FRIDAY, FEBRUARY 16
Yesterday, I spent four hours chaperoning a quartet of third-graders on a field trip to the Pacific Science Center. These kids attend a quality elementary school in one of Seattle's more fashionable neighborhoods, their parents read them stories every night before bedtime, and they get plenty of attention in class from both committed teachers and concerned parents.
Spend some time with kids like this, and you'll be amazed at the activity of their minds, and their vast potential.
But between the third grade and adulthood, something happens to the kids. Or rather, something doesn't happen. Nobody teaches them how to think. Approximately 50 percent of current third-graders will, when they're adults, believe in astrology. Approximately 45 percent of those third-graders will believe in ESP. Approximately 40 percent of those third-graders will believe that dinosaurs and humans lived simultaneously. Approximately 67 percent of those third-graders will believe that they actually had a psychic experience. And some percent -- anything greater than zero is inexcusable -- of those third-graders will believe that the moon landings were faked.
Quite simply, we're just not taught to think critically enough, or logically enough. As "proof," and in keeping with the typical subject matter of this column, I offer the following, from Sports Illustrated and CNNSI.com writer Tom Verducci's recent piece on the value of Derek Jeter:
The Yankees offense runs through Jeter, similar to the way it does through a key scorer on a basketball team. Consider this: the Yankees have scored 286 runs in the postseason since 1996. Jeter has scored or driven home 61 of them, or 21 percent of his team's offense.
No writer should have written that, no editor should have allowed that into print, and no reader should let something like that color their thinking. Because it's patently ridiculous.
I don't mean to pick on Verducci, and I don't mean to pick on Jeter. Verducci is a fine writer and reporter, and Jeter is -- let me say this for the hundredth time -- a fantastic player.
But when I read something like that, I'm immediately skeptical. And when I'm skeptical, sometimes I sit down and try to figure out why I'm skeptical.
In this case, what caught my eye was "21 percent of his team's offense." Now, I'm often accused of "using statistics" to justify an argument, but that's backwards. A sabermetrician starts out with a question rather than an argument, and he goes where the numbers take him. But when Verducci says that Jeter accounted for 21 percent of the Yankees' postseason offense, we should immediately realize that before he started writing, he already knew the argument. So then it was just a matter of piling up as much statistical "evidence" as he could find.
Think about it. First off, when "percent" is introduced into a discussion, one naturally assumes that the percentage involved is on a scale of 1 to 100. And since Jeter was responsible for "21 percent of his team's offense," we're led to believe that he accounted for a fifth of the Yankees' postseason run production. A fifth! From one player! And a shortstop, no less!
Ah, but it's not 21 percent of 100 percent. It's more like 21 percent of 195 percent, because for every 100 runs, you've got approximately 195 runs + RBI (every run is not accompanied by an RBI). You see what I mean? So by extension, Jeter's 61 runs + RBI actually represent about 11 percent of the possible total (61/558).
And you know what 11 percent is? It's the same as one-ninth. How many spots in the lineup? Nine. Jeter, occupying one of nine lineup slots, accounted for -- can you believe it? -- one-ninth of the offense.
Actually, Jeter was a little better than that. Turns out that Verducci's figures are a little off. By my count (and I counted twice), since 1996 the Yankees have actually scored 274 postseason runs rather than 286. And Jeter has actually scored (48) or driven home (21) 69 of them. Divide 69 into (274 runs + 258 RBI), and you get 13 percent. That's better than what Verducci's "method" might suggest, but let's be honest, folks ... 13 percent is not appreciably greater than 11 percent. Jeter has done about what you'd expect from a great hitter who's played as many postseason games as he has. No more, and no less. But of course, that's no mean feat when you remember that he's a shortstop. Like I said, Jeter's a fantastic player. But we only diminish his greatness when we drag up phony stats to "prove" it.
(By the way, Bernie Williams has 77 runs + RBI in the last five postseasons. Q: So why didn't Verducci argue that the Yankee offense runs through Williams? A: Jeter was the singular object of Verducci's affections, so he didn't even bother to check Williams.)
And yesterday, I ran across another of Verducci's pieces in Sports Illustrated. This one summed up, with a fair amount of detail and artistry, the sorry state of the Baltimore Orioles. Unfortunately, nestled among the hundreds of quality sentences, like a venomous asp within a field of sunflowers, was this ...
Ripken, meanwhile, signed a one-year, $6.3 million contract in November, though he's 40 and fighting a cranky back that contributed to a .310 on-base percentage last year, worse than all but 11 of the 116 American League players with at least 300 at bats.
This time, a slightly different mis-use of statistics. Whereas in the Jeter piece, Verducci essentially invented a bogus statistic that doesn't tell us anything, this time he plucked a particular, telling statistic to support his case. But while it's true enough that Ripken isn't really worth $6.5 million any more, and it's also true that Ripken's .310 on-base percentage doesn't measure up (and you know how I feel about on-base percentage).
Still, wasn't it just a year or three ago that every sportswriter in the land was lionizing Cal? His OBP wasn't so hot then, either. What's more, Verducci somehow forgot to mention that Ripken's .453 slugging percentage was better than those posted by Fred McGriff and Terence Long and Brian Daubach and John Olerud and Gerald Williams and 56 other American Leaguers with the requisite number of at-bats. Why didn't Verducci mention slugging percentage? Because it didn't help him make his point.
Quite often in the last week or so, I have been advised, "Get your head out of the computer and watch some ballgames." Well, I went to the ballpark 113 times last year. And lately I've come to realize that the great majority of correspondents who complain about my use of (some would say "reliance on") statistics are really objecting not to that use itself, but rather to whichever statistics-based conclusions happen to conflict with the correspondents' opinions.
I just go where the numbers take me. See, that's the difference between sportswriters and sabermetricians. As Bill James famously wrote, upon ?retiring,?
Sportswriters, in my opinion, almost never use baseball statistics to try to understand baseball. They use statistics to decorate their articles. They use statistics as a club in the battle for what they intuitively believe to be correct. That is why sportswriters often believe that you can prove anything with statistics, an obscene and ludicrous position, but one which is a natural outgrowth of the way that they themselves try to use statistics.
Fans often believe that, too, that you can "prove anything with statistics." But you can't, not unless you (1) cheat, and/or (2) are blessed with a gullible audience.
Like I said, I don't mean to pick on Verducci. Still, his limitations depress me a little, because he's not one of the old guys who already had a Hall of Fame vote when Bill James wrote his first Baseball Abstract. Verducci's a young man; by the looks of his photo at CNNSI.com, younger than I. So, quite frankly, he doesn't have any excuse for this silliness.
At some point, the train stopped at his station. But like too many, he just didn't bother getting on.
WEDNESDAY, FEBRUARY 14
Science isn't truth, it's merely our best stab at it. At the heart of scientific inquiry lies its thrilling contingency, the possibility that today's crackpot theory may be tomorrow's natural law ... As recently as the early 1960s, the theory of continental drift was considered so ludicrous that geology professors included it in their lectures on daffy moments in earth science.
-- Bruce Barcott, "The Measure of a Mountain"
In regard to the burning question of Derek Jeter's defensive prowess, or lack thereof, we've probably taken that about as far as we're going to, at least for a while. I can now say that I've been castigated on the air by real-life Yankee broadcasters, thus fulfilling a long-standing fantasy.
However, the burning question of how to measure defensive prowess remains with us. Range Factors -- or, more prosaically, putouts plus assists per nine innings -- are a starting point in any discussion of defense, rather than an ending point.
And as I wrote last week, I believe that if a shortstop were outstanding defensively, he would, at least over the course of a few years, tend to have higher Range Factors than his backups. If Ozzie Smith (for example) had a teammate who was making more plays than him, then that teammate would probably be playing regularly for another team, right? So shouldn't Ozzie have had better defensive statistics than the men who filled in for him?
Before I had a chance to run Ozzie's numbers, I got an e-mail from Baseball Prospectus co-author Clay Davenport, wherein he listed the Range Factor ratios (regulars vs. backups) for a bunch of current shortstops. Here's his message, which I found a tad confusing, but essentially understandable ....
Rob,
In case you need more support, here are ratings for 19 current current shortstops. In each case I used the ratio of their Total Chances per Nine Innings (Range Factors) to that of the rest of their team, weighted by the rest-of-team innings. That changes things from the straightforward add-em-up you used, but in most cases (like Jeter's) the results are very similar.
Only seasons with 500 innings at SS considered, and I needed to see three entries in the last five years.
Raw Raw
Rate Yrs Ind Rest Ratio
1 N Perez 118 3 5.18 4.54 114
2 R Sanchez 112 3 5.33 4.75 112
3 J Valentin 109 5 5.02 4.51 111
4 D Cruz 108 4 4.84 4.57 106
5 M Tejada 107 3 5.03 4.87 103
6 A Rodriguez 106 5 4.75 4.67 102
P Meares 106 4 4.91 4.76 103
8 E Renteria 105 5 4.66 4.58 102
9 R Ordonez 104 4 4.78 4.75 101
10 A Gonzalez 102 4 4.76 4.65 102
11 B Larkin 100 5 4.42 4.49 98
O Vizquel 100 5 4.69 4.73 99
K Stocker 100 5 4.76 5.06 94
W Weiss 100 5 4.77 4.65 103
15 M Bordick 99 5 4.85 4.57 106
16 Nomar G 98 4 4.73 4.80 98
17 D Jeter 97 5 4.38 4.53 97
18 R Clayton 93 4 4.92 5.28 93
19 R Gutierrez 91 3 4.54 4.83 94
There's definitely a trend towards the best shortstops having the better scores.
Thanks, Clay.
In case anyone's thinking that Clay is making things more complicated than they need to be, that's what I thought first, too. And then I realized that he's got a point. I don't bother with weighting anything, which can silly up my results. But as Clay points out, generally it doesn't make much difference, Kevin Stocker being the obvious exception.
OK, next I got the career data for Ozzie Smith (thanks to Ozzie's cousin, Dave Smith over at Retrosheet). Actually, I stopped at 1993, Ozzie's last season as a regular ...
1978-1993 Ozzie
Range 5.24
Others 4.88
Huh? Ozzie only .36 per nine innings better than his backups over 16 seasons? But wait a second. That's about 56 plays per season, and 56 plays is a whole bunch of plays. Conclusion: Ozzie Smith was a great shortstop for a long time. And we could have figured that out even if we never saw him play. (Not that we wouldn't want to. I saw more than 100 games in person last year, and I don't have any plans to retire from ballpark attendance.)
That's not something that most baseball men would believe, though. And speaking of baseball men, I contacted a few of them yesterday. I wanted to know how they went about evaluating defensive skills. The following opinions should not be considered representative, however, as I contacted organizations that tend toward forward thinking.
Rangers general manager Doug Melvin said, "I think, with defense, that's one of the reasons we pay scouts. They see the plays, they see if the guy gets a good jump. Does he turn the double play? Does he get back quickly on the short fly balls? I think that scouts are the best evaluators of defense."
At the same time, Melvin doesn't completely disregard the objective measures of defensive performance.
"We picked up Randy Velarde based on some statistical information," Melvin admitted, "but you have to see if the physical skills are still there, too. The numbers showed that he still had good range, plus our people said that he was still turning the double play really well."
Royals general manager Allard Baird takes something of a combination approach, too. "At the major-league level, I think you combine objective and subjective. You need to watch a guy to know how quick they are, but we look at Range Factors, too. We need to know if a player can finish a play. Some guys can get to the ball and make the exchange, but can't finish off the play."
In the course of our conversation, Baird pointed out a couple of things that I'd never considered much.
"With a club like ours that has limited strikeouts, defense become a little more important. Also, when you talk about range, positioning before the play is important. And a lot of that comes from advance scouting, which means that Range Factors can actually be tied to the organization rather than just the player."
I told Baird about Rey Sanchez's excellent fielding numbers, to which he replied, "I'm not surprised. I think that Sanchez is the most underrated defensive shortstop in either the American or National League."
Of course, Baird sees Sanchez play every day. I actually agree with Baird about Sanchez, but this points out a flawed argument that I've been hearing in favor of Jeter: "To appreciate him, you have to see him play every day." Fine, but that's a conversation ender rather than a conversation starter. And anybody, anywhere, can make the same argument. I'm sure there are people in Anaheim who will swear up and down that Gary DiSarcina is great, just as there are people in Kansas City who will swear that Sanchez is great. But they can't all be great. Greatness is, by definition, uncommon. And so I simply don't buy the "I see him play every day, you don't, so I'm right and you're wrong" argument.
Brad Kullman doesn't make arguments like that. Kullman, the the Director of Baseball Administration for the Cincinnati Reds, is one of the young baseball executives who doesn't curl up like a pill bug when he hears the word "sabermetrics."
"Watching a particular player, you can't watch the ball," Kullman says. "Getting that first step is a big key, and you just can't see that. But what you can do is watch the whole defense when the ball is hit, and see who reacts first no matter where the ball goes. I can remember seeing Pokey Reese when he was coming through the minors, he always seemed to be the first guy on the defense who moved."
And what about defensive statistics?
"I'm trying to figure out a way that we can use them," Kullman says, "but in general, we just don't look at the statistics."
It's funny. Whenever a new statistic comes along, people hold it to unreasonable standards. Yes, there are inherent biases in Range Factor. And no, we can't assume that a player with a high range factor is a great fielder, and a player with a low range factor is a poor one. But you know, a fair number of other statistics have some problems, too. Does anybody hold it against home runs that they're a lot easier to hit in Coors Field than in Comerica Park? Do people run down RBI because they're a lot easier to come by when you're following Kenny Lofton and Roberto Alomar in the batting order?
It's often argued that Range Factors don't "work" because particular players don't have good ones. Of course, it's really not fair to judge statistics by how well your favorites stack up -- is slugging percentage meaningless because Ozzie Smith didn't have a good one? -- but still, it's long been a black mark against Range Factor that Roberto Alomar's was subpar, relative to the league average.
Alomar won his first Gold Glove in 1991, and he's been winning them ever since. But his Range Factors were not impressive. In fact, they were generally well below the league average. But maybe he was playing behind fly-ball pitching staffs? Or maybe something else was going on. So what if we compare him to the other second basemen on his team?
Alomar Others
Range 4.98 4.73
I can hear it now ... "What?! Alomar only 0.26 plays better than his backups, people like Jeff Reboulet and Domingo Cedeno and Billy Ripken and Enrique Wilson? You're insane!"
Ah, but 0.26 plays per nine innings is not a trifling figure. Project those 0.26 plays per nine innings to 154 games, and you get nearly 40 plays.
That's a lot. Think about it like this: if Alomar had made 40 fewer plays last year, but collected 40 more singles, he'd have batted .375, and edged Nomar Garciaparra for the American League batting title.
The difference is ever more dramatic if we just look at the last five seasons. When Alomar played in Toronto, he annually ranked at, or near the bottom of, the Range Factor ratings. I spent many hours trying to explain his poor showings, but had little luck.
Here are two charts, the first describing Alomar's performance relative to his teammates in Toronto (1991-1995), the second his performance relative to his teammates in Baltimore (1996-1998) and Cleveland (1999-2000):
1991-1995 Alomar Others
Range 4.84 4.99
1996-2000 Alomar Others
Range 5.12 4.60
When he was with the Blue Jays, Alomar made fewer plays than his backups (mostly Domingo Cedeno and Alfredo Griffin). Granted, Cedeno was considered a fine defensive shortstop, and might well have been excellent at second base, too. Still, those numbers aren't what you'd expect if Alomar really was the best-fielding second baseman of all time, as some claimed even then.
But look at that second chart. Since leaving Toronto, Alomar's been 0.52 plays better than his replacements (who played 1,287 defensive innings, essentially a season's worth). That 0.52 comes to 80 plays per 154 games, and that's a huge, titanic number of plays.
So it's quite possible that Range Factors do tell us, when interpreted with even a small degree of sophistication, that Roberto Alomar is an outstanding second baseman.
TUESDAY, FEBRUARY 13
It's rare that somebody writes about baseball history and it hits the front pages. But two weeks ago, events nearly 50 years old did hit the front pages: first The Wall Street Journal, and then a few other great newspapers. The architect of all this? A talented, enterprising writer named Joshua Harris Prager, who turned some old, dusty rumors into a hard, cold fact: in the latter stages of the 1951 season, at the end of which they captured the National League pennant, the New York Giants employed a sophisticated system for stealing catcher's signs and relaying them to the batters.
The sign stealing began on July 20. From that point through the end of the season, the Giants played 28 games at their home ballpark, the horseshoe-shaped Polo Grounds, and won 23 of them. Over the same span, they went 29-13 on the road.
According to Prager, only about half the Giants hitters -- Bobby Thomson, among them -- did want the signs. If we figure 40 plate appearances per game, and half of those going to hitters getting the signs, we might (very roughly) estimate that approximately 560 plate appearances were conducted under questionable circumstances. That's a lot of plate appearances, and certainly leads to the question, "Does the sign stealing take anything away from the Giants' accomplishment?"
Before answering that question, it's worth noting that sign stealing, above and beyond the garden-variety, baserunner-on-second-peering-between-catcher's-legs version, has a long tradition, going back at least a century.
In 1900, the Philadelphia Phillies went just 30-40 on the road, but played brilliantly at home, going 45-23 at Philadephia Park (later renamed Baker Bowl).
As it turned out, the Phillies almost certainly employed an elaborate sign-stealing scheme, with the help of two part-time players.
Utility man Petie "What's the Use" Chiles often coached third base, where (as researcher Joe Ditmarr reports) "he had an unusual twitch in his legs at times and often stood in one position, right in the middle of a perpetual wet spot, in the corner of the coach's box."
Backup catcher Morgan Murphy rarely played, and when he wasn't in the lineup he was also absent from the bench and field area.
On September 17, with the Cincinnati Reds in town for a doubleheader, Chiles and Murphy were exposed. According to Ditmarr,
In the third inning of the first game, Tommy Corcoran, the Reds shortstop and captain, began frantically scratching with his spikes in the third base coaching area. Acting like a demonic chicken searching for grain ... Just below the surface Corcoran struck the lid of a small metal box. Opening the box exposed an "electric buzzer device" with protruding wires. It was thought that Chiles' cohort, Murphy, was stationed in the clubhouse behind the center field wall with some sort of pirate spyglass with which to steal the catcher's signs. Murphy, it was then assumed, would relay the information to Chiles' feet and he would verbally signal the batter as to whether the next pitch was to be a fastball or curve.
Not much came of Corcoran's discovery. But 12 days later in Pittsburgh, the Cincinnati shortstop sniffed out another scheme, this time in Pittsburgh. It was reported that the Pirates and Phillies knew of each other's chicanery, and had even agreed to not spy on each other. Of course, there's little honor among thieves, so we can imagine that both clubs were on the lookout for truce violations.
After the season, Philadelphia baseball writer Charles Dryden confirmed the details of the Phillies' sign-stealing scheme. Also after the season, Petie Chiles was arrested in Texas for involvement in a con-artist scheme and sentenced to two years of hard labor. In 1902, eight months before his scheduled release, Chiles escaped from custody. He was arrested for assault in 1903, reportedly played semi-pro baseball that same year, and that's the last we know of him. When and where he died remains a mystery.
In 1940, the Tigers edged the Indians by one game to capture the American League pennant, after a three-and-a-half-month duel. In the process, the Tigers beat Bob Feller -- who finished the season 27-11 and was generally considered the AL's best pitcher -- six times. There were rumors that the Tigers had sign-stealing spies in the center-field bleachers, and Feller echoed these rumors in 1990 when I asked him about it. "Yeah, they had a guy with binoculars out there," he said, "and he'd signal to the hitter what I was going to throw. Cost us the pennant."
In 2000, rumors abounded that the Blue Jays had rigged up a sign-stealing system in SkyDome. I asked two men very close to the team about this, and one of them snorted, "These guys? There aren't five of 'em smart enough to use a system if there was one."
These incidents are just a small sampling. In his piece, Prager also mentioned schemes from the early 1960s (Milwaukee's County Stadium) and the 1980s (Chicago's old Comiskey Park), and of course there must have been many more attempts, if not successes.
A few points about the Giants in 1951:
Everybody knows that after July 19, the Giants went 52-18 and surged to the National League pennant. And now everybody knows that, over that same span, they employed a sophisticated system for stealing signs. But how many people know how effective that system actually was?
Dave Smith of Retrosheet has the game data for each game of the Giants' 1951 season, so he checked the "before and after" numbers. The results are, to say the least, surprising.
Home OPS Road OPS
Thru July 19 814 725
After July 19 761 758
Yes, the Giants actually hit worse at the Polo Grounds after they started cheating. As Smith points out, the real improvement came in their road hitting, and especially the pitching. Before July 19, the Giants pitchers posted a 3.47 ERA at home, 4.49 on the road. After July 19, they lowered those figures to 2.90 and 2.93. So the pitching improvement is the real story of the Giants' second-half comeback.
Of course, this doesn't mean that stealing the signs didn't help them. Perhaps without cheating, their home OPS decline would have been more severe. And of course, had the Giants won just one fewer game, there would have been no pennant, because there would have been no playoff series with the Dodgers.
So let's make no mistake -- the Giants did cheat. True, it wasn't until 1961 that a rule was instituted banning the use of mechanical devices for spying on the opposition, which means that manager Leo Durocher and the rest of the club didn't do anything violating the letter of the law.
But the Giants cheated, and they knew it. I've got a lot of baseball books in my basement, and a fair number of them were written by men who knew what was going on in 1951.
Giants center fielder Willie Mays has done a couple of autobiographies, including a fine 1966 book with Charles Einstein. There's no mention of sign stealing.
Giants shortstop Alvin Dark dictated a book, "When in Doubt, Fire the Manager." Not only is there no mention of sign stealing, but Dark says of the Giants' comeback, "It couldn't be pinned to any precise moment, yet suddenly we were a different team."
And then there's Leo Durocher. Understand, Durocher was not one to shy away from controversial statements. Spend just a few minutes leafing through Durocher's book -- the wonderful "Nice Guys Finish Last" -- and you'll read about Giants owner Horace Stoneham's alcoholism and Ernie Banks' single-minded devotion to his public image. Yet there's no mention of sign stealing in "Nice Guys Finish Last." In fact, Durocher claims that he told Thomson to expect a fastball, both on the first pitch (that Thomson took for a strike) and the second (that he hit over the fence).
Giants right fielder Monte Irvin did a book in 1996 called "Nice Guys Finish First." There's no mention of sign stealing.
Nearly every other Giant has been quoted in various books -- I've got one called "The Miracle at Coogan's Bluff," and another called "The Home Run Heard 'Round the World" -- and none of them includes a mention, an inkling, even the tiniest hint of any chicanery. Now, I certainly don't blame any of the Giants for failing to volunteer such information. After all, when finally confronted with pointed questions by Josh Prager, they all 'fessed up. But earlier, not one player wanted to be the one to spill the beans. Personally, I have no ill feelings for any of the Giants, any more than I have ill feelings for Gaylord Perry or Whitey Ford. Baseball's always been like Wall Street or tax time: It ain't cheatin' if you don't get caught.
In answer to the question, "Does the sign stealing take anything away from the Giants' accomplishment?" I would direct you to the Giants themselves. Clearly, they believed that it did. Knowing that, you can decide for yourself.
P.S. One thing bothered me about Josh Prager's article ... If the Giants won the National League pennant thanks to chicanery in 1951, then what about 1952? And '53 and '54 and '55 (Durocher was gone in '56)? Prager told me that according to Giants pitcher Al Corwin, they did not steal signs in 1952 (when they finished in second place, six games behind Brooklyn) or 1953 (fifth place), but they did cheat in 1954, when they won both the National League pennant and the World Series.
FRIDAY, FEBRUARY 9
Regarding my last column, about Derek Jeter and his peers ...
Rob:
I agree with you that Jeter is not quite the player that A-Rod or Nomar is, but you neglected to mention a factor in Jeter's favor: his superior durability. Over the past three years (1998-2000) Jeter has averaged 687 at-bats plus walks per year, compared to 648 for ARod and 603 for Nomar. Over the same span, Jeter has averaged 1,327 defensive innings played per season, again noticeably more than A-Rod (1,263) or Nomar (1,204).
Also, Jeter's OBP is best of the three, and, as you know, OBP is slightly more important than SLG. Ordinarily this would be nitpicking, but in such a close comparison I feel that minute analysis is warranted. Jeter is still the least potent of the three as a hitter, but the gap is a little smaller than would appear from OPS.
And finally, Jeter is a year younger than Nomar (though a year older than A-Rod). Taking into account his lower age and superior durability, I think a case can be made that Jeter's projectable future is as valuable as Nomar's ... especially since the latter's defense is nothing special, either.
Regards,
Tom Bell
Chicago
A few people wrote me about Jeter's durability, and a few people wrote me about Jeter's OBP, but Tom's was the only message that included both. And they're good points. I do tend to stress durability, and it's certainly true that Jeter has been the healthiest of the three. It's also true that OBP is somewhat more important than slugging percentage. In fact, (OBP*1.4) + (Slugging) correlates better with run production than simple OPS (OBP + Slugging). So here are relevant three-year totals for the three shortstops:
Derek Nomar Alex
Games 455 418 438
ModOPS 1083 1160 1113
("ModOPS" stands for Modified OPS, where OBP is weighted 40 percent more heavily than slugging.)
This closes up the ranks quite a bit, as Jeter draws close to Rodriguez in the hitting stat, and looks significantly better than Garciaparra when it comes to durability. Throw in their home ballparks, and the one-year difference in age, and I might be convinced that Jeter will, indeed, be just as valuable as Garciaparra over the next decade or so.
But the real thrust of Wednesday's column was this: there is no evidence that Jeter is the outstanding defensive player that we're told he is. That doesn't mean he's not a great player. All things considered, he's still an A (and not a B+, as I argued Wednesday). He's just not the A+ that some writers and broadcasters advertise.
Because to be an A+, he'd have to be an outstanding defensive shortstop, and I maintain that there's virtually no evidence to suggest that. To that end, I'll ask the same question today that I asked Wednesday: If you have an outstanding defensive shortstop, would you expect him to make more plays than his replacements, or fewer?
I think you would expect him to make more plays than his replacements. It's such an obvious conclusion that I worry that I'm wasting your precious time when I stress the point.
And as we saw Wednesday, Derek Jeter has actually made fewer plays than his replacements. He's made 4.27 plays per nine innings, and his replacements -- among them, Luis Sojo, Andy Fox and Alfonso Soriano -- have made 4.31 plays per nine innings.
Given that the thrust of my argument wound up being that Jeter is not an outstanding defensive shortstop, I didn't bother checking anyone else. But I was on ESPN Radio that evening, and in the course of castigating my callous disregard for intangibles and whatnot, co-host Chris Moore lambasted me for failing to run the same numbers for Garciaparra and Rodriguez.
Moore's criticism was unfounded, because I never claimed that either Garciaparra or Rodriguez are great defensive shortstops. Nevertheless, I decided to check anyway, throwing in another famous American League shortstop as well.
Jeter Nomar Alex Omar
Range 4.27 4.58 4.63 4.62
Others 4.31 4.65 4.56 4.57
For Jeter, the sample size includes five seasons, 6,767 of his defensive innings, and 461 defensive innings for his replacements.
For Garciaparra, the sample size includes four seasons, 4,956 of his innings, and 821 innings for his replacements.
For Rodriguez, the sample size includes five seasons, 6,291 innings, and 889 innings for his replacements.
And for Omar Vizquel, the sample size includes five seasons, 6,479 innings, and 752 innings for his replacements.
I suspect that if you look at the table again, something will strike you ... there was virtually no difference between the number of plays made by the regular shortstops and their replacements.
While this certainly doesn't prove anything, what this suggests is that the number of plays made by a shortstop -- and, by extension, players at every position -- are highly dependant on the pitching staff, and other outside factors.
Let's check another pair of shortstops, Rey Sanchez and Rey Ordonez, both of whom are generally regarded as excellent fielders.
Sanchez Ordonez
Range 5.28 4.64
Others 4.81 4.38
For Sanchez, the sample includes two seasons with Kansas City (1999 was his first as a regular since 1995), 2,327 defensive innings, and 533 defensive innings for his replacements.
For Ordonez, the sample includes five seasons, 5,179 innings, and (a whopping) 2,085 innings for his replacements.
The differences in plays made between these guys and their replacements are still not huge, but at least they're noticeable.
We shouldn't be too hasty in reaching any conclusions. That said, I now suspect that neither Derek Jeter nor Nomar Garciaparra nor Alex Rodriguez nor Omar Vizquel is truly an outstanding defensive shortstop, while both Sanchez and Ordonez are at least very good.
So why do many people continue to tout Jeter and Vizquel (and others) as great defensive players? It's very simple, really. Those guys often look outstanding, making barehanded grabs (Vizquel) or powerful throws from the hole (Jeter). Unfortunately, the most important thing about playing defense -- the quickness of the fielder's first step -- is something that virtually nobody notices, because we're too busy watching the ball come off the bat.
So instead, we tend to rate defensive players on what happens at the end of the play, rather than the beginning. We rate them on style.
And sometimes we get lucky, and it works.
Ozzie Smith was probably the most spectacular shortstop that we've seen, and that's why he won 13 Gold Gloves. But he also created an immense number of outs with his dazzling glovework. Ozzie led the National League in Range Factor seven times.
Bill Mazeroski was probably the most spectacular second baseman of the last half-century, and that's why he won eight Gold Gloves. But he also created an immense number of outs. Maz led the National League in Range Factor eight times.
Those guys weren't brilliant defensive players because they looked great; it just worked out that way.
Gary Huckabay writes, "We wouldn't dream of judging hitters based on the appearance of their swing, but the public and broadcasting mainstream wouldn't dream of judging defense any other way."
Gary's got a great point. I'll never forget the day I watched Rob Ducey take batting practice. Most impressive pre-game display I've ever seen, this side of Mark McGwire. But nobody considered or considers Rob Ducey a superstar, because he didn't produce the numbers of a superstar. Add up his sweet line-drive swing and his mediocre performance, and you get the 24th or 25th guy on the roster.
But if you look good in the field, they'll give you a Gold Glove, actual performance be damned.
And if you can hit, too? By God, they'll give you the keys to New York City.
WEDNESDAY, FEBRUARY 7
February 7, 2001.
The day we answer the burning question, "So how much is Derek Jeter worth, anyway?"
In recent years, and especially in recent days, it's been the fashion to sing the praises of The Magnificent Derek Jeter. The Greatest Yankee Shortstop, even better than the Scooter. Maybe as good as Nomar Garciaparra. Maybe even as good as Alex Rodriguez, the $250 Million Man.
Jeter's better than Rizzuto. But is he close to being as good as Garciaparra and/or Rodriguez? Here are some key career numbers for each of them:
Alex Nomar Jeter
OPS 935 955 862
Garciaparra has certainly benefited from playing half his games in a good hitter's park, but the advantage certainly can't account for a 93-point difference in OPS. Same story with Rodriguez, who's actually spent the last season-and-a-half playing in a great pitcher's park (and before that, the Kingdome wasn't nearly as hitter-friendly as widely believed). As for the "little things," Jeter is a good baserunner, but no better than Rodriguez. And of course he's not often asked to bunt.
As hitters, Rodriguez and Garciaparra both earn A's, and Jeter gets a B+. Nothing to be ashamed of -- there are plenty of Hall of Fame shortstops who didn't hit like Jeter does -- but he certainly doesn't compare favorably with his slugging peers.
Well, if it's not his hitting, it must be his fielding, right? People rave about Jeter's defense, with Tim McCarver leading the wild-eyed hyperbole parade. However, there is virtually no objective evidence to support the notion that Jeter is a Gold Glove-quality shortstop.
The simplest fielding metric is called Range Factor: Putouts plus Assists, per nine innings.
And Jeter's Range Factor is, in a word, execrable. It's execrable every year. Here's where he ranked, in each of the last five seasons, among major league shortstops who started at least 100 games:
1996 20th of 24
1997 16th of 24
1998 23rd of 25
1999 21st of 21
2000 23rd of 23
Those numbers, elegant in their simple consistency, speak for themselves.
But of course, Range Factor is subject to all sorts of outside influences. Heinous things like pitching staffs and infield surfaces and gosh knows what else. There's a way to account for those outside influences, though. We can compare Jeter to the other Yankees who played shortstop in the same seasons that he did. Unfortunately for the purposes of this little argument, Jeter has been exceptionally durable, but the table below lists Jeter's Range Factor, along with the composite Range Factor of the other Yankee shortstops of the last five seasons.
Jeter "Others"
Innings 6767 461
Range 4.27 4.32
Now, this certainly isn't conclusive. But if Jeter were truly a superior defensive shortstop, wouldn't you expect him to make more plays per nine innings, rather than slightly fewer, than slow-footed players like Luis Sojo and Clay Bellinger?
As I said, Range Factor is subject to various outside influences. Fortunately, Clay Davenport (one of our colleagues at Baseball Prospectus) goes to great pains to adjust for those influences. Davenport's method rates Jeter as 23 runs worse than the average American League shortstop (given the same playing time) last season. That was the worst in the majors.
It was also the worst showing of Jeter's career, but then he's never done well by this measure. In 1999 he was minus-12; in '98, minus-3.
And you know, it's not just Davenport. Fielding statistics are not, despite what you might have heard from your favorite Luddite, meaningless. They can be quite meaningful in the hands of bright people. And as it happens, any number of bright people -- Bill James, Tom Tippett, etc. -- have designed their own methods for evaluating defensive statistics, and I believe that they all reach the same conclusion: Derek Jeter is, at best, an adequate defensive shortstop. He simply doesn't make many plays, and that's true even if you adjust for the left-handedness (or not) of the Yankee pitchers, and it's true even if you also adjust for the tendency of the Yankee pitchers to allow ground balls (or not).
Is Jeter the worst defensive shortstop in the major leagues? Maybe he is, maybe he isn't. There is some evidence to suggest that he is, but I'm sure there's evidence to suggest that he isn't, too. My point is that there's no evidence to suggest that he's an outstanding defensive shortstop, or even a good one.
Wait, that's not precisely true. There is plenty of anecdotal evidence in Jeter's favor. McCarver thinks he's great, and so do a lot of other people. But you know, even Tim McCarver doesn't know everything. McCarver sees that Jeter is adept at snagging pops in short left field, and McCarver sees that Jeter does, indeed, boast a powerful arm. But other observers have seen that Jeter doesn't get a good jump on grounders -- the single most important skill for any middle infielder -- and others have cast doubt on his footwork, especially when he fields balls up the middle.
I simply don't believe that Jeter is a good fielder; the idea that he is a good fielder will likely endure as one of the great baseball myths of our time.
After reading all this, somebody out there will still be arguing, "Yeah, but Jeter's a winner. How many World Series have those other guys won?"
The answer, of course, is that they haven't won any. Zero, compared to Jeter's four.
Vlad Guerrero hasn't played for a World Series winner yet. Does that mean he's not as good as Paul O'Neill? Jeff Bagwell doesn't have a ring yet. Does that mean he's not as good as Tino Martinez? Randy Johnson hasn't reached the Promised Land yet. Does that mean he's not as good as Andy Pettitte?
Of course it doesn't. It's a silly, circular argument. Derek Jeter is great because the Yankees win. The Yankees win because Derek Jeter is great. Round and round we go, and where the specious logic stops, nobody knows.
So what is Jeter worth? If you're the Yankees, he's worth whatever it costs to keep him. Because you've got bottomless pockets, and because Derek Jeter is a very good baseball player.
Those are good reasons. But let's not jump to the ill-founded conclusion that Jeter is in the same class as Rodriguez and Garciaparra. Because he's not.
MONDAY, FEBRUARY 5
You may have heard that the Orioles are altering the playing field at Camden Yards this season. According to Orioles spokesman Bill Stetka, the outfield fences will be seven feet farther from the plate, "give or take a couple of feet." In other words, it might be seven feet, or it might be five, or it might be nine. And there's obviously a big difference between five and nine feet, so we'll just have to wait and see about that.
As I wrote a few weeks ago, Oriole Park at Camden Yards is decidedly a pitcher's park, at least relative to the other American League stadiums. From 1998 through 2000, there were nine percent fewer runs scored in Orioles home games than in Orioles road games (not including interleague play). Nine percent is significant; over those three years, only Oakland's Coliseum favored the pitchers more.
The Orioles, you see, had the perfect ballpark: a pitcher's park that everyone thought was a hitter's park. Why is that perfect? Because such a ballpark limits the wear and tear to pitchers (especially the juvenile hurlers), but is attractive to hitters (especially the stats-sensitive sluggers).
But now a franchise that can actually spend with the rich kids might have a bit more trouble signing the big boppers. Oh well, a wise man once told me that among the hallmarks of an idiot organization is the tendency to monkey around with its ballpark's dimensions. And any list of idiot organizations almost certainly must include the Baltimore Orioles.
What's more, while the club's intent to help the pitchers might well make sense, it remains to be seen what the effects of the alterations will be. For one thing, while it'll be harder to hit home runs at Oriole Park, it'll presumably be easier to hit doubles and triples, due to the larger gaps. What's more, moving the plate toward the backstop results in less foul territory. I've never seen any sort of definitive study on the relationship between foul ground and batting average, but I'm certain that there is a relationship.
A few random notes about ballpark effects:
A few weeks ago, I noted that in 2000, Wrigley Field played as the best pitcher's park in baseball; 19 percent fewer runs were scored in Cubs home games than in Cubs road games. The numbers were so extreme last season that they even made Wrigley look like a pitcher's park for the previous three seasons combined. However, if we go back a year, and look at the three-year span 1997 through 1999, Wrigley looks like a good hitter's park, with eight percent more runs scored in Cubs home games. I remain fairly convinced that last year's numbers were an aberration, perhaps the result of wacky weather and fluky player performances.
While large one-season variations in (apparent) ballpark effects are certainly possible, they're also fairly rare, absent structural changes. Coors Field, for example, has been quite consistent over the last three years. Over the last three years, 63 percent more runs have been scored in Rockies home games than in Rockies road games (NL games only). And within those three years, it breaks down like this: 2000, +65 percent; 1999, +52 percent; 1998, +60 percent.
Camden Yards is in the news, but Cincinnati's Cinergy Field is worth watching as well. Already a good hitter's park, Cinergy might become a great hitter's park in 2001. Due to construction on the Reds' new ballpark, the outfield fences will be coming in about 20 feet, home plate back about 10 feet. The combination means less foul territory and fences 10 feet closer to the batter. There will be a 30-feet-high fence (the hitter's background) in straightaway center, and the fences in the power alleys and possibly down the lines will likely be raised from eight feet to 14 feet. We should expect big power years from not only Griffey, but perhaps guys like Sean Casey and Dmitri Young.
In their first seasons, Houston's Enron Field and San Francisco's Pacific Bell Park played as complete opposites. Pac Bell was a great pitcher's park (just as Candlestick had been), suppressing run production by 16 percent. Enron was a great hitter's park (not like the Astrodome), apparently promoting run production by 19 percent. It's only one season, of course, so we don't know for sure about either ballpark.
FRIDAY, FEBRUARY 2
If there's one thing Cardinals manager Tony La Russa and Cardinals pitching coach Dave Duncan have done well over the years, it's work with veteran pitchers. When La Russa and Duncan were in Oakland, Dave Stewart resurrected his career. Bob Welch won 27 games. Shoot, Storm Davis even had a big year. On the other hand, in all their years in Oakland, La Russa and Duncan never developed a young starting pitcher, and they've not had much luck since coming to St. Louis in 1996, either.
Thus, a number of analysts thought the Cardinals were going in the right direction prior to the 2000 season, when GM Walt Jocketty acquired three veteran starting pitchers: Darryl Kile, Pat Hentgen and Andy Benes. At the time, it looked like a gamble. Here's what the three had done in 1999:
Kile Hentgen Benes
Team Rockies Blue Jays Diamondbacks
Record 8-13 11-12 13-12
ERA 6.61 4.79 4.81
Kile pitched horribly in Colorado, while Hentgen and Benes posted nearly identical poor numbers in Toronto and Arizona, respectively. If these three did not perform better, as a group if not individually, it would be a long season in St. Louis.
The Cardinals did, of course, win a division title ... but we can hardly credit Hentgen and Benes. Kile, yes. But the other two didn't improve even a single iota between them.
1999 2000
Kile 8-13, 6.61 20- 9, 3.91
Hentgen 11-12, 4.79 15-12, 4.72
Benes 13-12, 4.81 12- 9, 4.88
Kile enjoyed the biggest turnaround of any pitcher in the major leagues. But Hentgen and Benes virtually matched their 1999 ERAs, and their records improved only because both received more run support than they had in 1999.
Who had the best ERA among Cardinals starters? Certainly not Benes or Hentgen. Not Kile, whose 3.91 mark ranked 16th in the National League. Not Garrett Stephenson, who went 16-9 but posted a 4.49 ERA. In terms of ERA, the best St. Louis starter was rookie Rick Ankiel, whose 3.50 ranked ninth in the NL, right behind Tom Glavine. And speaking of young pitchers, rookie Britt Reames came up late in the season, started seven games, and posted a 2.88 mark.
Reames (traded to Montreal) and Hentgen (signed with Baltimore) are gone, the latter replaced in the rotation by Dustin Hermanson ... a 28-year-old veteran who posted a 4.77 ERA last season. If the Cardinals are to be a truly outstanding team, La Russa and Duncan will have to continue the development of Rick Ankiel and get quality work from the veteran starters.
Can Kile do it again? Last season was his 10th in the majors, yet only the second in which he won more than 15 games. The Cardinals rotation is neither strong nor deep, which means somebody has to come through with a big season, as Kile did last year. Ankiel seems like the best bet, but of course he's a question mark following his postseason meltdown.
There are other questions about this club.
Can Craig Paquette and/or Placido Polanco and/or Shane Andrews hold down the hot corner? With the trade of Fernando Tatis, these two are supposed to compete for the third-base job this spring. But Paquette and Andrews both will struggle to keep their OBPs above .300, and Polanco will struggle to keep his slugging average above .400. And all three bat right-handed, so any sort of standard platoon isn't an option.
At 30, Jim Edmonds established career highs in virtually every important statistical, and is an obvious candidate for a big-time decline in 2001. Remember, too, that throughout his career Edmonds has had trouble staying healthy. Speaking of staying healthy, can Mark McGwire play 140 games? Because this year the Cards won't have Will Clark if (when?) McGwire goes down.
The Cardinals won 95 games last season, but it seems highly unlikely that they'll reach that lofty total in 2001. But let's say they drop to 89 wins (a reasonable assumption); is anybody else in the National League Central good enough to beat them?
The Cubs, Brewers and Pirates are all long shots. Astronomical, Keanu-Reaves-winning-an-Oscar shots.
The Reds shouldn't be completely discounted, because their lineup includes Ken Griffey, Barry Larkin and Sean Casey. But their rotation may include Elmer Dessens and Osvaldo Fernandez.
So that leaves the Astros. Yes, they were incredibly unlucky last year, and they're likely to improve by seven or eight games due to that alone. That gets the Astros to 80 victories, but what about the other 10 they'll need to challenge the Cardinals? The presumed returns to good health by Craig Biggio and Billy Wagner will help. And Jose Lima certainly can't be as bad as he was last year. Will all that be enough? The Astros finished 23 games behind the Cardinals last season, and that's too many to make up unless one team or the other had undergone a major overhaul since September. And neither team has. Houston certainly has a chance, but St. Louis certainly must remain the favored club in the NL Central.
All (or most) things considered, I give the Cardinals a 45 percent chance of winning the Central, the Astros a 35 percent chance, and the Reds a 10percent chance, with the other three clubs dividing up the remaining 10 percent.
| |
 |