The Most Democratic Pitcher
Don’t try to strike everybody out. Strikeouts are boring. Besides
that, they’re fascist. Throw some ground balls — it’s more democratic. — Crash Davis,
Bull Durham
On May 7 of this year, Chien-Ming Wang started for the Yankees against the Rangers
in a game played in Arlington. In typical fashion, Wang got three
groundball outs in the first inning and went on to beat the Rangers, 8-5. Wang ended up with the following line:
IP H R ER BB SO HR 6 7 3 3 0 2 0
Of his 16 outs on batted balls, 10 came on grounders, a typical number
for this groundball pitcher. The two strikeouts in six innings pitched was also
typical for Wang &mdash in fact, his groundball tendencies and low strikeout rate are Wang’s
two defining characteristics. Wang made what seemed like a slew of starts just
like this one: May 12 against Oakland: 8 IP, 0 runs, 0 K, win; May 27
against KC: 7 IP, 4 runs, 3 K, win; July 8 versus Tampa Bay: 8.1 IP, 1
run, 2 K, win; July 17 against the Mariners: 7 IP, 2 runs, 1 K, win;
and on and on and on. Pitching six or seven innings, a fair number of hits allowed, few
walks, very few strikeouts, even fewer home runs: ground balls and
wins — that was Wang’s MO in 2006.
Wang ended up as the most reliable pitcher on the Yankees staff this
year. I guess Mussina had stats that were just as good, but by the end
of the season, Wang was clearly the Yankees’ number one starter. You
will recall it was Wang that started Game 1 of the ALDS against the
Tigers (the only game the Yanks won, as it turned out). Chien-Ming
ended up the season with a record of 19-6 and an ERA of 3.63 in 218
innings pitched. The ERA was good enough for eighth in the American
League and he tied for the lead, with Johan Santana, in wins. However,
he struck only 76 batters, or 169 fewer than Santana. He was the only
pitcher in the top 10 ERA finishers to strike out fewer than 100
batters. Basically, it’s very, very hard to lead the league in wins
and strike out so few batters.
A Junk Stat: Strikeouts Per Win
I think I came across this stat the first (and only) time in a Rob
Neyer article back in the late ’90s. It’s a junk stat: not useful for
much, except for having fun.
Wang’s 2006 ratio of 4 strikeouts per win is extremely low. In fact, in the
last 15 years or so, not a single starting pitcher has
recorded such a low K/W ratio (min 10 game started, 10 wins). Here are
the top ten low-K/high-W seasons from 1990 to 2005:
Most Extreme Strikeout/Win Seasons Since 1990 +------------------+------+------+------+------+-------+------+-------+ | Name | Year | Team | G | GS | W-L | SO | K/Win | +------------------+------+------+------+------+-------+------+-------+ | Rueter, Kirk | 2003 | SFN | 27 | 27 | 10-5 | 41 | 4.10 | | Welch, Bob | 1992 | OAK | 20 | 20 | 11-7 | 47 | 4.27 | | Doherty, John | 1993 | DET | 32 | 31 | 14-11 | 63 | 4.50 | | Gullickson, Bill | 1991 | DET | 35 | 35 | 20-9 | 91 | 4.55 | | Gullickson, Bill | 1992 | DET | 34 | 34 | 14-13 | 64 | 4.57 | | Morgan, Mike | 1999 | TEX | 34 | 25 | 13-10 | 61 | 4.69 | | Welch, Bob | 1990 | OAK | 35 | 35 | 27-6 | 127 | 4.70 | | Robinson, Ron | 1990 | ML4 | 22 | 22 | 12-5 | 57 | 4.75 | | Halama, John | 2001 | SEA | 31 | 17 | 10-7 | 50 | 5.00 | | Tewksbury, Bob | 1990 | SLN | 28 | 20 | 10-9 | 50 | 5.00 | +------------------+------+------+------+------+-------+------+-------+
Note that a preponderance a these seasons is bunched up near the 1990
cutoff. That’s because strikeouts per plate appearance have been
growing steadily since about 1920. So, generally speaking, as you go further back in time, you
find more low-K pitchers. Wang’s 2006 would be at the top of
this table, of course.
The Low-K Pitchers
When you look at Wang throw, he doesn’t appear to be a very
low-strikeout kind of pitcher. He seems to have good “stuff”, throws
in the low-nineties, but, in fact, he only strikes out a bit over 3 batters per
game. That’s the lowest
K/G rate of any qualifying pitcher in baseball
in 2006. The surprising thing is that Wang
was so successful despite the low strikeout rate.
It’s useful to compare Wang to other low-K pitchers from 2006; here
are the 10 qualifying pitchers with the lowest K-rates:
Low Strikeout Pitchers of 2006 +------------+-------+------+------+------+------+------+ | Pitcher | IP | SO | K/G | ERA | HR/G | GB% | +------------+-------+------+------+------+------+------+ | Wang C | 218 | 76 | 3.31 | 3.63 | 0.52 | 62.8 | | Silva C | 180.1 | 70 | 3.38 | 5.95 | 1.84 | 43.6 | | Cook A | 212.2 | 92 | 3.94 | 4.24 | 0.73 | 57.8 | | Redman M | 167 | 76 | 4.03 | 5.71 | 1.01 | 44.4 | | Trachsel S | 164.2 | 79 | 4.21 | 4.99 | 1.23 | 41.5 | | Byrd P | 179 | 88 | 4.29 | 4.88 | 1.27 | 38.5 | | Marquis J | 194.1 | 96 | 4.33 | 6.03 | 1.58 | 42.9 | | Buehrle M | 204 | 98 | 4.39 | 4.99 | 1.61 | 44.2 | | Benson K | 183 | 88 | 4.42 | 4.82 | 1.66 | 41.3 | | Pineiro J | 165.2 | 87 | 4.53 | 6.37 | 1.20 | 47.5 | +------------+-------+------+------+------+------+------+
If you exclude Wang, this group had a composite ERA of 5.23 and no
other pitcher on the list comes close to Wang’s 3.63. Of course,
looking at the last two columns in the above table, we get an idea of
why Wang is outpitching his low-K counterparts: he gives up many fewer
home runs, which in turn is a consquence of his high groundball
percentage.
So, everything seems to make sense: Wang strikes out very few batters,
but he compensates for it by being very stingy with the home run. I
didn’t mention it before, but his control is very good, too: his walk
rate was only 2.2 per game. Everything looks pretty good.
But some people see the low strikeout rate and worry. They worry because
studies have shown that pitchers with low strikeout rates generally
have shorter, less successful careers than pitchers with high
K-rates. Bill James, in the New Historical Baseball Abstract, wrote a
lengthy article, entitled “Bird Thou Never Wert”, on the subject. The
title is a reference to Mark “The Bird” Fidrych,
who had a phenomenal
rookie year in 1976 (19-9, 2.34), despite striking out only 3.7
batters per game. Fidrych was struck by injuries after that season,
and did not reach 500 career innings pitched. James is saying that
Fidrych would likely not have had a long career even had he remained
healthy: the strikeout rate was just too low.
The Future of Chien-Ming Wang
So, what kind of career can we expect Wang to have? Well, one way of
answering this question, or at least thinking about the question, is
to look for similar players from baseball’s past and see how they
turned out. So that’s what I did.
I looked for primarily starting pitchers of the last 50 years or so
who 1) accumulated between 200 and 400 innings in their first two
years in the majors, and 2) had an
ERA+ between 100 and 135. Wang has logged about 330 IP with an
ERA+ of 117 in his first two years. I find 161 such pitchers and
for each of them, I
calculated their peripherals like K/9, HR/9 and the like. Actually,
since strikeout and home run rates vary quite bit over the period I’m
considering, I need to work with normalized stats. Normalized stats
are obtained by dividing the regular stat by the league average and
multiplying by 100. If the stat is supposed to be small (like ERA),
then we divide the stat into the league average. In this way, all
normalized stats have 100 as the average, anything above 100 is “better”
than average and anything below is “worse” than average. As an
example, a pitcher with a normalized K-rate of 110 strikes out 10%
more batters than the average pitcher.
So, for each of the 161 pitchers, I calculate the normalized versions
of strikeouts per nine innings, which I call K/9+. I also calculate
the normalized home run and walk rates, HR/9+ and BB/9+, and the G/F
ratio (not normalized). The following table shows the 10 pitchers with
the lowest (normalized) strikeout rate, along with Wang’s number so
you can compare:
Low-K Pitchers to be Compared to Wang +------------------+---------+-------+------+------+------+------+ | Name | Years | IP | ERA+ | K/9 | K/9+ | G/F | +------------------+---------+-------+------+------+------+------+ | Wang, Chien-Ming | 2005-06 | 334.3 | 117 | 3.31 | 52 | 1.78 | +------------------+---------+-------+------+------+------+------+ | Holt, Chris | 1996-97 | 214.3 | 112 | 3.99 | 58 | 1.55 | | Lamp, Dennis | 1977-78 | 253.7 | 112 | 3.02 | 58 | 1.91 | | Thurmond, Mark | 1983-84 | 294.0 | 124 | 3.24 | 58 | 1.21 | | Cocanower, Jaime | 1983-84 | 204.7 | 104 | 3.21 | 63 | 1.86 | | Ruhle, Vern | 1974-75 | 223.0 | 104 | 3.11 | 63 | 1.17 | | Lemongello, Mark | 1976-77 | 243.7 | 103 | 3.40 | 64 | 1.02 | | Bunker, Wally | 1963-64 | 218.0 | 124 | 4.00 | 65 | 0.91 | | Dunne, Mike | 1987-88 | 333.3 | 108 | 3.83 | 65 | 1.17 | | Straker, Les | 1987-88 | 237.0 | 106 | 3.76 | 65 | 1.01 | | Grimsley, Ross | 1971-72 | 359.0 | 100 | 3.66 | 66 | 0.75 | +------------------+---------+-------+------+------+------+------+
Wang’s normalized strikeout rate is lower than any of these guys. I
don’t know about you, but I don’t find these “comps” very inspiring, perhaps because
I hardly know who any of them are. Well, Ross Grimsley was an All Star
and Bill James wrote about Wally Bunker in the essay on DIPS in the
New Historical Abstract, and a couple other names are vaguely
familar, but I hope for Wang’s sake that he doesn’t end up on the
career similarity lists of guys like Cocanower, Straker or Lemongello.
But, strikeouts, as we’ve already discussed, isn’t the whole story
when it comes to Wang. We should also be looking at home runs allowed
and perhaps groundball to flyball ratio and probably walk rate,
also. Instead of running a table for each of those categories, I’m
going to combine all of them to give one “similarity score”. Actually,
I call my score “chi2” (pronounced chi-square, rhymes with high-chair)
and its value is small when two players are very similar. I used this
same technique when profiling
Jeff Francoeur some time ago. So, without further ado, here are
the 10 most similar pitchers to Chien-Ming Wang, based on their first
2 years of pitching:
Top 10 Most Similar Pitchers to Chien-Ming Wang +--------------------+---------+-------+------+------+-------+-------+------+-------+ | Name | Years | IP | ERA+ | K/9+ | BB/9+ | HR/9+ | G/F | chi2 | +--------------------+---------+-------+------+------+-------+-------+------+-------+ | Wang, Chien-Ming | 2005-06 | 334.1 | 117 | 52 | 120 | 196 | 1.78 | n/a | +--------------------+---------+-------+------+------+-------+-------+------+-------+ | Fontenot, Ray | 1983-84 | 266.7 | 109 | 75 | 113 | 233 | 1.86 | 3.43 | | Hibbard, Greg | 1989-90 | 348.3 | 120 | 68 | 135 | 191 | 1.15 | 4.17 | | Holt, Chris | 1996-97 | 214.3 | 112 | 58 | 129 | 136 | 1.55 | 4.18 | | Magrane, Joe | 1987-88 | 335.7 | 134 | 92 | 107 | 199 | 1.83 | 4.58 | | Ruffin, Bruce | 1986-87 | 351.0 | 115 | 69 | 113 | 148 | 2.07 | 4.61 | | Pichardo, Hipolito | 1992-93 | 308.7 | 108 | 67 | 117 | 155 | 1.27 | 5.21 | | Hamilton, Joey | 1994-95 | 313.0 | 134 | 81 | 135 | 139 | 1.47 | 5.56 | | Lawrence, Brian | 2001-02 | 324.7 | 108 | 94 | 144 | 147 | 1.59 | 5.85 | | Reuschel, Rick | 1972-73 | 366.0 | 131 | 114 | 147 | 175 | 1.66 | 7.59 | | Stieb, Dave | 1979-80 | 372.0 | 111 | 85 | 102 | 152 | 1.35 | 7.71 | +--------------------+---------+-------+------+------+-------+-------+------+-------+
All but one of these guys (Reuschel) had below-average K rates, they
all had plus control and were all stingy with the home run. Each one
was also a groundball pitcher, for the most part. Of course, the
method was designed to find pitchers like this, so it’s no
surprise. So, who is this Ray Fontenot guy, the most Wang-like of the
pitchers in our sample? Actually, there are some curious parallels between
Fontenot and Wang.
At the age of 25 Fontenot was called up mid-season by the Yankees
and made 15 starts; he won 8 games with an ERA+ of 118.
At the age of 25 Wang was called up mid-season by the Yankees
and made 17 starts; he won 8 games with an ERA+ of 111.
Fontenot pitched his second full season
with the Yankees, as did Wang. Current fans of the Yankees will hope
the similarity ends there, because Fontenot spent his third season pitching for
the Cubs, going 6-10 and seeing his ERA+ drop to 92. Thereafter banished to the
bullpen, Fontenot pitched one more year and then was out of
baseball. Will Wang follow a similar career path?
Well, Fontenot’s just one guy, you can’t make any predictions based on
that. Let’s have a look at the careers of these 10 most Wang-like pitchers:
Career Results for Top 10 Wang Comparables +--------------------+--------+------+------+------+------+ | Name | IP | W | L | ERA+ | chi2 | +--------------------+--------+------+------+------+------+ | Fontenot, Ray | 493.7 | 25 | 26 | 98 | 3.43 | | Hibbard, Greg | 990.0 | 57 | 50 | 99 | 4.17 | | Holt, Chris | 736.7 | 28 | 51 | 93 | 4.18 | | Magrane, Joe | 1096.7 | 57 | 67 | 103 | 4.58 | | Ruffin, Bruce | 1268.0 | 60 | 82 | 99 | 4.61 | | Pichardo, Hipolito | 769.7 | 50 | 44 | 105 | 5.21 | | Hamilton, Joey | 1340.7 | 74 | 73 | 94 | 5.56 | | Lawrence, Brian | 934.0 | 49 | 61 | 95 | 5.85 | | Reuschel, Rick | 3548.3 | 214 | 191 | 114 | 7.59 | | Stieb, Dave | 2895.3 | 176 | 137 | 122 | 7.71 | +--------------------+--------+------+------+------+------+
Reuschel and Stieb are not really very similar to Wang: note the jump of almost two in
chi2 between the 8th and 9th pitchers. (Oops, I probably should have excluded Lawrence,
since he is still active.) Anyway, only the not-very-similar Stieb and Big Daddy
managed to win 100 games in their career. Of the first eight
listed, 6 of them ended up with below average career ERAs.
At this
point, I would normally proceed by taking a group of players that I deem similar to
Wang, look at their average career and compare them to the others, the
un-Wang-like pitchers. However, the answer you get when you do this is
going to depend on who you include in the list of Wang comps. You can
see from the table above, that the 9th and 10th pitchers on this list
are going to change the results quite a bit, depending on which group
you put them in.
Well, let’s just plunge ahead and see what happens. Let’s take all
10 of these guys and declare them similar to Wang. If we then look
at the average career of these the group of these 10 (I’m going to
call them the Wangers) and compare them
to the average career of all the others, this is what we find:
Average Career of Wangers vs. Non-Wangers, Top 10 Comps +-------------+----------+------+---------+------+------+ | | Number | IP | ERA+ | W | L | +-------------+----------+------+---------+------+------+ | like Wang | 10 | 1407 | 102 | 79 | 78 | | unlike Wang | 151 | 1266 | 102 | 75 | 68 | +-------------+----------+------+---------+------+------+
We see no real difference between the two groups, if anything the
Wangers have a slightly longer career on average. This would indicate that
Wang’s profile doesn’t really impact how long or successful his career
might be.
On the other hand, if you decide that Reuschel and Stieb shouldn’t be in the Wang
group, you get a different result: namely that the Wangers have shorter and less successful
careers than the non-Wangers. Here are the numbers including eight
pitchers in the Wang-like group:
Average Career of Wangers vs. Non-Wangers, Top 8 Comps +-------------+----------+------+---------+------+------+ | | Number | IP | ERA+ | W | L | +-------------+----------+------+---------+------+------+ | like Wang | 8 | 954 | 98 | 50 | 57 | | unlike Wang | 153 | 1292 | 103 | 77 | 70 | +-------------+----------+------+---------+------+------+
So, the answer you get depends on where you draw the line. It’s not very satisfying, I admit, but that’s
the way it is.
The Other Extreme
Here are the 10 pitchers in the sample who were least like Chien-Ming Wang:
Ten Worst Comps for Wang +-------------------+---------+-------+------+------+-------+-------+------+-------+ | Name | Years | IP | ERA+ | K/9+ | BB/9+ | HR/9+ | G/F | chi2 | +-------------------+---------+-------+------+------+-------+-------+------+-------+ | Eckersley, Dennis | 1975-76 | 386.0 | 120 | 170 | 85 | 99 | 0.47 | 54.57 | | D'Acquisto, John | 1973-74 | 242.7 | 102 | 141 | 66 | 107 | 0.57 | 43.67 | | Bibby, Jim | 1972-73 | 236.7 | 100 | 142 | 63 | 104 | 0.73 | 41.88 | | Sanderson, Scott | 1978-79 | 229.0 | 114 | 144 | 109 | 96 | 0.49 | 41.65 | | Lemaster, Denny | 1962-63 | 323.7 | 111 | 124 | 90 | 69 | 0.61 | 41.20 | | Johnson, Bob | 1969-70 | 215.7 | 123 | 153 | 101 | 120 | 0.59 | 39.31 | | Bere, Jason | 1993-94 | 284.3 | 122 | 137 | 72 | 111 | 0.65 | 38.73 | | Smoltz, John | 1988-89 | 272.0 | 103 | 117 | 91 | 83 | 0.51 | 38.34 | | Montefusco, John | 1974-75 | 283.0 | 121 | 156 | 104 | 144 | 0.56 | 37.81 | | Benes, Andy | 1989-90 | 259.0 | 105 | 124 | 93 | 88 | 0.55 | 37.80 | +-------------------+---------+-------+------+------+-------+-------+------+-------+
These were flyball pitchers who punched out a ton of batters, but also walked quite a few and gave
up there share of home runs. Interesting to see Eckersley at the top of the list: we remember him, of course, as the one-inning
relief specialist, he was quite the high-octane starting pitcher when he came up as a 20-year-old in 1975. The career stats of these
Un-Wangs is quite a bit better than the Wangers:
Career Stats of Un-Wang-Like Pitchers +-------------------+--------+------+------+------+-------+ | Name | IP | W | L | ERA+ | chi2 | +-------------------+--------+------+------+------+-------+ | Eckersley, Dennis | 3285.7 | 197 | 171 | 116 | 54.57 | | D'Acquisto, John | 779.7 | 34 | 51 | 80 | 43.67 | | Bibby, Jim | 1722.7 | 111 | 101 | 99 | 41.88 | | Sanderson, Scott | 2561.7 | 163 | 143 | 102 | 41.65 | | Lemaster, Denny | 1787.7 | 90 | 105 | 96 | 41.20 | | Johnson, Bob | 692.3 | 28 | 34 | 103 | 39.31 | | Bere, Jason | 1111.0 | 71 | 65 | 86 | 38.73 | | Smoltz, John | 2929.3 | 177 | 128 | 126 | 38.34 | | Montefusco, John | 1652.3 | 90 | 83 | 103 | 37.81 | | Benes, Andy | 2505.3 | 155 | 139 | 104 | 37.80 | +-------------------+--------+------+------+------+-------+
Final Thoughts
As I mentioned in my article on Francoeur, I’m not totally convinced that the similarity method actually works. It might work,
but then again it might not. But it’s a fun analysis, it conjures up players from the past, maybe guys you haven’t thought about
in a while (like Big Daddy Reuschel). And perhaps it does tell us something about what we can expect from a guy like Wang going
forward. No, I can’t draw any firm conclusions from the numbers, but it seems to me that the cards are stacked ever-so-slightly against
Wang having a 100-win career. Wait, let me clarify: the cards are stacked against any pitcher having a 100-win career: only 42 of the 161
pitchers in our sample won 100 games or more. And these guys all looked promising after two seasons in the majors.
But, it looks to me that the odds are a bit worse for a guy
like Wang who strikes out so few batters. We’ll see, I guess.
References & Resources
-
I tend to be a believer in DIPS, so I did not include hit rates in
my comparison. I have tried doing so, though, and the basic results
don’t change. Only one player in the top 10 changes: Reuschel drops out and some guy named
Randy Tomlin takes his place. - G/F Ratio — my G/F ratio is slightly different than others you’ll come across. That is because full groundball/flyball
info is not available going back fifty years. So, I had to devise my own version using Retrosheet
data. It certainly is good enough for this analysis. - The details of the chi2 calculation were given in my article on Francoeur here. Of course, the batting statistics used there have been replaced by
the pitching statistics discussed above.