Schrodinger’s Bat: Wily Mo Redux

“His power is unbelievable. He’s like a ‘Little Papi,’ and I know that when his time comes, he will be a good player. Nobody hits the ball like Mo–nobody.”
—David Ortiz on his now former teammate

In my inaugural column in this space back in March of 2006, I took a look at the prospects of outfielder Wily Mo Pena. Days before, he had been traded from the Reds to the Red Sox for Bronson Arroyo. Several times since then, readers have asked whether I had anything else to add given his 2006-2007 performances. So, at long last, today we’ll take another look at the fortunes and prospects of Wily Mo.

Catching Up

Many thought that 2007 would be Pena’s breakout year. After sustaining a fractured hamate bone which put him on the shelf for two months in 2006, and a strained left quadriceps that held him out of action for a month in 2005, the hope was that he could build on his career highs in OBP and batting average (.301/348/.487). This hope came despite the fact that those gains were largely fueled by a .411 batting average on balls in play.

Matters changed a bit with the acquisition of J.D. Drew over the winter. the plan was to use Pena against left-handed starters, and spot start him against right-handers. The assumption may also have been that he could come off the bench against the occasional situational lefty, but alas, as Nate Silver commented, this might not show Pena’s best side. In any case, we know what often happens to the best laid plans and Pena got off to a rough start; he hit .172 in April while striking out in more than half of his at-bats. He rebounded in May–although he hit only one home run and two doubles–but then slumped again in June and July, hitting a combined .220/.272/.378. This all but sealed his fate in Boston; he became expendable so that the Sox could add a starting pitcher. The trade rumors that swirled around Pena primarily involved his being dealt to the White Sox in a package that would net Jermaine Dye. When that fell through at the trade deadline, Pena cleared trade waivers and was finally dealt to the Nationals on August 17th, where Jim Bowden, apparently trying to corner the market on former Reds outfielders, picked him up for a player to be named later; the Sox also picked up some cash.

Pena left town graciously, and with Washington has continued to exhibit those same skills which make him at once both promising and maddening. He homered in four of his first ten games, but overall has collected only six other hits while striking out fourteen times. I was reminded to take up the topic of Pena again after seeing this combination in action during the Nats’ three-game series against the Rockies August 24-26. After collecting a single off Jeff Francis and victimizing Taylor Buchholz with a homer on the 24th, he struck out in both his first two trips on the 25th while facing rookie right-hander Ubaldo Jimenez. In both cases, Pena fell behind in the count 1-2 by taking a strike and fouling off a pitch before chasing sliders low and away. The second strikeout, using PITCHf/x, looked something like this from Pena’s perspective:

The first pitch was a 76 mph curve that missed high and inside, the second an 85 mph slider on the inside corner, the third a 96 mph fastball low and in, and the fourth was another slider low and away clocked at 89 mph. Of course, this at-bat highlights what for Pena has been his two biggest problems. For starters, he’s had trouble finding success against right-handed pitchers. Over the course of his career his splits have been .270/.341/.467 against lefties, and .247/.294/.471 against righties. Closely related is his difficulty in hitting breaking balls, as testified to by ESPN’s Inside Edge, which has him hitting .306 and .333 on fastballs against right- and left-handed pitchers, respectively, while hitting .186/.216 against curves and .192/.194 against sliders.

To get a feel for how pitchers work Pena we can take a look at the following two charts of the location of fastballs (FB), breaking balls (BB, which would include sliders, curveballs, and some cutters), and changeups (CH) thrown by right-handers and southpaws:

As you would expect, right-handed pitchers generally try and work Pena down and away with breaking balls and up with fastballs. Southpaws, on the other hand, try and keep the fastball in, apparently trying to keep him from extending his arms. The following table shows the mix of pitches by count, which again illustrates how right-handers feed him a steady diet of breaking balls:

Versus Righties               |  Versus Lefties
Count     FB     BB     CH    | Count     FB     BB     CH
1-0       50%    50%     0%   |  1-0      79%    18%     4%
2-1       40%    60%     0%   |  2-1      61%    33%     6%
2-2       57%    29%    14%   |  2-2      50%    50%     0%
3-0      100%     0%     0%   |  3-0      47%    24%    29%
3-1        0%   100%     0%   |  3-1      46%    54%     0%
3-2       50%    33%    17%   |  3-2      67%    17%    17%
0-0       43%    45%    11%   |  0-0      40%    30%    30%
0-1       33%    43%    24%   |  0-1      60%    40%     0%
0-2       20%    40%    40%   |  0-2      30%    60%    10%
1-1       40%    55%     5%   |  1-1     100%     0%     0%
1-2       36%    43%    21%   |  1-2      33%    67%     0%
2-0       50%    50%     0%   |  2-0      71%    14%    14%
Total     41%    46%    13%   | Total     57%    32%    11%

Looking Forward

For those unfamiliar with the analysis I did at the start of the 2006 season, the crux of it was a comparison of players of similar age, opportunity, and power (since Theo Epstein had drawn the correlation between power at a young age and future performance in Pena’s case), and included a look at how their careers developed. What I found was that the players clustered around Pena, on average, went on to have careers that spanned roughly 15 years and were very productive at least into their early 30s. At the top of that comparable list were Willie Stargell, Tony Perez, Andre Dawson, and Jim Rice; at the bottom Nick Esasky, Pete Incaviglia, Dave Roberts (the 1970s version), and Cory Snyder, with a good median projection being someone like Jesse Barfield.

The huge caveat to Pena turning into Stargell or even Barfield that we discussed was Pena’s strikeout rate. Using the 14 players who were most comparable to Pena based on strikeout-to-walk ratio, we found (with an accompanying graph, of course) that their average strikeout rate decreased from around 23 percent of plate appearances to about 18 percent from their age-23 through age-25 seasons, and thereafter held fairly steady. Being successful and getting to those 15-year careers appears to be correlated with reducing the frequency with which you strike out. As with pitchers who have what I termed “old-pitcher skills,” hitters who strike out frequently have less room for error, so they must rely more heavily than other players on the other two of the three true outcomes (home runs and walks) in order to contribute. When those start to slip, their production bottoms out in a hurry.

Where Pena is concerned, two important points follow:

1. Pena’s strikeout rate started at a much higher level than even the fourteen comparables, at around 30 percent of plate appearances at ages 21 and 22.
2. His combined rate in 2005 and 2006 was around 32 percent in his age-23 and -24 seasons, and this season it sits at almost 33 percent, albeit with limited playing time. This is not a good trend.

This new information forces us to put Pena in a different category, so now we’ll run a new set of comparables that includes players since 1955 who at ages 24 or 25 struck out in 30 percent or more of their plate appearances while racking up over 300 at-bats. It’s not a very long list, and those meeting the criteria sorted by year are shown in the following table:

Name              Year  Age    PA   AVG   OBP   SLG   OPS   ISO  BB/PA SO/PA
Dave Nicholson    1964   24   350  .204  .329  .364   693  .160  .149  .360
Dave Kingman      1973   24   351  .203  .299  .479   778  .275  .108  .348
Dave Kingman      1974   25   393  .223  .300  .440   740  .217  .089  .318
Rob Deer          1986   25   546  .232  .335  .494   829  .262  .126  .328
Bo Jackson        1987   24   434  .235  .295  .455   749  .220  .069  .364
Bo Jackson        1988   25   468  .246  .286  .472   758  .226  .041  .312
Pete Incaviglia   1988   24   467  .249  .321  .467   788  .218  .077  .328
Tony Clark        1996   24   411  .250  .299  .503   802  .253  .068  .309
Melvin Nieves     1996   24   484  .246  .322  .485   807  .239  .087  .326
Melvin Nieves     1997   25   405  .228  .311  .451   762  .223  .081  .388
Ruben Rivera      1999   25   475  .195  .295  .406   701  .212  .114  .301
Russell Branyan   2001   25   361  .232  .316  .486   802  .254  .102  .366
Jared Sandberg    2002   24   401  .229  .304  .444   748  .215  .090  .347

This list of ten players is obviously different than the previous one, and collectively has a much lower horizon, with an average career length of 10 years. The graph below illustrates the group trends in walks, strike outs, and playing time:

From the graph, we can divine some good news for Wily Mo, as the collective strikeout rates dropped from over 30 percent at age 24 to down around 25 percent by age 27, and then essentially stabilized. On the other hand, walk rates remained relatively constant throughout, while playing time peaked at age 26, then descended rather precipitously: only four of the ten were still in the majors at age 33, and only two at age 35.

Finally, we’ll take a look at the aggregate production of these ten using OPS normalized for park and league (NOPS/PF) by age.

When most of the group is intact at age 27, the NOPS/PF (shown in red) reaches its peak at 108 (eight percent above average), and then as players are selected out of the league we see spikes at ages 29, 31, 33, and 35. To correct for the selection bias, the black dotted line depicts the performance curve when crediting the players who were selected out with replacement-level performances.

So what does all of this tell us about Pena? Unfortunately for Nationals fans, the end result is that unlike our forecast in the spring of 2006, when the upside for Pena seemed to be a Barfield-type career, the projection now approximates something more like a cross between Pete Incaviglia and Tony Clark at the high end, with a less tasty mix of Melvin Nieves and Ruben Rivera representing the low end. In other words, he could be contributor on a good team, but likely not a player you build around.