Statistical Analysis FAQ
How
can we use statistics to analyze players?
If
you’ll give me a moment to speak, I’ll be glad to tell you. Though the
truths that follow apply at all levels, keep in mind that the emphasis here is
to use statistics to arrive at an educated guess of whether or not a minor
leaguer will go on to have a productive Major League career.
What
can statistics tell us?
Statistics can lead us to make a lot of assumptions about a player without
ever having seen him. We can assume whether a player is fast or slow based on
his stolen base numbers. We can assume the quality of his bat based on his
batting average. We will generally assume a pitcher is bad if he has a high ERA.
Sometimes these assumptions are deadon. Sometimes they are greatly misleading.
Statistics provide significant material to infer the qualities of a player, or
lack thereof, because of his numbers. Taking a step past the quick and easy
inferences and heading toward interpretation is what Baseball Examiner is
all about. How do we properly interpret statistics without inferring an
inaccurate assumption?
Hey! I
thought I was the one asking the questions here!
Sorry.
So,
how do we properly interpret statistics without inferring an inaccurate
assumption?
Good
question! There are several factors to include in the proper interpretation of
statistics. We’ll address those factors one by one.
When
analyzing players, do statistics tell us everything we need to know?
No! There
is plenty of helpful information in statistics, but there is plenty it does not
tell us. It does not tell us the strength of a player’s arm, the softness of
his hands, or the mindset of his approach at the plate or on the mound. The
scouts will give us more information regarding a player’s tools than we can
ever infer or interpret, but statistics are a form of litmus test to see whether
a player is fulfilling or exceeding his potential.
So,
which stats lead us to a more proper interpretation of a player’s skills?
Let’s
start by discussing which stats are often misleading. When analyzing pitchers,
perhaps the most misleading stat is record. A pitcher’s record does not tell
us how well or poorly he has performed. It is a team stat. Teams win and lose,
not individuals. Sure, the better a pitcher does the more likely he’ll win.
But so many other factors are to be considered in the outcome of a game. Does
the team have a powerful lineup? Or is it sickly? Pitchers can also get no
decisions due to a lateinning comeback after they’ve left the game for
outings which they deserved either a win or loss. Record is a team stat.
The
better the pitcher is, the better he’ll give his team a chance to win. But he
can only do so much. Ben Sheets’ 2004 earnedrun average is nearly two full
runs less than Mark Mulder’s (2.70 to 4.43). But Mulder was 178 and Ben
Sheets was 1214. Pitchers win or lose with their team.
Well,
if record is a misleading stat, then surely ERA will tell us whether he is any
good.
That’s
not a question, but I’ll respond anyway. ERA is indeed a better indicator than
is pitching record, but it is still not sufficient. ERA is more effective on a
quick glance basis because it is a generalization. ERAs under 3.00 are more
impressive than ERAs under 4.00 which are more impressive than ERAs under 5.00.
But Baseball Examiner is not in the business of quick glance generalizations. We
want to get more into the nitty gritty than that.
The
problem with ERA is that there are other weighing factors outside of what the
pitcher can control. Is the pitcher helped by a good defense thus making him
look better than he ought? Or is his ERA tainted by a bad defense? Sure, a bad
defense will allow unearned runs that don’t count toward the pitcher’s ERA,
but a bad defense will also allow base hits to sneak through and baserunners to
advance on plays where no error is ruled. A pitcher may have to get 4 or 5 outs
to escape an inning without a single error being officially committed. And
it’s all earned!
A
starting pitcher’s bullpen will also affect his ERA. Example: two pitchers
throw identical games, the same pitch sequences, the same ball and strike
counts, etc. Pitcher A goes 6 1/3 innings, allowing 2
earned runs on 7 hits and 3 walks with 6 strikeouts. Pitcher B goes 6 1/3 innings, allowing 2 earned runs on 7 hits and 3 walks with 6 strikeouts. Who was
more impressive? Neither. With identitical performances, they both have an ERA
of 2.84 for this outing. However, they both exit the game one out into the 7^{th}
inning with the bases loaded. Pitcher A has a bullpen that comes in and slams
the door. His 2.84 ERA remains intact. Pitcher B has a bullpen that throws
gasoline on the fire and all 3 inherited runners score bringing his earned run
total to 5 and his ERA for the outing to 7.11. Now who appears to have
had a more impressive outing? 2.84 or 7.11? Don’t fall for it! They were
identical performances!
It’s
possible that over the course of a long season, a bad defense and bad bullpen
can make a better pitcher’s ERA higher than an inferior pitcher’s ERA
that has been aided with a good defense and good bullpen. And we haven’t even
discussed ballpark effects!
If ERA
can even be misleading, then what stats do matter?
Well,
we’ve already stated that a bad defense can give up additional base hits. But
hits per inning is still getting us in the right direction. More effective than
ERA is looking at how many baserunners a pitcher allows. The less frequently
that opposing batters reach base, the less likely they are to score runs. The
two main ways batters reach base are via hit or walk. So, what helps a pitcher
be successful? Limit the hits and walks allowed.
If
there are factors outside of what a pitcher can control to affect his
statistics, then what can he control?
If a
pitcher doesn’t want to take the chance that the defense will blow a play, the
best thing to do is get an out without making the defense have to do a thing.
That is what we call a strikeout.
On the
flip side, giving up a homerun allows baserunners (and of course runs) that the
defense never even had a chance to make a play on. It’s better to allow the
defense an opportunity to make (or break) a play than give up runs with no play
at all.
The
pitcher who, relative to innings pitched, keeps his hits, walks and homerun
totals low while striking out a bunch is a guy we’re going to focus in on.
So, if
that’s how we interpret a pitcher’s stats, what about hitters?
Studying
a hitter’s stats is not too disimilar, only reversed. If a pitcher wants to
keep batters off base, in the yard, and turning back to the dugout without
putting the ball in play, then a hitter wants to get on base, knock the ball out
of the yard, and put the ball in play before turning back to the dugout.
Translated: Onbase Percentage, Slugging percentage, fewer strikeouts per
atbat.
What
about batting average?
Batting
average is a good thing. It’s just that it doesn’t tell you enough. It tells
you the number of hits per AB, but it doesn’t tell you what kind of
hits. Are they all singles? Or are several extrabase hits included? Is Ichiro
Suzuki’s .372 batting average (with all but 14% of his hits singles) more
valuable than Barry Bonds’ .362 (55% extrabase hits including 45 homeruns)?
Would you rather have Albert Pujols batting .330 (with 99 extrabase hits) or
Juan Pierre batting the same (with only 37 extrabase hits in 86 more atbats)?
Slugging percentage tells you more about each hit than batting average does.
As we
discussed with pitchers, what is the other common way to reach base other than
hit? Draw walks. Onbase percentage is useful because the opposite of OBP is
virtually out percentage. Being out only 65% of the time (.350 OBP) is better
than being out 70% (.300 OBP).
Are
you saying batting average is meaningless?
No way! I
just want to know more about each plate appearance than a hit ratio. What kind
of hits are they? How often does he make an out? In fact, batting average is
still utilized because it is already incorporated into both OBP and SLG.
A .280
batting average contributes 80% of a .350 OBP. With a .500 SLG, a .300 batting
average makes up 60% of the SLG, representing all the singles the hitter gets.
The other .200 of the SLG is a result of his extrabase hits. Therefore, in the
increasingly popular onbase+slug statistic (OPS), batting average is counted
twice! So, looking at OBP and SLG are not eliminating batting average. It’s
already included! But OBP and SLG tell us more.
Is
there anything else about a hitter’s stats we should know?
I’m
glad you asked. We’ve already discussed both walks and strikeouts. But we
discussed them separately. Now let’s put them together.
Are
you saying strikeouttowalk ratio is important?
Well,
it’s not the ratio that is as important as it is each of them in relation to
atbats. Example: players A and B each have 500 AB. Player A has 100 strikeouts
and 100 walks; a 1.0to1.0 ratio. Player B has 35 strikeouts and 35 walks; a
1.0to1.0 ratio. They have the same strikeouttowalk ratio, but those numbers
tell us the two players have a drastically different approach to hitting. Player
A is patient at the plate yet takes big cuts. Think Jim Thome or Lance Berkman.
Player B is less likely to have the same kind of power stroke and may be more
unpatient at the plate so he doesn’t go deep into counts. Think Orlando
Cabrera or Endy Chavez. No one had better confuse Jim Thome for Endy Chavez
simply because they may have similar strikeouttowalk ratios!
Then
how do I best analyze strikeouts and walks?
We’ve
already started doing that with the Thome and Chavez example. We can interpret
from those numbers a hitter’s approach at the plate. Let’s start with walks.
A
hitter’s walk total tells us about his plate awareness. A good eye with a
patient approach will always draw the occasional base on balls. One walk per 10
atbats seems to be an adequate ratio. Anything less than that is cause for
concern. The more a hitter draws base on balls the more he shows that he not
only has a good eye at the plate but the patience to keep from swinging at balls
out of the strike zone.
A player,
then, who lacks the discipline to avoid swinging at bad pitches will, of course,
swing at bad pitches. His strikeout total will rise. If the strikeouttowalk
ratio is more than 2.0 or 2.5to1.0, we have another cause for concern. Put
our two causes for concern together (less than 1 bb per 10 ab and a
strikeouttowalk ratio worse than about 2.5to1.0) and we’re throwing up a
red flag as an alarm to stay away!
An
example of a red flag candidate is 30 walks and 110 strikeouts in 500 atbats.
Translation: he whiffs a lot with little patience or discipline at the plate.
Sure, he might hit some homers. But he had better! If he’s not hitting homers
he’s not doing much to help offensively.
Guys can be productive at the Major
League level with 30 walks and 110 strikeouts in 500 atbats! What gives?
Remember
that our emphasis is to judge whether a minor league player will go on to have a
productive Major League career. A player with strikeout and walk totals like
that in the minor leagues will get torn up, eaten apart and spat back out in the
bigs. Heck, even players who have solid strikeout and walk totals in the minors
will come to the big leagues and struggle to duplicate the same ratios.
For
examples, see Reggie Sanders (118 strikeouts and 33 walks in 446 atbats) and
Jason LaRue (10826 in 390). Sanders does provide offensive contributions with
homers and stolen bases despite his strikeouttowalk ratio. LaRue doesn’t do
as much as Sanders offensively but plays a more premium position defensively. In
the minors, LaRue had a strikeouttowalk ration of nearly 2to1 and Sanders’
ratio was 1.74to1. Against major league pitchers in 2004, they combined for
close to four strikeouts per base on balls. The Majors is a whole different game
and if a player doesn’t have a good approach against several pitchers who will
never see the big leagues, he’ll likely struggle against those who are already
there.
But there
are also exceptions when a player with a horrid strikeouttowalk ratio still
develops into a terrific hitter. See Alfonso Soriano. He provides MVPtype
numbers with a horrible strikeouttowalk ratio. But let me add this point: in
the 2003 World Series, Soriano was exposed for being the freeswinger that he is
and he had such a miserable series that he was eventually benched and
subsequently traded. Even his huge numbers could not hide his lack of discipline
at the plate. A disciplined approach at the plate is essential for a minor
leaguer to expect success in the Majors. If his approach is poor, he will
be exposed and the results are often poor.
What about power? How can we tell if a minor
leaguer will hit homeruns in the Majors?
Don’t
necessarily trust minor league homerun totals to predict power potential. Oddly
enough, guys with low homerun totals in the minor leagues can end up being
powerful Big League bashers. Power is typically the last tool to develop. And
most minor league bodies haven’t completely filled out yet. Even among those
with great power potential, minor league homerun totals may be quite modest.
Therefore, there must be an alternative to homeruns for predicting future power.
There is.
What
would a hitter need in order to realize homerun power when his body has filled
out and he’s experienced enough with the bat to incorporate a loft into his
swing? Answer: a good approach at the plate and a producer of hard line drives
that go for extra bases.
The best
indicators of potential future power are strikeouttowalk ratio and extrabase
hits. Guys that put up large double totals are candidates for future homerun
potential once he learns how to turn his gap power into overthefence power.
Couple that with a disciplined eye at the plate, and you’ve got a guy that is
willing to wait for his pitch and drill it.
Of
course, a lot has to do with his age. This is mostly true for younger players.
The older he becomes, the more he realizes his actual ability and there is less
wondering about his potential.
What does age have anything to do with a minor leaguer’s prospect
status? If he’s putting up good numbers, he’s going to be a good player,
right?
Age plays
a tremendous role in determining a player’s status as a prospect.
Imagine the following scenario involving three high school players who post identical
numbers: Player X is a senior playing for the varsity team and he is considered
a quality player. Player Y is a freshman also playing for the varsity team and
produces the same statistics. Player Z is a senior with identical performance
but he is playing for the junior varsity team. Who is the best player?
Player Y
would be considered the best of the three because he is producing at a quality
level exclusively against kids older than he is. If he’s already good against
juniors and seniors when he’s still just a freshman and it’s not expected of
him, how much better will he be by the time he is a senior? Player X is
producing at a quality level against those of his same age and level of
experience. Good for him! Player Z puts up the same numbers as the other two,
but he’s a senior doing it against sophomores and freshmen. He’s just a big
bully feasting on less developed players. It’s expected that he would produce.
Shame on him if he didn’t.
The same
is true in the minor leagues. Don’t be starstruck just because a minor leaguer
put up gaudy numbers. He could be a big bully playing against younger
competition.
So how
can I know if a minor leaguer is considered young or old for his level of
competition?
Each
minor league level has its typical range of age. There is TripleA, DoubleA,
two divisions of Class A (high and low), and several shortseason leagues. Since
TripleA is filled with so many veterans, the age range is extremely diverse. So
let’s start at the bottom and work our way up.
There are
three divisions of shortseason ball: Rookie leagues, Advanced Rookie leagues,
and shortseason Class A leagues (They are termed shortseason because the
leagues play only two months – July and August). Rookie leagues are typically
filled with former high schoolers who have recently signed since being drafted;
either just weeks before in June, or the year before. Therefore, most of the
players will be 18 or 19 years old. Also in the Rookie leagues are young
foreignsigned players like those from the Dominican Republic. Foreign born
players can sign a professional contract as young as age 16. A 17year old
having success in Rookie ball is promising. A 20year old is expected to do well
against teenagers.
Advanced
Rookie ball may also see several recent high school draftees, age 18 or 19, but
also should face some recent college draftees. Most of the collegians will be 21
or 22 years old.
Shortseason
A ball should exhibit several recent college draftees or players signed out of
high school who have one or two years of Rookie ball experience.
Therefore, the typical age range in shortseason A ball is 2022.
The
lowest level among fullseason leagues is low Class A featuring several players
between ages 21 and 23. A decent prospect should not be older than 22 if he is
still in low Class A ball. If a teenager puts up good numbers at this level,
look out! He could be dynamite.
A year
later, most of those players will be in high Class A ball. And of course
they’ll be a year older, typically ages 22 to 24. The 24year old still in
Aball is not likely to be making much noise as a prospect. Many of baseball’s
better prospects are already in the Major Leagues by age 24.
A player
who is in lowA at age 21 and is moving a level a year will reach DoubleA at
age 23. These days, DoubleA can sometimes be a stepping stone straight to the
Majors, even jumping TripleA. A 23year old producing in DoubleA still shows
good promise. A 21year old producing well in DoubleA should be a knockout!
Just ask the Marlins. Entering the 2003 season, Dontrelle Willis and Miguel
Cabrera started in DoubleA at ages 21 and 20, respectively. Before their season
ended they had helped lead their Parent club to a World Series title.
Using
another example while sticking with the Marlins theme, World Series MVP Josh
Beckett was promoted to DoubleA just a month after his 21^{st} birthday
in the 2001 season. Just three examples of what kind of core individuals it
takes to win a World Championship. All three were 21 or younger by the time they
reached DoubleA.
To
restate, the age range in TripleA is so broad it’s difficult to narrow down.
But let’s look at a typical path to reach the Majors for a midlevel prospect
who was drafted out of college. Selected in the June Draft as a junior and
subsequently signed, he could get immediate experience in a shortseason league
at age 21. In his first full professional season he will start in lowA ball at
age 22. The expectation for minor leaguers is to advance a level with each
season. Following that path, he’ll be 23 in highA ball, age 24 in DoubleA,
25 in TripleA and potentially get an opportunity to succeed in the Majors at
age 26. Anything older than that path should be considered below the bell curve
and less likely to be considered prospects at all. The more special a player is,
the younger he will be at each of those levels as he likely makes a rapid ascent
to the Major Leagues.
Therefore,
age plays a significant role in the prospect status of a minor leaguer.
Now
do I have everything I need to know in order to recognize a good minor league
prospect?
If we
knew everything there was to know in order to distinguish those with the
greatest promise from those of ordinary skill, we’d be working in Major League
front offices. But then again, they can’t even seem to figure it out all the
time. None of us really, truly ever know for sure. And that’s what makes it so
much fun.
