Yaz v. Manny
It was one of those conversations that starts up in front of the TV,
somewhere in the middle innings when the home team has the game in
hand, and we can let our concentration lapse a little.
Somebody said, “So, do you think Manny will go down as the greatest
Red Sox left fielder ever?” Three heads turned and six eyes looked
curiously at the speaker, whose name was Steve.
“After Williams, I mean.” Ah, okay.
The place was a small town in Massachusetts and the time was early
August 2008. Naturally, Manny Ramirez, who had been recently traded
to the Dodgers after spending nearly eight years in a Red Sox uniform,
was on people’s minds.
 |
 |
| Who was the greater Red Sox left fielder? (Ramirez photo by Icon/SMI) | |
“So, it comes down to Manny versus Yaz,” I said logically. “I
dunno, Manny has the offense, but he’s pretty bad in everything
else. Carl Yastrzemski, was a good defender, at least that’s what his reputation
is. What do you think, Dad? What’s your recollection of Yaz on
defense.”
The oldest of this little family gathering offered this: “Well, I
didn’t follow the American League that much back in those days, but
Yaz was known as an excellent defensive player. Rifle arm.”
The arm part made sense to me, since I had written an article called
“The Best Outfield Arms of Our Time” for the Hardball Times
Baseball Annual 2007 and Yastrzemski ranked as one of the top
two arms of the last fifty years, together with Clemente.
“But Manny is possibly the best right-handed bat in history.” Steve had the
floor again. “Plus, I don’t think he’s that bad on defense. He
makes some nice plays off the wall. And his own arm isn’t bad. Didn’t
he lead the league in assists a couple of years ago?”
Uh, oh. The Boston erstwhile 12-run (!) lead was now down to just two
runs. Our attention returned to the game and we saved the Yaz
vs. Manny debate for another day.
Hitting the ball
So, who was better, Manny or Yaz? Now, trying to figure out the
second-best Red Sox left fielder of all time might seem like an
esoteric exercise, but, hey, it’s the off-season and what else is
there to do? Besides, when I dig into a topic like this, I often turn
up something interesting. Before we dive into the numbers, just one
point: I’m going to compare Manny’s career to Yaz’s at the same age,
i.e., 36. Both came to the big leagues at age 21, so I’m going
to be looking at 16 years of big league play for each player. All
numbers shown below for Yaz refer to the period 1961-1976 (he retired after
the 1983 season).
Let’s tackle offensive production first, since it’s generally the most
important aspect for position players and the easiest to quantify. Here
are the offensive numbers for our two players:
G AB R H 2B 3B HR RBI BB SO BA OBP SLG SB CS OPS+
----+-----+----+----+---+---+---+----+----+----+-----+-----+-----+----+---+----+
Yaz 2421 8848 1402 2559 489 50 338 1343 1428 1118 .289 .387 .470 150 103 136
Manny 2103 7610 1444 2392 507 18 527 1725 1212 1667 .314 .411 .593 37 31 155
Obviously, Manny has the better raw offensive stats. Yaz played more
in those 16 seasons, in part because Ramirez did not become a full
time player until his third year, whereas as Yaz hit the ground
running, getting 583 at-bats as a 21-year-old. But durability seems to
be Yaz’s only edge here. Manny’s slash stats are superior across the
board and he destroys Yaz in home runs and RBIs. Oh, Yaz has an edge in
triples and stolen bases.
But what about context? Yaz’s prime came in the offense-suppressed
’60s, while the start of Manny’s career coincided exactly with the
offensive explosion that started in 1993. Or to put it another way,
the average American League team scored 4.0 runs per game during Yaz’s
career, but that number is 4.85 runs per game during Manny’s years.
Yaz produced fewer runs, but so did everybody else and it took fewer
runs to win ballgames in Yaz’s time.
The OPS+ number given above, though, takes context into account. It
adjusts for ballpark effects and the level of league offense and even
so, we see Manny with a clear advantage, 155 to 136. But OPS doesn’t
tell the whole story, so let’s look at a more comprehensive statistic
for measuring production: Batting Runs.
Batting Runs, also known as “Linear Weights” (a terrible name)
was developed by sabermetrician Pete Palmer back in the 1980s. It
estimates overall offensive production, taking into most things that a
player can do with a bat (not stolen bases, though). The version available at Baseball Reference, which takes into account park
effects and league offensive levels, is particularly useful for
our purposes. The following table compares Batting Runs for
Yastrzemski and Ramirez at the same age:
Batting Runs (Yaz through 1976) Age Yaz Manny 21 -8.5 -5.4 22 18.4 11.2 23 42.1 35.0 24 20.2 41.3 25 40.4 39.9 26 18.7 39.7 27 72.3 63.2 28 55.2 62.2 29 28.8 48.6 30 68.0 56.8 31 13.7 54.2 32 13.1 46.3 33 33.8 43.8 34 34.5 49.4 35 11.0 20.9 36 15.9 55.5 Total 477.6 662.7
So, Manny’s superiority in the raw stats is also evident in these
context-neutral Batting Runs, where he leads Yaz by nearly 200
runs. But, this isn’t yet the final word on batting, since we still
need to translate these runs into wins. Now, a reasonable rule of
thumb is that 10 extra runs is worth one extra win. But that rule
doesn’t really work in the low-scoring environment of Yaz’s prime.
That’s okay, though, Baseball Reference also provides Batting Wins,
taking into the effect of the run environment.
What you end up with is 49.6 Batting Wins for Yastrzemski and 60.4 for Ramirez,
an 11-win advantage for Manny.
Ok, that’s the easy part, evaluating the hitting contributions of
these guys. But, baseball is more than just swinging the bat. I now
turn to three other parts of the game that we can do a reasonable job
evaluating statistically: baserunning, defensive range and throwing
arm.
Baserunning
Proficiency on the basepaths is an aspect of offensive production that is not included in Batting Runs.
Neither of these players were fast and it shows in
their stolen base numbers. Yaz would run occasionally, although not
at a high success rate. In the period we are considering, he stole 150
bases in 253 attempts. Manny on the other hand rarely attempts a
straight steal, and many of his 68 attempts, 37 of which were
successful, were likely hit-and-run plays.
We can assign a run value to the stolen base (.22 runs) and the caught
stealing (-.38 runs) to find that these stolen base attempts amounted
to -.7 wins for Yaz and -.3 wins for Manny. But running the bases is
more than just attempting steals—there is the whole business
of taking the extra base on a hit or an out.
Although several people have worked on baserunning metrics, including
Dan Fox
(here
at THT and later at Baseball Prospectus) and Bill James in his
Baseball Annual, I was unable to locate any baserunning results for
Yaz (beyond stolen bases) and I haven’t seen full data for Manny,
either. So, I decided to analyze the Retrosheet data myself to get an
estimate of the baserunning skills of both of them.
I don’t have the space here to go into a detailed description of my method, perhaps
I’ll do that in a future article. The basic idea is to measure how
often a player goes first-to-third or second-to-home on a single,
first-to-home on a double or moves up a base on a flyout. An average run value
is assigned to each outcome (e.g. extra base taken, thrown out, etc.)
for each of these situations and a player is compared to a
league-average baserunner. The end result is a measure of baserunning
expressed in runs above average.
As an example, consider the following table, which shows the base
reached when the runner was on first base (with second base
unoccupied) and a single was hit:
Runner final destination, starting on 1B, Single +----------+------+-----------+-------+-------+-------+ | Player | Opps | ThrownOut | 2B | 3B | Home | +----------+------+-----------+-------+-------+-------+ | Yaz | 443 | 0.032 | 0.589 | 0.348 | 0.032 | | Manny | 314 | 0.016 | 0.707 | 0.268 | 0.010 | +----------+------+-----------+-------+-------+-------+
In this situation, Manny was more cautious than Yaz — he stopped
at second 71 percent of the time, while Yaz was content with second base only
59 percent of the time. However, Yaz was thrown out twice as often as
Manny. Interestingly, there’s a small chance of reaching home on this
play and Yaz did it three times more frequently than Manny.
However, we cannot compare Manny directly to Yaz, because over the
years runners have become less willing to take risks on the basepaths.
This is seen in the next table, which shows the results for the league
as a whole, for the Yaz years and the Manny years:
League averages, starting on 1B, Single +-------------+-------+-----------+-------+-------+-------+ | Era | Opps | ThrownOut | 2B | 3B | Home | +-------------+-------+-----------+-------+-------+-------+ | Yaz years | 81318 | 0.019 | 0.626 | 0.343 | 0.012 | | Manny years | 97858 | 0.013 | 0.675 | 0.300 | 0.012 | +-------------+-------+-----------+-------+-------+-------+
Here we can clearly see that in Yaz’s time players took more chances on
the bases, taking the extra base more often, but also getting thrown
out more frequently. So, we must compare Yaz and Manny not to each other,
but to an average baserunner during the era in which they played.
We end up with the following:
Baserunning runs +----------+-------------------+------+------+ | Player | Situation | Opps | Runs | +----------+-------------------+------+------+ | Yaz | runner 1B, single | 443 | 1.5 | | Yaz | runner 1B, double | 155 | -0.3 | | Yaz | runner 2B, single | 273 | 1.5 | | Yaz | runner 2B, flyout | 254 | -0.1 | | Yaz | runner 3B, flyout | 173 | -0.7 | +----------+-------------------+------+------+ | Yaz | Total | 1298 | 1.9 | +----------+-------------------+------+------+ | Manny | runner 1B, single | 314 | -2.0 | | Manny | runner 1B, double | 167 | -8.8 | | Manny | runner 2B, single | 277 | -5.4 | | Manny | runner 2B, flyout | 194 | 1.1 | | Manny | runner 3B, flyout | 147 | -2.9 | +----------+-------------------+------+------+ | Manny | Total | 1099 |-18.0 | +----------+-------------------+------+------+
We see that Yaz was about average on the basepaths, while Manny
was below average, but not at all terrible (remember, this is 16
seasons’ worth of baserunning). Adding in the stolen base runs and
converting to wins, we get for Yaz -.5 wins and for Manny -2 wins.
Frankly, I thought that Manny would fare worse in the baserunning
analysis. He isn’t fast and doesn’t look particularly good on the
basepaths. The fact that he also stands at home plate and watches his
long fly balls doesn’t really help his baserunning rep. It’s
noteworthy that my baserunning analysis does not (and cannot) measure
how often a single is stretched into a double (or a double “shrunk” to
a single).
In any case, we see here that baserunning has only a tiny effect on our evaluation
of these two players. This is generally true, although of course some
players are so good (or so poor) on the basepaths, that their
baserunning performance becomes important in their overall evaluation.
Where we stand
We’re about halfway done with this little exercise, so let’s take
stock. After accounting for batting and baserunning, here is
how our two heroes stack up:
Wins Above Average
Yaz Manny
------------------------------
Offense 49.6 60.4
Baserunning -0.5 -1.9
Range ?
Arm ?
------------------------------
Total so far 49.1 58.5
Manny has almost a 10-win lead and I think we can safely conclude that he was
the better offensive player. Next time we’ll tackle defense to see if
Yaz can close the gap.