Schrodinger’s Bat: In the Arms of an Angel

“[Angel Berroa] is still going to be a heck of a player. [Andres] Blanco has got a long way to go even before he considers himself in Angel Berroa’s category.”
—Royals Manager Buddy Bell in September of 2006

Friday March 23rd likely spelled the end of the Angel Berroa era in Kansas City. Historically, the Royals are not exactly known for quality at that position–it has featured such luminaries as Jackie Hernandez, Freddie Patek, U.L. Washington, Onix Concepcion, Angel Salazar, Kurt Stillwell, David Howard, Greg Gagne, Jay Bell, Mendy Lopez, Rey Sanchez, and Neifi Perez before Berroa. Nevertheless, Berroa’s decline and fall, leading to the trade for (ironically) Tony Pena Jr. and Berroa’s subsequent demotion to Omaha, has been especially depressing for Royals fans for a number of reasons.

Once upon a time Berroa was the bright spot in the Johnny Damon trade of January of 2001, a deal that also moved A.J. Hinch to KC (and put Ben Grieve in Tampa Bay). Berroa immediately justified the optimism of then-GM Allard Baird by putting up a good showing in High-A and Double-A, netting 60 extra base hits as a 21-year-old. Yes, he needed to develop some plate discipline, and yes, he was still erratic with the glove, but he’d made some strides in his patience at Wichita and cut his error total from 54 in 2000 to “just” 30 in his two minor league stops. His performance earned him Baseball America’s First Team Minor League All-Star shortstop and the Royals’ Minor League Player of the Year honors. The raw tools–the power, range, and arm–were all there and simply needed honing.

But 2002 would prove to be a big disappointment. A knee injury cost him the first two months of the season and when he returned he hit an appalling .215/.277/.360 in 77 games in Omaha and no better in 20 games in Kansas City. And if that weren’t enough, he was one of a number of foreign-born players in the wake of 9/11 to find themselves suddenly several years older. Just like that, he was a 24-year old who had yet to show he could hit Triple-A pitching. The injury also reportedly restricted his range in the field, although his raw chances don’t show a steep decline.

We all know what came next. Berroa parlayed an injury-free season and a surprising Royals showing into a four-vote margin over Hideki Matsui to grab Rookie of the Year honors in the American League. His .287/.338/.451 league-average performance (his OBP boosted by 18 HBPs) and adequate defense added up to a 7.8 WARP. His plate discipline wasn’t anything to brag about, but he did see an almost-league-average 3.68 pitches per plate appearance.

Still, there were concerns. His offensive performance was a bit out of line with his minor league track record, and his defense wasn’t necessarily all it was cracked up to be. Despite looking good in metrics based on aggregate statistics such as range factor, his defense was actually below average, and John Dewan’s original Plus/Minus system had him making 13 fewer plays than would have been expected. He did especially poorly on balls to his right; Dewan’s system ended up ranking him 33^rd defensively among big league shortstops. But as after 2001, there were things on which to build going into 2004.

It was clear that the Royals thought they had their shortstop of the present and future when, despite suffering from undiagnosed migraines that necessitated a spinal tap to rule out meningitis in mid April, they signed him to a four-year, $11 million contract with a club option for 2009. Berroa responded to the good news by slumping badly, and getting himself demoted from his leadoff spot to eighth in the order.

Berroa caught fire in June, but after again slumping in July he was sent to Wichita to work on his deficiencies. Apparently sufficiently chastened, he was brought back two weeks later, and he finished the season strong to raise his batting average to .263 and his on-base percentage to .308, tiptoeing above replacement level. His power was clearly diminished, but the team was impressed with how he handled the demotion, and charitably chalked up much of his decline to dealing with his early-season illness.

However, his line-drive rate fell from a healthy 21.3 percent in 2003 to 18.7% in 2004, while both his ground ball and popup rate increased, and just as problematic was that his pitches per plate appearance decreased to 3.58. On defense, things were no better. His aggregate stats caught up with his Plus/Minus rating as he was credited with -20 this time; he was once again ranked 33^rd.

After Berroa recovered from a bruised left knee suffered in spring training 2005, Allard Baird was again confident that he could do the job and become the shortstop this team needed him to be. When asked about the possible switch of Andres Blanco to second base, Baird noted “I’ve got a shortstop here who is a pretty good shortstop.” Shortly thereafter, Berroa seemed to find his groove, and in May of 2005 he reeled off an eleven-game hitting streak, and he was once again in the leadoff spot.

It didn’t last. He struggled in June, and as July dawned had driven in just 18 runs in over 300 plate appearances. Most disappointing was his hacktastic 3.35 pitches per plate appearance and his flailing after offspeed pitches, a trend that caused that caused his line-drive rate to fall to 17.8 percent, while his pop-up percentage went up to almost 10 percent. In the field he contributed to the worst defense in baseball by making 26 fewer plays than would have been expected. With over 1,700 major league plate appearances under his belt, two consecutive replacement-level seasons, three consecutive seasons at the bottom of heap at his position defensively, and with his age-27 season now behind him, his promise was all but gone.

Nevertheless, not even the spring of 2006 would be his undoing with the Royals organization. Given the investment they’d already made and the fact that there were simply no other options in the system that promised anything better, Berroa was a secure incumbent. Even though his perennial decline continued (.234/.259/.333 equating to 17.4 runs less than replacement level, a 16.3 percent line-drive rate, 3.37 pitches per plate appearance), Dayton Moore was still saying as late as last September that “(t)here are tools there, and I’ve learned that you stay with tools. You have to be patient with players, but player have to produce.” Well, tools or not, twelve strikeouts in his first 26 at-bats in spring training this year stretched Moore’s patience to the breaking point.

Even with $8.5 million still owed on that contract, the response by the organization was a trade for a 26-year-old shortstop with a .285 career minor league OBP who will be lucky to approximate Berroa’s offensive output (despite enjoying a fine spring with the bat that he carried into opening day) albeit with better defense. Sounds suspiciously like Andres Blanco. And to get that player, they gave up a pitcher in Erik Cordier, who although he’ll have to sit out 2007 recovering from Tommy John, could still be a major league contributor down the road. That’s throwing in the towel in style.

Some players don’t live up to expectations, and others flame out early after an initial flowering. The case of Angel Berroa is a strange one. Did the illness in the spring of 2004 cause a permanent change in approach? Did he put too much pressure on himself after signing the big contract? Is he simply a player who can’t or won’t make adjustments? Or is it some combination of physical, mental, and emotional elements all wrapped up together? Whatever the case, it appears for now that the Royals have stopped looking for answers.

A War of Attrition

As regular readers know, one of the features of our PECOTA forecasts are the Breakout, Improve, Collapse, and Attrition percentages. Attrition Rate measures the probability that a hitter’s plate appearances or a pitcher’s opposing batters faced will decrease by at least 50% in the coming year relative to his forecasted playing time. And while it is a good indicator of injury risk, it also may reflect decreased playing time due to poor performance and managerial decisions. What you may not know is that there is also a Drop Rate, which is the percent chance that a player will not receive any plate appearances or innings pitched in the future (the current year is always zero percent) based on their comparables who disappear from the ranks of major league players entirely.

In the world of major league baseball there is constant selection pressure being applied to players, akin to the natural selection that all organisms feel to varying degrees in the natural world. While in the natural world a species must enjoy differential reproductive success to survive and overcome that pressure, in the baseball world a player must usually consistently perform above a certain level in order avoid being selected out of the league, or at least out of a regular job. The Attrition and Drop Rates are measures of that pressure.

Given that Berroa’s performance had declined so precipitously since his rookie campaign, it’s no wonder that his Attrition Rate was a dangerously high 29% going into spring training, and that his drop rate was 15 percent for 2008, rising to 53% by 2011. I say “was” because his demotion to Omaha effectively shoots his Attrition percentage close to 100 percent, as reflected in the updated projected playing time for the Royals.

But in pondering this fallen prospect, I wondered where Berroa’s decline (at least offensively) ranks from a historical perspective, and whether other players survived similar selection pressure to go on to have productive, or at least lengthy, major league careers.

To examine this question I created a data set that included all players who came to the plate at least 400 times (as a measure of a player who held down a regular job) between the ages of 24 and 30, and who had their offensive output decline for three consecutive years. This age range was chosen since one of the most interesting aspects of Berroa’s decline is that it came between the ages of 25 (2003) and 28 (2006), at a time where the normal performance trajectory for all players is in its steepest upward slope. As you might imagine, an older player might very well see a decline of this type but yet be less likely to be selected for significantly less playing time because of their longer track record. In order to approximate offensive output I’m using the fairly quick and dirty method of calculating a park-adjusted Normalized OPS (the ratio of the player’s OPS to the league OPS) referred to as NOPS/PF so we can stretch back to 1901.

After running the numbers, there were 96 instances of a player suffering three consecutive years of decline which encompassed 85 distinct players (the complete list can be found on my blog). The longest of those declines was actually three consecutive ones that add up to a five-year span by Jimmie Foxx from ages 24 through 29 (1932 through 1937), when his NOPS/PF values were 160, 157, 150, 147, and 133. His performance then rebounded in 1938 and 1939, when he put up values of 147 and 150 respectively. Similarly Jimmy Williams (1901-1905), Ty Cobb (1912-1916), Tito Francona (1959-1963), Richie Hebner (1972-1976), Steve Kemp (1979-1983), Willie Upshaw (1983-1987), Greg Jefferies (1993-1997), Raul Mondesi (1997-2001), and Alex Rodriguez (2000-2004) all experienced four consecutive years of decline. The steepest three-year falloff was that of Andres Galarraga, who averaged an 11% drop in production from 1988 through 1991 when his NOPS/PF went from 130 to just 88 in his final year with the Expos. Kevin Mitchell (1989-1992), Chick Galloway (1922-1925), and Bobby Murcer (1971-1974) all declined more than 9 percent per year as well. On the other end of the spectrum, Kirk Gibson makes the list, although was actually very consistent from his age-27 through age-30 seasons, with values of 123, 122, 117, and 116, registering just a 0.9% decline per year. Overall, the players on the list averaged a 5.8% decline over the three-year span.

What is most interesting about these examples and most of the other declines (which include players like Lou Gehrig, Tris Speaker, Eddie Mathews, and Duke Snider, to name a few) is that they began from such a high level. To illustrate the difference between the average player who experienced such a decline and Berroa, consider the following graph.

Here we can see that the average player on this list started with production that was almost 25 percent greater than league average, and even after three years of decline, was still four percent better than their league. Angel Berroa, on the other hand, began in 2003 with production that was slightly below league-average, and after experiencing more modest declines in 2004 and 2005, sank to rock bottom in 2006 at just 73% of league average with a -17.4 VORP. Obviously players who start at a higher plateau do not experience the selection pressure of those who start lower, since even in decline they are typically better than league average, and because their greater output provides them a larger cushion in terms of future opportunities.

When we restrict the list to players along the same plane as Berroa, we find the following players who began their decline with an NOPS/PF of less than 100 as shown in the following table.

                   Yr1                   Yr2         Yr3         Yr4
Name             Age   Year   PA  NOPS    PA  NOPS    PA  NOPS    PA  NOPS Future PA
Rafael Ramirez    24   1982  669    99   668    98   629    87   595    85    2810
Angel Berroa      25   2003  635    98   554    92   652    90   503    73
Brian Hunter      25   1996  553    94   738    90   636    82   527    73     768
Bud Harrelson     25   1969  457    93   682    92   618    91   490    86    2077

That’s it. From 1901 through 2006 only three other players posted three consecutive declining offensive seasons while retaining their jobs:

• Not coincidentally Rafael Ramirez is listed as Berroa’s highest comparable, with a Score of 67. Ramirez enjoyed a nice 1982 campaign with the Braves, and essentially repeated that performance in 1983. Like Berroa he had his problems defensively, although more so with catching the ball (38 errors in 1982 and 39 in 1983) than with his range. In 1984 his output declined as his walk rate and slugging percentage went south. In his age-27 season in 1985 he duplicated his 1984 production, marking a third consecutive year of decline. He would tread water for his final two years in Atlanta before being traded to the Astros in 1988, where he would enjoy his best offensive season. All told, he played seven seasons following the decline, performing (except for 1988) at his 1984 level while accumulating almost 3,000 plate appearances in that time.
• Mets shortstop Bud Harrelson started out worse relative to his league than any of the four players, and then declined an average of just 1.9 percent per year. He also owned the best defensive reputation (he won a Gold Glove in 1971 and was an All-Star in both 1970 and 1971) and by far the best eye of the bunch, walking 95 times in 1970 (Year Two in his line). After a 1972 season in which he would hit .215/.313/.266, he would rebound to his pre-decline level and play for eight more seasons, and accumulate just over 2,000 additional plate appearances.
• Brian L. Hunter received exactly one vote for Rookie of the Year in 1995 with the Astros on the strength of his .302 average and 24 stolen bases. It was pretty much downhill from there. In 1996 he would post a 92/17 strikeout-to-walk ratio while hitting .276 and stealing 35 bases. After being traded to the Tigers in a nine-player deal over the winter, he developed a little plate discipline and stole 74 bases. Still, in his new league and park his NOPS/PF fell a few points. That decline continued in 1998 with the Tigers, and again in 1999 when he spent most of the season with the Mariners. He would go on to play just four more seasons and come to the plate fewer than 800 times.

It shouldn’t be surprising that three of the four players are shortstops, since by virtue of it being a more demanding defensive position it puts less pressure on a player’s offense. The other player continued to be afforded playing time because he played a demanding position (center field) and because he had another skill (stolen bases) that, while certainly overvalued, was sought after.

If you’re an Angel Berroa fan, this is not an encouraging outcome, although if you’re an Angel Berroa fan, you are perhaps by definition used to discouraging outcomes. What history tells us is that it would be almost unprecedented if Berroa were to play as many as a handful of additional seasons, and will almost certainly never be a regular player again.

On only a couple of occasions has anyone survived the selection pressure at nearly the same performance trend that Berroa has put up. And in both those cases the shortstops were clearly better defenders. What is truly remarkable is that Berroa garnered as much playing time as he did for as long as he did. Players who consistently perform as poorly as he has are usually selected out of a regular job in a ruthlessly efficient manner. Roger Metzger, Frank Duffy, Ivan DeJesus, and Jose Lind are examples of players who lost their jobs after two years of similar decline, and a lesser percentage of players rebound and go on to play a little longer in a backup role (guys like Jose Pagan and Dale Berra).

In the case of Berroa, what preserved him was the perfect storm of a financially strapped organization, little depth in the system at the position, no pressing need to upgrade in order to be competitive, and an ill-advised contract. All of these factors contributed to this unhappy result, and to what appears to be this sorry end.

Special thanks to long-suffering Royals fan Ron Hostetter for planting the idea for this column.