Here’s a pair of hypothetical line-ups:
Line-Up A Line-Up B Roy Campanella, C Ron Hassey, C Orlando Cepeda, 1B Willie Aikens, 1B Joe Morgan, 2B Steve Lombardozzi, 2B Mike Schmidt, 3B Luis Salazar, 3B Rogers Hornsby, SS Spike Owen, SS Ted Williams, OF Dan Iorg, OF Willie Mays, OF Jeff Leonard, OF Ty Cobb, OF Larry Herndon, OF
OK, so Hornsby played only 356 career games at shortstop, but you’d squeeze him in somewhere if you had him and Morgan. Given a choice between these two line-ups, which would 99.9 percent of managers select?
Does this affect your answer?
Line-Up A PA Avg OBP Slg --------------------------------- Campanella 126 .237 .310 .386 Cepeda 90 .207 .233 .368 Morgan 219 .182 .324 .348 Schmidt 155 .236 .310 .386 Hornsby 52 .245 .288 .327 Williams 30 .200 .333 .200 Mays 99 .247 .323 .337 Cobb 68 .262 .294 .354
Line-Up B PA Avg OBP Slg --------------------------------- Hassey 38 .323 .447 .452 Aikens 49 .375 .490 .725 Lombardozzi 36 .344 .417 .469 Salazar 27 .333 .333 .593 Owen 49 .366 .469 .415 Iorg 26 .522 .577 .826 Leonard 30 .370 .433 .815 Herndon 34 .310 .412 .552
Those are the playoff statistics of those 16 players. Based just on those numbers, you’d assume that Line-Up B would regularly trounce Line-Up A. But nobody doubts that Line-Up A is comprised of much better players than Line-Up B.
There are a couple of alternatives to explain those numbers. You might conclude that 25, 50, 100, or even 200 plate appearances don’t reveal how good a player really is. Or you might conclude that some players have specialized playoff ability and other players don’t.
No strong relationship exists between playoff and career statistics across the population of major-leaguers. For players with at least 25 playoff plate appearances, their playoff and career batting averages vary by an average of 54 points, or 20 percent.
This table shows the points and percent by which playoff and career batting average, OBP, and slugging average diverge for players with at least 25, 50, 100, and 200 playoff plate appearances:
No. of ----Points--- ---Percent--- PA Players Avg OBP Slg Avg OBP Slg -------------------------------------------- 25+ 856 54 57 93 20% 17% 23% 50+ 480 46 49 80 17% 14% 19% 100+ 160 36 37 63 13% 10% 15% 200+ 28 32 28 48 11% 8% 10%
In other words, even for the 28 players with 200+ playoff plate appearances, batting average varies by an average of 32 points. While the differences are huge across the board, the variance does diminish as the number of plate appearances increases. Point being, given enough opportunities, on average a player’s playoff statistics slowly converge with his career statistics.
Only a handful of players ever reaches 200 playoff plate appearances, though, and even the all-time leader in playoff plate appearances, David Justice with 408, still has well short of a season’s worth. Once a player reaches 200 playoff appearances, however, the numbers start to level off. The batting averages generally cease to look so astronomical or catastrophic, and the OBPs and slugging averages conform to more normal ranges.
But let’s say there’s such a thing as specialized playoff ability. How would you prove it? Would you assume the player who compiled these numbers in three series has it?
PA Avg OBP Slg -------------------- 27 .208 .296 .250 23 .217 .217 .304 29 .182 .379 .318
Probably not. But let’s put those numbers in context:
Series PA Avg OBP Slg ------------------------------- 1991 ALCS 21 .474 .524 .474 1992 ALCS 28 .423 .464 .692 1992 WS 27 .208 .296 .250 1993 ALCS 28 .292 .393 .333 1993 WS 27 .480 .519 .640 1996 ALDS 19 .294 .368 .471 1996 ALCS 23 .217 .217 .304 1997 ALDS 11 .300 .364 .500 1997 ALCS 29 .182 .379 .318 1999 ALDS 21 .368 .429 .579
You might’ve guessed that those are Roberto Alomar’s playoff statistics. Alomar was the hero of the 1992 ALCS. Little noted is how awfully he played in the 1992 World Series. And in 1996 and 1997, Alomar excelled in the ALDS. He was missing in action in the ALCS both those years, though. Did Alomar forget his specialized playoff ability?
Or how about a player from this year’s sweep of the Astros, Brian Jordan?
Series PA Avg OBP Slg ------------------------------- 1996 NLDS 13 .333 .385 .583 1996 NLCS 26 .240 .269 .480 1999 NLDS 18 .471 .500 .706 1999 NLCS 28 .200 .286 .440
Hero or goat? Depends on the series. So specialized playoff ability, if it exists, looks to be fleeting, or at least inconsistent.
Even if there’s no such thing as the specialized ability to raise the level of play in the playoffs, could it be that some players, in the pressure of the playoffs, simply don’t live up to their otherwise established abilities? Perhaps.
In sizeable numbers of plate appearances, Justice and Morgan have playoff batting averages 61 and 89 points worse, respectively, than their career batting averages. But Morgan also had a few series when he was decent or even excellent:
Series PA Avg OBP Slg ------------------------------- 1972 NLCS 20 .263 .300 .579 1972 WS 30 .125 .300 .208 1973 NLCS 22 .100 .182 .150 1975 NLCS 14 .273 .429 .545 1975 WS 32 .259 .375 .296 1976 NLCS 13 .000 .462 .000 1976 WS 17 .333 .412 .733 1979 NLCS 14 .000 .214 .000 1980 NLCS 19 .154 .421 .385 1983 NLCS 17 .067 .176 .067 1983 WS 21 .263 .333 .684
Did he take an anti-pressure elixir in the 1975 NLCS or the 1976 World Series? You could discount those series as aberrations. But if you’re going to do that, then why not call all 219 of his playoff plate appearances an aberration and look instead at the 11,329 career plate appearances he made in the regular season?
The playoffs are different, but how much different? In the regular season there certainly has to be some pressure involved in performing before tens of thousands of fans in attendance and hundreds of thousands more fans watching on TV, getting paid huge salaries, and performing at the most competitive level of sports in the world. But somehow these rarefied professional athletes crack when you turn it up a notch?
Not likely. The evidence seems to indicate far more strongly that playoff statistics, compiled in small samples, aren’t indicative of specialized playoff
ability, or inability, any more than they’re representative of overall ability. In other words, they’re random.
Ability is manifested in frequency: the frequency of hits, of home runs, of RBI. It’s not manifested in the distribution of that frequency. In other words, a player’s ability dictates how often he gets hits or hits home runs. Ability doesn’t determine when those events will occur.
In 600 at-bats, a .300 hitter will get 200 hits. Yet there’s no ability that dictates the distribution of those 200 hits. Since playoff statistics represent just a tiny part of a player’s overall performance, the tyranny of random distribution can make good players look bad and bad players look good.
But what are the odds of it happening in, say, four consecutive playoff series? Probably not that insignificant. If you cut up a player’s career of 5,000 plate appearances into 250 snippets of 20 plate appearances each, drawing from a hat, could you pull out four snippets where a player batted really poorly? Of course, just as you could pull out four snippets where he batted really well.
Randomness is difficult to accept, because there’s nothing to combat it. If there’s no such thing as specialized playoff ability, then it’s not as if a team like the Astros, now 2-12 in the playoffs the last five years, can go out and sign players with proven specialized playoff ability so they can reach the promised land.
When the Astros acquired Moises Alou, he was coming off a World Series in which he batted .321 with a .387 OBP and .714 slugging average. The very next postseason, though, Alou batted just .188 with a .188 OBP and .188 slugging average, despite a stellar regular season. And everyone forgets that in the NLDS and NLCS before the 1997 World Series, Alou batted .214 and .067, respectively.
Or consider the signing of 2000 World Series hero Jose Vizcaino, who many people wanted to start against the Braves, “since the playoffs are the reason the Astros got him.” Vizcaino’s playoff record included a .286 batting average, .295 OBP, and .381 slugging average, however. Better than many of the Astros, to be sure, but not exactly remarkable. Vizcaino was 1 for 6 against the Braves in 2001.
Playoff experience is not a factor in performance. Otherwise, it would be expected that playoff statistics would improve as a player appeared in more series. This doesn’t appear to be the case. A player in his first series is just as likely to succeed, or fail, as a player with several series under his belt. The virtue of playoff experience is mythology.
If whether a player’s established ability reveals itself in the playoffs is random, then what’s a team to do? Hiring a new manager, acquiring playoff heroes, or adopting different offensive strategies aren’t likely to do the trick if the team just plain doesn’t hit.
As Lance Berkman put it after yesterday’s loss: “People are going to paint us as chokers. You’ve got to realize, there wasn’t a man in this clubhouse who choked. We weren’t so nervous we couldn’t swing, or that we couldn’t take good swings. It just didn’t work out. That’s all you can say about it. What’re you going to do?”
The only way to fight the tyranny of random distribution is to continue to use the strength of frequency: good hitters hit more frequently than bad hitters. You put your best line-up out there and let them get you to the playoffs and strive to succeed there. Maybe somebody believes that the Astros would be better off with Julio Franco, Marcus Giles, and B.J. Surhoff instead of Jeff Bagwell, Craig Biggio, Lance Berkman, and Moises Alou.
Most people should smell that notion for what it is.
Wrangler
The recent culmination of the vilification of Larry Dierker got me reading a book I bought a few years ago, The Bill James Guide to Baseball Managers from 1870 to Today. Something James wrote in the Introduction sounded particularly familiar:
People like to talk about baseball managers, about Tony LaRussa and Bobby Cox and Joe Torre. The talk focuses almost entirely on who is a good manager and who is a lousy manager. The average fan has a one-dimensional image of a manager: He’s good, or he’s bad. If he’s real good, he’s a genius. If he’s real bad, he’s an idiot.
When the discussion turns to why a manager is good or why he is bad, you realize how little solid information is being used. On a talk show, 97% of all explanations as to why the local manager is an idiot will begin with the words “Well, one time he …” One time he bunted with the number-six hitter, sent up a pinch hitter, Felipe Alou walked the pinch hitter and then the number-eight hitter grounded into a double play; see, Felipe Alou outsmarted him there. One time he took out his starting pitcher with a three-run lead in the seventh inning, and his relievers didn’t have anything.
Much of what the average fan uses to form his impressions of a manager is no doubt valid … But people who discuss baseball managers — I include fans and professionals — have, in my experience, almost no conceptual framework within which to store these observations. Working with a one-dimensional concept of a manager’s job, we use what we learn to push managers up and down the scale, toward the “idiot” pole, or toward the “genius” pole. This is all we know.
The caricature was stunningly accurate. It also made clear why I took issue so strongly with much of what’s been said about Dierker. Where baseball is concerned, in the absence of some objective evidence, I’m skeptical of subjective judgments. But judgments about managers are inherently subjective.
I don’t really have an opinion on whether Dierker is a good or bad manager. He’s definitely not the best manager in baseball, and he’s almost certainly not the worst. Particularly with just anecdotes and opinion to go on, I’m wary of taking a strong position on the issue.
He’s done some foolish things. He talked too much about his players to the media. He was too orthodox in the use of his pitchers and the construction of his line-up. There was never a reason not to have Craig Biggio bat leadoff. He’s the best leadoff man since Rickey Henderson. Julio Lugo shouldn’t have batted higher than seventh no matter how speedy he is. Daryle Ward deserved as much playing time as it’s possible for a fourth outfielder to get. Billy Wagner could be used in close games that didn’t involve ninth-inning save situations.
But Dierker also did some things fine. He challenged his starting pitchers, and they appeared to respond. He didn’t feel the need to kick his players in the ass all the time. Some people found that to be a bad thing, but you can go overboard with that approach. When Don Baylor joined the Cubs, one of his priorities was to light a fire under Sammy Sosa. In case nobody noticed, Sammy Sosa needing a kick in the ass is not one of the Cubs’ bigger problems. Also, see, e.g., Terry Collins.
Tactically, he made errors. All managers do. But how often he made tactical mistakes is colored by any particular fan’s preferences. Some people think little ball is still viable. Others don’t. Fans differ on whether a manager should use a quick or slow hook. Or whether the manager should seek the platoon advantage or just bring in the better overall reliever or pinch hitter. There’s really no right or wrong answer to those questions. Except that many people believe the result in a particular situation determines whether a decision was the right one, which isn’t logical.
Regardless of what anyone thinks, Dierker’s almost certainly gone. He may resign rather than wait to be fired. It’s sad to see him leave under such circumstances, though. He’s been a loyal member of the organization for a long time, and, at least until recently, he was at least beloved if not admired or respected.
Don’t be surprised if the next manager, whomever he may be, doesn’t please everybody either. There’ll be a honeymoon, but the first five-game losing streak of 2002 will bring folks out of the woodwork to castigate the new skipper. It is hoped that he doesn’t face the same level of vulgar, hateful, sometimes irrational criticism that Dierker has endured. No matter how badly a man has done his job, there’s simply no excuse for some of the things Dierker’s been called or accused of.
Maybe the Astros will win a postseason series with a new manager. Barring a 2000-type collapse, they certainly should compete for a playoff berth again. For all the postseason misery, watching the Astros play in the first round four of the last five years sure beats watching hockey or preseason basketball like the fans of the 22 teams who didn’t make it to the playoffs. Perhaps next year the hits will finally break their way.
The impending managerial change might do the Astros some good. The Astros stood a better chance to win Game 1 against the Braves had Dierker pitched Octavio Dotel rather than Mike Jackson in the eighth inning. Dierker has been pilloried for other moves as well. There are no promises that a new skipper might not make mistakes, too, but maybe a change in itself will help.
James also wrote that “There is one indispensible quality of a baseball manager: The manager must be able to command the respedct of his players. This is absolute; everything else is negotiable.” Dierker has almost certainly lost that, may very well have lost it a while ago.
All the managerial replacements in the world aren’t going to make up for a near total outage on offense, though.