First Year Player Draft. Justin Upton (1), Alex Gordon (2), Jeff Clement (3), Ryan Zimmerman (4), Ryan Braun (5) and Troy Tulowitzki (7) were those selected ahead of him who have already established themselves as starting players in the major leagues.

Coming out of high school, Bruce split his first season in 2005 between two rookie league teams, posting isolated power numbers over .200 at each. In 2006, at age 19, he spent the whole season with Dayton in A ball, and again showed good power with 42 doubles and 16 home runs. 2007 saw a climb across three levels, from Sarasota in High A to Chattanooga in AA to Louisville in AAA. He hit over .300 at each, combining for 46 doubles and 26 home runs. Hitting .364 with 10 HRs for AAA Louisville by late May 2008, Bruce got the call to join the Reds. He didn’t disappoint, reaching base in his first six plate appearances. After 16 games, Bruce is still hitting .382 with 4 HRs.

In looking at Bruce’s professional record so far, what type and how good of a player does he look to be?

In calculating Major League Equivalencies from minor league stats, adjustments must be made for park factors, the same as in the major leagues, but then also for the lesser competition at each level of the minors.

At each level, I have taken the batting totals for each player and normalized them for park factors. Then I grouped the players who played on more than one level, and summed the weighted means of the records on each level. The level factors are the ratios of each component stat (SI, DO, TR, HR, BB, SO) from one level to another.

Comparing the projected MLE to actual major league stats, the following root mean square errors were observed:

(ISO is Isolated Power, or SA-BA. BABIP is Batting Avg on Balls In Play, or H-HR divided by balls in play)

BA OB SA ISO BABIP BB SO
.017 .023 .031 .024 .024 .009 .034

70% of the time, the projection will be within this error margin of the “true” major league value that is being estimated. 95% of the projections should be with twice twice the error rates from the true values.

Actual BA OB SA AB H DO TR HR HP BB SO SB CS DP
2005 18 R .270 .331 .500 122 33 9 2 5 0 10 31 4 6 2
2005 18 R .257 .358 .457 70 18 2 0 4 0 11 22 2 2 2
2006 19 A .291 .355 .516 444 129 42 5 16 0 40 106 19 9 7
2007 20 A+ .325 .372 .588 268 87 27 5 11 2 22 67 4 4 4
2007 20 AA .333 .405 .652 68 22 7 1 4 0 8 20 2 1 0
2007 20 AAA .305 .353 .567 187 57 12 2 11 1 14 48 2 2 1
2008 21 AAA .364 .373 .630 184 67 9 5 10 4 12 45 8 1 4
2008 21 ML .382 .468 .632 68 26 5 0 4 1 9 12 2 2 1
 
MLEs BA OB SA ISO BABIP BB% SO% AB H DO TR HR HP BB SO
2006 19 A .254 .309 .469 .215 .304 .071 .241 457 116 42 2 17 3 35 120
2007 20 A+ .298 .339 .563 .264 .366 .058 .264 279 83 25 5 13 2 17 79
2007 20 AA .279 .353 .533 .255 .410 .103 .327 66 19 8 1 3 0 8 24
2007 20 AAA .254 .301 .477 .223 .317 .061 .291 192 49 11 1 10 1 12 60
2008 21 AAA .302 .319 .521 .219 .368 .035 .273 193 58 8 3 9 0 7 56
2008 21 ML
.382 .468 .632 .250 .423 .127 .152 68 26 5 0 4 1 10 12
 
Actual BA OB SA ISO BABIP BB% SO% AB H DO TR HR HP BB SO
2006 19 .254 .309 .469 .215 .304 .071 .241 457 116 42 2 17 3 35 120
2007 20 .280 .327 .529 .248 .351 .064 .282 537 150 41 6 26 3 37 163
2008 21 .323 .360 .550 .227 .384 .060 .239 261 84 13 3 13 1 17 68
 
MLEs BA OB SA ISO BABIP BB% SO% AB H DO TR HR HP BB SO
Cartwright .275 .323 .507 .232 .330 .066 .258 553 152 42 5 25 3 39 155
Sickels .280 .320 .475 .195 .339 .055 .238 559 157 37 4 22 3 33 143
 
Percentile Rank 74% 48% 96% 97% 97% 33% 1%

My projection line includes a regression to the league mean for the BABIP after all the lines were summed. For Bruce, this cost him 5 points of BA, SA and OB. John Sickles recently did a year-by-year career projection for Bruce. The line above represents the mean of Sickels’ 2009-2011 projections. Sickels projects a higher BABIP, but lower ISO, BB and SO rates than I do, but that does not make any major adjustments to the batting line. The totals for the various stats represent what Bruce would be expected to do over 600 plate appearances for each projection.

Bruce has shown very steady ISO, BB and SO rates, with all three easily with one standard deviation (SD) of his career total for the last three seasons. His BABIP has shown an increase each of the past three seasons, in amounts more than one SD, although through less than half a season for 2008, it is very likely that the rate will fall to near his three year mean. If Bruce can maintain a BABIP near .340 he can (the all-time career leaders are

Bruce has an unusual combination of stats. Ryan Howard and Wily Mo Pena are the only players to have as high a SO rate as Bruce’s projection while maintaining as high a career BABIP (through 2007). Very few players have been able to walk at a below average rate, strike out well above average, and still have well above average BABIP and ISO rates. I found six comparable players, although all have lower ISO, higher BB and lower SO rates than are projected for Bruce, which suggests that if he continues his high SO rate he will not maintain his high BABIP and ISO rates in the long run.

Career leaders in BABIP are Derek Jeter .361, Rod Carew .359, Ichiro Suzuki .357, Miguel Cabrera .354 and Matt Holliday .352.

Bruce’s most comparable batters:

Player BA OB SA ISO BABIP BB% SO%
Jay Bruce .275 .323 .507 .232 .330 .066 .258
Josh Phelps .273 .341 .476 .204 .335 .072 .256
Geoff Jenkins .277 .337 .496 .219 .324 .066 .230
Bill Hall .264 .317 .467 .203 .321 .072 .246
Preston Wilson .264 .326 .468 .205 .314 .076 .246
Curtis Granderson .280 .342 .493 .214 .344 .083 .238
Craig Wilson .262 .350 .474 .211 .335 .081 .280

This is a set of players with very similar batting results but a mixed bag of career paths. Geoff Jenkins is nearing the end of a solid but unspectacular career, while Curtis Granderson is a rising star. Bill Hall has alternately been brilliant and struggled, while Preston Wilson and Craig Wilson lost their starting jobs before the age of 30. Despite a fine minor league record and good results when he has played in majors, Josh Phelps has struggled to stay on a major league roster.

Jay Bruce has shown more power than any of this group, and that will be boosted another 10% by playing for Cincinnati in the National League Central division. Can Bruce maintain a high BABIP while striking out 25% of his plate appearances? By the end of 2009, he will still only be 22, but should have enough Major League playing time in the books to show us where he is headed.

Comments (5) -> “Profile - Jay Bruce”

  1. Matt Mitchell
    18 June 2008 06:07
    1

    Brian,

    Care to share any more details on how you do your calculations? There are a lot of black boxes here where numbers go in and other numbers come out.

    It is interesting to contrast your list of comparables to that of Baseball Prospectus, which had Cliff Floyd, Eric Chavez, Ian Stewart, and Manny Ramirez listed. Your list does not seem quite as optimistic.

  2. KJOK
    18 June 2008 20:05
    2

    I’m also curious about which MLE formula you’re using…

  3. Brian Cartwright
    19 June 2008 03:14
    3

    The basic model for the MLE is find a group of players who played for both Louisville and Cincinnatti. If Player A has 600 PA and 150 SO in AAA, and 300 PA and 80 SO in MLB, put 75 SO (300/600*150) in the AAA column, 80 in the MLB column. Repeat this for all the players in the selection, in all categories. Always project the larger PA down to the smaller. Then sum the AAA columns and the MLB columns, and get the ratio from AAA to MLB.

    The more advanced method is to do park factors and first normalize all the major and minor stats. Instead of going from one team to another team, go from one league to another. Then apply the league factor to the park normalized stats.

    BP’s weighted mean PECOTA for 2008 is .270/.338/.513, which barely differs from my line. They give him a BB% of .087, which is slightly above average and well above my .064, although their percentile histogram puts his BB% at 35%, which is almost exactly the same as mine. BP has his SO% at .236, which is close but a little lower than mine.

    I believe BP regresses at each season, which I believe puts someone too close to the league mean. I don’t regress until I get the final total.

    As far as comparable players, I calculated Jay Bruce’s MLE BABIP, ISO, BB% & SO%, then searched for all players whose career major league numbers were within one SD on all four components. BP makes comparisons to other players MLEs, but I only looked at others major league totals. BP also uses age, which I did not, but I can in the future.

    Cliff Floyd and ERic Chavez look to be good comparables IF Bruce’s BB rate is as high as BP projects (normal or slightly above), but I don’t see that from Bruce’s numbers, nor does BP’s histogram. Manny Ramirez had similar BABIP and ISO in his early 20’s, but always had a high walk rate and a better SO rate than Bruce. I did present the numbers of my comparables so you could see how they matched, which BP does not.

  4. Matt Mitchell
    21 June 2008 15:57
    4

    Thanks! That explanation helps a lot.

  5. Brian Cartwright
    21 June 2008 19:21
    5

    There’s Chone, and BP has Pecota…I’m thinking I should name mine Castillo (CArtwright’s STandard Interpretation of League to League Observations?)

    I updated my Pirate minor league database since 1996 to include 2008 so far, and was looking at the MLEs.

    Jose Castillo babip iso bb so
    Minor 283/336/422 323 138 061 167
    MLE 254/301/388 297 133 053 194
    Major 257/300/388 298 131 046 185

Reply