Wednesday, March 02, 2005

Questioning baseball memories

It's always fun when two of my major interests, history and baseball, overlap. I'm just about done reading Doris Kearns Goodwin's Wait Till Next Year, a memoir of growing up a (Brooklyn) Dodgers fan in the '50s. It's a good read, and timely, what with spring training up and running. My biggest complaint is that the nostalgic tone gets in the way of an accurate account of how things actually were. The following (baseball!) example is indicative.

For all of us, the love was personal and familiar. We spent arguing about whether Duke Snider, Willie Mays, or Mickey Mantle was the best center fielder. The handsome, smooth-fielding Duke Snider was the most consistent home-run hitter of the three, but Mays had a balletic grace and a joyful fury, while the switch-hitting Mantle had the greatest raw power and speed.

Now, this may be how Kearns Goodwin remembers things, but her description of their respective abilities is off. For those of you who don't care about baseball statistics but are more interested in my take on Kearns Goodwin's nostalgia, skip down a bit until the numbers disappear. Otherwise, read on.

Kearns Goodwin makes several claims about the baseball abilities of Snider, Mays, and Mantle here. To sum them up:

-Duke Snider: "smooth-fielding", "most consistent home-run hitter of the three"
-Willie Mays: "had a balletic grace and a joyful fury"
-Mickey Mantle: "had the greatest raw power and speed"

Some of these descriptions have quantitative analogies. "Smooth-fielding" indicates mistake-free defense, so we're looking at fielding percentage (FP). Hitting home runs (HR) is fairly straightforward, though you could quibble about the precise metric to use. There's no obvious way to quantify "balletic grace and [...] joyful fury". There are a few stats that measure power: slugging percentage (SLG) and isolated power (ISO, calculating by subtracting batting average from slugging percentage) being the most obvious. Measuring speed is tougher, but stolen bases (SB) and triples (3B) seem to correlate well with footspeed. There's also a fairly obscure stat, the Power/Speed Number (P/S#) measuring, you guessed it, power and speed.

So we'll be looking at HR, SLG, ISO, 3B, SB, P/S#, and FP. Let's see how our three heroes fare.

The period in question is the time span covered in the book, 1949-1958. Right away we run into a problem: Mays and Mantle didn't enter the majors until 1951.* So we won't be able to compare the absolute numbers. Instead, we need to look at averages for some of the stats.

Home runs
Snider: 326
Mantle: 249
Mays: 216

Unsurprisingly (since he played two more seasons), Snider holds the lead in this category. But it's not very fair to compare this way. how about if we look at how many home runs they hit per season played?

Snider: 32.6
Mantle: 31.1
Mays: 30.9

Things are a whole lot closer now. Snider still has the edge, but not by much. But wait! Mays only played 34 games in 1952; Mantle only played 96 in 1951, his rookie year. A better way to measure consistent home-run hitting is to check how often home runs are hit.

Mantle: 15.8
Mays: 16.2
Snider: 16.6

Uh-oh. Snider doesn't look like the home-run king anymore. On average, Snider went to bat 16.6 times for each home run he hit, worse than both Mays and Mantle. Put another way, had Mays and Mantle gotten the same number of at-bats as Snider (and maintained their home-run rate), they would have hit more dingers than the Duke of Flatbush. Incidentally, if you look at the career numbers for each, Mays and Mantle have the same 162-game average for home runs, 36. Snider is down at 31.

While Duke Snider hit more home runs than either Mays or Mantle from 1949 to 1958, he wasn't the most consistent home-run hitter of the bunch; Mickey Mantle was.

Mays: .591
Mantle: .577
Snider: .564

Mays: .274
Mantle: .262
Snider: .258

Recall that Kearns Goodwin wrote that Mantle had the "greatest raw power". This simply doesn't square with the numbers. Willie Mays led Mantle and Snider in both slugging percentage and isolated power. Mays had the most power of the bunch, not Mantle.**

Mays: 152
Mantle: 77
Snider: 60

Mays: 74
Snider: 62
Mantle: 50

Stolen bases are a mark of speed because you need to be quick to beat the catcher's throw, triples because it's not easy to run the 270 feet from home plate to third base in the seconds that pass between hitting the ball and it being returned from the outfield to the infield. Mays absolutely crushes Mantle and Snider in stolen bases, and has a considerable lead in triples. And remember that these are the absolute totals; throw in the fact that Mantle and Snider played a lot more games than Mays in this period, and it's obvious that Willie Mays was the fastest of the three.

Power and Speed
Mays: 178.4
Mantle: 117.6
Snider: 101.3

Yup, when you consider power and speed together, Willie Mays comes out on top, not Mantle. No surprise here.

Snider: .986
Mays: .981
Mantle: .979

Finally, one that Kearns Goodwin got right! If we take "smooth-fielding" to mean handling balls cleanly, fielding percentage is clearly the way to go. Duke Snider was, in fact, the most "smooth" fielding of the three.

There's a whole lot more to fielding, of course, than avoiding errors. As you might guess, Mays's speed allowed him to track down a lot of balls in center field that Snider and Mays had no chance at. But Kearns Goodwin made no claims about Snider being the best fielder of the three, just the smoothest, so no need to bring up range factor, zone rating and all that.***

Let's sum up.
"smooth-fielding" - Duke Snider, as Kearn Goodwin wrote.
"most consistent home-run hitter" - Mickey Mantle, not Duke Snider.
"raw power" - Willie Mays, not Mickey Mantle.
"speed" - Willie Mays, not Mickey Mantle.

Of the four quantifiable abilities Kearns Goodwin attributes to Snider, Mays, and Mantle, she only correctly assigns one of them (the defensive one, which turns out to be the least significant of any of them). And, let's be honest, any discussion of these three guys that doesn't immediately place Willie Mays and Mickey Mantle several steps above Duke Snider is way off-base. There's a reason why Mays and Mantle were inducted into the Hall of Fame in their first year of eligibility, while it took Duke Snider more than ten tries. Mays and Mantle were the best center fielders of the 1950s, if not of all time.

Whew. That was a lot of number crunching just to show that someone's memories of the 1950's aren't entirely accurate. This isn't an isolated case. There are plenty of anecdotes and descriptions throughout the book that don't quite ring true. And not just baseball ones; I simply chose this example since it's a clear-cut case where Kearns Goodwin's description of the past is inaccurate. Well, that and I like fiddling around with baseball statistics. The point is that Kearns Goodwin doesn't get the history right.

Why does all this matter? After all, memoirs aren't exactly known for their accuracy (I'm sure there are plenty of pithy things people have said about memoirs; sadly, I don't seem to know any of them). And what's the problem with looking back fondly on growing up a fan of the Brooklyn Dodgers? Why I'm making such a fuss about a nostalgic memoir that gets the details wrong. There are a few reasons.

If I were feeling partisan, I'd point out that people like Maggie Gallagher (yes, that Maggie Gallagher) have seen Kearns Goodwin's book as a bulwark against "the tenured radicals [who] attempt to rewrite our nation's history". Yes, I know, the canard about tenured radicals is a sure sign that you're dealing with a right-winger who doesn't know anything about academia. While nostalgia is generally harmless, it can have grave consequences when political conservatives play up some golden age of the past (typically the result of some post hoc gilding) to support reactionary policies. But I'm not sure what golden age Gallagher sees in the book. While Kearns Goodwin describes her youth as a time of naivete (neighborhood kids playing in the street, big to-do when the first family on the block gets a television, and the like), that naivete is destroyed by the end the narrative as Kearns Goodwin's mother dies and she discovers that her grandfather committed suicide, along with a growing awareness of the civil rights movement. While Kearns Goodwin's version of the early 1950s is generally glowing and happy-go-lucky, it's pretty clear throughout the book that not everything turns out well. Truth be told, it's hard to see Gallagher's comment about "tenured radicals" as anything but an irrelevant jibe against academia.

So it's not that I'm upset with Kearns Goodwin for providing conservatives with fodder. She doesn't; conservatives are misreading "Wait Till Next Year" if they view it as a panegyric to the 1950s and a defense of the decade against historians who have reminded us that '50s America was not a utopia.

My real complaint about Wait Till Next Year is far more esoteric. The problem with Doris Kearns Goodwin's book is that she uncritically accepts her memories as accurate records of the past.

I have no problem with her remembering that Mickey Mantle had more speed than Willie Mays or that Duke Snider was a more consistent home-run hitter than Mays and Mantle. I do have a problem, however, with her presenting those statements as historical fact. There might be very good reasons why Kearns Goodwin's memories aren't accurate. Maybe she watched a Yankees game where Mantle stole two bases and hit a triple. Her praise of Snider is even more understandable; she rooted for the Dodgers, so it's natural that she saw its stars in an overly positive light. I'm genuinely curious about why she remembers things this way.

Miriam Burstein and John Holbo have recently complained about academics overusing "interrogation," but I think this is a clear case where a bit of interrogation is exactly what needed. Memoirists (what an awful word) should do more than simply ask themselves, "What do I remember?" And they should even go beyond, "Is what I remember accurate?" They should ask themselves a far more interesting question: "Why do I remember this the way I do?" It's not an easy question, but the hard questions always seem to have the most fascinating answers.

Doris Kearns Goodwin is an historian, albeit one who has run into some problems in recent years. She's capable of source criticism, and she's smart enough to realize that memories are just as much historical sources as diaries, newspapers, and photographs. As a historian writing a memoir, she's well-equipped to answer the questions posed above. My disappointment with Wait Till Next Year lies in Doris Kearns Goodwin's naive acceptance of memories as facts. Her memoir could have been an insightful exploration of the nature of memories. Instead it's an engaging, if straightforward, account of childhood, baseball, and the '50s. Nothing wrong with that, but I'd expect more from someone of Kearns Goodwin's stature.

*Mays didn't play in the majors in 1953. I'm not sure whether this was due to injury or he spent the year in the minors. So we're dealing with 7 years of Mays data, 8 years of Mantle data, and 10 years of Snider data.
Update (8 March 2005): Reading this Hardball Times article by Steve Treder, I've discovered that Willie Mays missed most of the 1952 and all of the 1953 season due to his service in the Korean War. The forgotten war indeed; you see a player missing stats for, say, 1942-44, you immediately think of WWII, but the Korean War didn't even cross my mind as an explanation for Mays's big-league absence.

**You could argue that Mays's slugging percentage is inflated since he hit more triples than the other two (triples are often seen having more to do with speed than power). But even if you count triples as doubles for the purposes of slugging percentage, Mays comes out on top. Besides, Kearns Goodwin claims that Mantle had power and speed going for him... no fair taking speed out of the equation to make him look better.

***Besides, I don't know all that much about defensive statistics. And there's probably not great fielding data from fifty years ago.


At Mar 7, 2005 9:09:00 AM , Anonymous Anonymous said...


I think the thing you're missing here is that Goodwin is talking about what she remembers (so right off the bat, you get a certain ability to gloss things with that proviso) but also a memory of when she was young--when you're entitled to get things like the respective abilities of ballplayers wrong. I remember thinking highly of Ron Cey and the other Dodgers (LA) of his time (except for Steve Garvey, I always thought he was an insincere prick even when I was a kid) but some of the stuff I would have said about them at the time would have been a mangled teenager parsing of sportswriter lingo.

Tim Burke

At Mar 7, 2005 3:02:00 PM , Blogger Danny said...

Right. I'm scared of even thinking about my opinions about ballplayers at that age. Howard Johnson? Dave Magadan? Yes, my baseball hopes were once foisted on the likes of them.

But I wasn't so much bothered by Kearns Goodwin's inaccurate memories as her failure to reflect upon them. That's the kind of thing I'm interested in. Her picture of the '50s didn't really illuminate much for me, so I was left wanting more. Like I said, it's a pretty esoteric complaint; it's quite possible that if she'd written the book to address my critiques, it wouldn't resonate so successfully for so many readers.

At Mar 7, 2005 4:25:00 PM , Blogger Danny said...

Also, there are plenty of other cases throughout the book where KG's memories aren't entirely accurate. I chose this one because, well, I like fiddling around with baseball stats. My point probably would have been clearer with a different example.

At Mar 8, 2005 3:59:00 PM , Blogger Steve said...

Don't want to be a sabermetric snob here, but I think this analysis is flawed without factoring in the parks these guys played in. I'm not old enough to remember Ebbett's Field, but my memories of Yankee Stadium and the Polo Grounds (pre-Shea Mets) are intact.

I don't have the numbers handy, but I would think the Polo Grounds might be an easier place to hit a triple. Homers, too, for dead pull hitters.

Also, I don't believe Stolen Bases are as sure an indication of speed. Speed is important, but base stealing is more than raw speed. Managerial philosophy might tend to hold SB totals down as well.

I don't know if this latter effect is measurable, but I think we need to have more numbers if we're going to make these arguments statistically.

At Mar 13, 2005 7:16:00 AM , Blogger Danny said...

You're absolutely right, Steve. The statistical analysis here is pretty quick-and-dirty. A more accurate analysis would have to consider park factors. has park-adjusted lgSLG stats. Much to my surprise, Duke Snider consistently comes out on top for this trio, with Mantle at the bottom for each year. It looks as if I wasn't giving Kearns Goodwin's enough credit for her praise of Snider, though her take on Mantle's power seems even more misplaced.

I'm also troubled by using stolen bases as a measurement of speed, but that and triples seem to be the best "basic" stats for measuring speed. Still, for Mays to lead Snider in Mantle in both categories while playing considerably fewer games leads me to believe that he was the fastest of the three.

A more sophisticated analysis certainly is necessary to gain a nuanced understanding of Mays's, Mantle's, and Snider's respective abilities. I'd be rather surprised, though, if that analysis vindicated Kearns Goodwin's memories.

At Mar 13, 2005 11:20:00 PM , Blogger Steve said...

Since we're talking specifically about triples, Lg SLG in the Polo Grounds isn't really what we're looking for. Lg triples would be, but that number isn't readily available on the internet, as far as I know.

An interesting way of measuring speed that I've started using is to compare the runs scored (excluding homeruns) relative to the runs scored (excluding those scored on HRs) that you would expect for an average player, given the potency of the offense around the given player. This approach has the advantage of being calculable with fairly readily available stats. I'll try to do this analysis for the first 3 full years of the players' careers and report back.

At Mar 14, 2005 3:47:00 AM , Blogger Danny said...

Cool. I look forward to seeing what you come up with.

At Mar 19, 2005 2:08:00 PM , Blogger Steve said...

Compute A = the number of times a player was a baserunner.
Compute B = the number of runs the player scored, excluding HRs hit by the
player and HRs hit by the team to drive him in. (estimate this last
based on team stats)

Take the ratio of those numbers.

For the first 5 full years of Snyder (49-53), Mays (51, 54-57) and Mantle
(52-56) those ratios are:

Snyder .348
Mays .342
Mantle .364

These numbers represent the percentage of times a player scored, given the
opportunity of being on base, when not scoring because of a homerun.

Now, we should take into account how good the team was at driving the player
in. To do this, compute the team's slugging, minus homeruns, and also minus
the stats of the players themselves:

BRO .347
NYG .315
NYY .339

At this point, it's not clear what to do with these numbers and retain
statistics with actual physical meaning. It's clear that the Yankees
were better at driving in runs than the Giants in the years examined.

(And, man, how 'bout them Dodgers! I would not have guessed that those
teams were more powerful than the early 50s Yankees.)

If you just divide the runs numbers by the teams slugging pct, Mantle
comes out better than Mays, with Snyder last (1.07 to 1.03 to 1.0), but
I'm not going to pretend that that's a statistically proper thing to do.

At Aug 26, 2007 8:54:00 PM , Anonymous Anonymous said...

1- Mays was in the military in '52 and '53.
2- Anybody who saw Mantle play will tell you he had more raw power than anyone of his time. Not everything can be reduced to stats.
3- Did you look at SB% as an indication of speed? The percentage of successful might help.
4- Mantle and Mays played CF at parks 460-480 foot CF fences.
Which requires much more ability than playing in a bandbox like Ebbets Field.
5- Did you take into account the park effects for each player? Mays hit a few more homers at home, Snider hit many more homers at home. Mantle hit more on the road.
Which indicates the park effects for all three.


Post a Comment

Links to this post:

Create a Link

<< Home