When scouting, it is generally better to scout based off
what you see (bat speed, approach, swing path, etc.) vs. what the numbers say.
But to minor league numbers hold any relevance? Is there a correlation between
a player’s minor league stats and his future major league stats? I took a
sample of 40 players, who have debuted since 2000, played with the Rangers at
some point in their career, and have at least 200 major league at-bats. I took
their statistics at the Low-A, High-A, AA, and AAA levels and ran tests to find
both the average difference between the stats at each level compared to the
player’s MLB stats and the strength of correlation between the numbers for AVG,
OBP, and SLG.
Here are the means, or averages, the standard deviations, or
average difference from the mean, and r values, or strength of correlation (0-.5
is a weak correlation, .5 to .8 is a moderate correlation, and .8 to 1 is a
strong correlation) for each of the levels for batting average,
on-base-percentage, and slugging percentage. If the r value is at .5 or higher,
that shows that the numbers have some relevance. I have also listed the current
slash lines for the players on each roster that have 200+ PAs at any minor
league level, and then translated that line to their projected major league
career slash line based on the averages. Remember that there are tons of
players who are well better or well worse than the average, so the projected
stats are more of a fun tool than something to read anything into.
Keep in mind that the numbers for High-A are biased and the
average drop should not be so extreme. Much of the sample that I used were
hitters that went through the hitter-friendly California League, back when the
Rangers High-A team was out in Bakersfield, whereas now these players are
playing in the pitcher-friendly Carolina League. The drop shouldn’t be so big
for these hitters, so take that into consideration.
Low-A Batting Average:
|
mean=-.027
|
Standard deviation=.037
|
r=.334 (weak)
|
Low-A OBP:
|
mean=-.034
|
Standard deviation=.039
|
r=.4626 (weak)
|
Low-A Slugging Percentage:
mean=-.025
|
Standard deviation=.077
|
r=.508 (moderate)
|
In Hickory (-.027/-.034/-.025):
C Kellin Deglan (.231/.331/.393) to (.194/.284/.340) using
High-A stats
C Kevin Torres (.241/.279/.324) to (.214/.245/.299)
1B Ronald Guzman (.282/.331/.398) to (.255/.297/.373)
SS Luis Marte (.222/.248/.280) to (.195/.214/.255)
2B Nick Urbanus (.198/.268/.234) to (.171/.234/.209)
1B Nick Vickerson (.237/.377/.360) to (.210/.343/.335)
OF Lewis Brinson (.246/.325/.429) to (.219/.291/.404)
OF Nomar Mazara (.227/.304/.362) to (.200/.270/.337)
High-A Batting Average:
|
mean=-.037
|
Standard deviation=.032
|
r=.4208 (weak)
|
High-A OBP:
|
mean=-.047
|
Standard deviation=.035
|
r=.5069 (moderate)
|
High-A Slugging Percentage:
mean=-.053
|
Standard deviation=.06
|
r=.771 (moderate)
|
In Myrtle Beach (-.037/-.047/-.053):
C Jorge Alfaro (.259/.331/.443) to (.232/.297/.418) using
Low-A stats
C David Lyon (.239/.321/.416) to (.212/.287/.391) using
Low-A stats
SS Hanser Alberto (.213/.253/.287) to (.179/.214/.241) using
AA stats
1B/OF Preston Beck (.249/.347/.352) to (.212/.300/.299)
2B Chris Bostick (.282/.354/.452) to (.255/.320/.427) using
Low-A stats
3B Joey Gallo (.245/.334/.610) to (.218/.300/.585) using
Low-A stats
OF Royce Bolinger (.285/.330/.426) to (.248/.283/.373)
OF Zach Cone (.262/.326/.461) to (.235/.292/.436) using
Low-A stats
OF Odubel Herrera (.257/.289/.339) to (.223/.250/.293) using
AA stats
OF Nick Williams (.293/.337/.543) to (.266/.303/.518) using
Low-A stats
AA Batting Average:
|
mean=-.034
|
Stardard deviation=.03
|
r=.3556 (weak)
|
AA OBP:
|
mean=-.039
|
Stardard deviation=.038
|
r=.4054 (weak)
|
AA Slugging Percentage:
mean=-.046
|
Stardard deviation=.07
|
r=.523 (moderate)
|
In Frisco (-.034/-.039/-.046):
C Pat Cantwell (.253/.318/.316) to (.216/.271/.263) using
High-A stats
C Tomas Telis (.263/.292/.348) to (.229/.253/.302)
C Zach Zaneski (.237/.310/.364) to (.203/.271/.318)
1B Trever Adams (.257/.334/.386) to (.220/.287/.333) using
High-A stats
2B Edwin Garcia (.252/.308/.315) to (.215/.261/.262) using
High-A stats
2B Rougned Odor (.281/.326/.465) to (.247/.287/.419)
3B Ryan Rua (.291/.371/.480) to (.257/.332/.434)
1B/OF Jordan Brown (.302/.348/.451) to (.274/.313/.413)
using AAA stats
OF Chris Grayson (.192/.299/.295) to (.155/.252/.242) using
High-A stats
3B/OF Drew Robinson (.257/.367/.405) to (.220/.320/.352)
using High-A stats
OF Jake Skole (.201/.319/.286) to (.164/.272/.233) using
High-A stats
OF Jake Smolinski (.258/.345/.401) to (.230/.310/.363) using
AAA stats
AAA Batting Average:
|
mean=-.028
|
Standard deviation=.023
|
r=.4686 (weak)
|
AAA OBP:
|
mean=-.035
|
Standard deviation=.027
|
r=.6687 (moderate)
|
AAA Slugging Percentage:
mean=-.038
|
Standard deviation=.06
|
r=.631 (moderate)
|
In Round Rock (-.028/-.035/-.038):
C/1B Brett Nicholas (.289/.357/.474) to (.255/.318/.428)
using AA stats
3B Alex Buchholz (.267/.328/.398) to (.239/.293/.360)
2B Kensuke Tanaka (.323/.403/.404) to (.295/.368/.366)
INF Guilder Rodriguez (.272/.360/.291) to (.244/.325/.253)
OF Jared Hoying (.249/.284/.479) to (.221/.249/.441)
OF Brad Snyder (.288/.352/.504) to (.260/.317/.466)
OF Ryan Strausborger (.232/.297/.360) to (.198/.258/.314)
using AA stats
What I was expecting to see was a very low strength or
correlation for the lower levels, and higher ones at the upper levels. Low-A
went as expected, with the lowest strength of correlation for both batting
average and slugging percentage. However, Double-A had very weak correlations,
with the weakest in OBP and second weakest in average and slugging. Triple-A unsurprisingly
had fairly strong correlations, and High-A very surprisingly, at least to me,
had pretty strong correlations, including the strongest correlation of any stat
at any level with an r value of .771 in slugging percentage.
What these numbers showed is that there definitely is a
correlation between a player’s minor league stats and his future major league
stats, but not enough of one to overlook a player’s raw tools, or to overvalue
a player’s minor league numbers. In other words, high-performing minor league
players such as Nate Gold (who never made the majors) don’t even come close to
consistently translating that success to the major league level, but they do
have a higher tendency to do so than a lesser performer at the minor league
levels, which makes sense.