Winners
Dayn Perry, baseball writer for Fox Sports and Baseball Prospectus - as well as an
occasional Primate – has added book writer to his resume. The title of his debut tome
best expresses its purpose:
Winners:
How Good Teams Become Great Ones (and It’s Not the Way You Think). Though
the concept inspiring, it is similar to 2003’s
Paths of
Glory by Mark Armour and Dan Levitt. Still, enough differences between the two exist to
prevent a reader of the elder book from experiencing déjà vu while flipping through Perry’s work.
Whereas Armour & Levitt mainly structured their book by looking at a dozen teams over
the last 100+ years, Perry approached the subject by doing a group analysis of examines
the 124 play-off teams from 1980-2004 (except for 1981 when Bowie Kuhn had a
psychotic break and thought MLB was the NFL, and 1994 when there was no play-offs)
to reveal what underlying traits these squads had in common. Also, Perry organized it
thematically instead of the other’s team-by-team approach, with chapter like “The
Slugger,” “The Closer” and “The Money Player” in
Winners designed to show how
important these aspects of a team are, and what makes them important.
A crucial fact to recognize about the book is its target audience. Perry is not writing aiming
at the previously converted stathead, but rather at the general baseball fan. He uses
sabermetric stats – VORP, SNLVAP, DER, DIPS and so forth – but he has no intention
of having in-depth discussions of the math behind them. At the outset of the book he
even jokingly refers to them as acronyms which sound like German profanity. This does
force him to walk a tightrope, though. If he gives too much mathematical background on
these Reich curses, then his audience will find themselves wishing they were reading the manual
for their VCR. However, if he gives too little, then they will not have any idea what he is talking
about. Ultimately, he performs this high-wire act admirably. At some points a little
more background for the reader would help, but on the whole the book does not fall apart.
The chapters focus on pitching (three actually, one for starters, closers, and middlemen,
respectively), hitting, fielding, base stealing, the front office, team age, and luck. That is
pretty good coverage, but he did not include anything on managers. Though it is not a
crippling oversight, it would be interesting to see how the managerial tendencies of
October-bound teams compared with other teams.
In each chapter, Perry spends most of his time regaling the reader with tales of the teams
and their players. Perry is in his element with these colorful anecdotes because he is a
wonderful writer with a nice way with words. He peppers the pages with wry
observations and amusing descriptions. He characterizes Will Clark’s 1987 stolen base
percentage “an international incident on the base paths,” brushes off Deion Sanders as “a
novelty outfielder,” rightfully claims that Dennis Eckersley looked more like a bassist for
Foghat than a baseball player, describes the defensive ability of the 2000 A’s as “tastes
like . . . triples!” and my personal favorite – says Johnny Damon’s unwashed look
appeared to be “the residue of some unholy coupling of a porn star and a neo-Luddite
hostage taker.”
Even aside from the humor, the man can tell a story. Each chapter tells a host of stories
about players and post-season teams, and all keep the reader’s attention. One can relive
the glory of a young Dwight Gooden as well as his sad aftermath. Or remember the achievements
of minor-leaguer-skipping Bob Horner. Finally, he feels no need to “dumb down” his writing.
He has a large vocabulary, and he knows how to use it. If that means people have to dive to their
dictionaries to look up penumbras, porcine, endomorphs, opacity, patois, and oeuvre, or wonder what
Plato’s world of forms were, or what it means to be a “hydra-headed center fielder”, then so be it.
While this could cost him readers as some might feel too much goes over their heads,
Perry should not be criticized for wielding a wide array of words. Instead, he should be applauded
for his willingness to assume that his readers possess more than an eighth grade education.
As for the analysis itself, much – though not all of it – will be rather familiar to the
typical Primate. Stolen bases are overrated, but speed is valuable because it can help the
team in other ways. You can learn a lot about a pitcher based on his defense independent
stats. Errors and fielding percentage are overrated, but range and getting to the ball are
underrated.
Though the research is not path breaking, at times it is both interesting and thought provoking.
One of his most interesting bits of research concerned the trade line. He contends, and
convincingly shows that though most playoff teams do make trades, their impact is often
minimal, and the best trades are often the ones done well-before the deadline. He also
makes a fantastic case for Walt Jocketty as the premier in-season dealmaker. In
discussing closers, he demonstrates that while modern day ninth inning specialists’ value is
lessened by their paucity of innings compared to the firemen of yore, their leverage is
actually higher. He makes an interesting point on age, showing that play-off teams are a
bit older than other teams on average. Speaking as someone who thinks that the
sabermetric community tends to be too dismissive of anyone over the age of 30, I liked
that point. Even in some of the topics that are very familiar to Primates, he makes convincing
arguments. For example, while arguing that stolen bases are overrated, he does a nice job
using Rickey Henderson’s career to show how little of his worth as a player came from
those mountains of stolen bases.
Perhaps his most interesting bit of analysis comes in the section on sluggers. Here he
stands sabermetric orthodoxy on its head by arguing that SLG’s importance exceeds that
of OBP. For evidence, he divides up baseball history into a series of 25 year periods
(though the first and last periods are both a little longer than a quarter century) and shows
that in the current age, almost all proceeding periods, and the entirely of MLB history a
better correlation exists between winning and SLG than with OBP. Very interesting stuff, but
he does not make his case as strongly as he could. At other points in the book, he briefly
delves into a statistical analysis controversy or changes in opinion within the
sabermetric community on given topics (such as the importance of defense for the overall
success of a ballclub). These discussion really need more elaboration to summarize the shifting
sabermetric positions, because Perry argues completely against the wind. Alas, his only depiction
of the pro-OBP position is a reference to Michael Lewis’s Moneyball. Yup, that’s the only
passage. Ah well. On the whole, his analysis is good, though (for better and for worse) no one
will ever confuse Dayn Perry with Tangotiger or mgl.
Though Perry masterfully handles the stories and provides good analysis, the whole is
less than the sum of its parts in Winners. The stories, though always entertaining,
generally do very little to bolster his points. Exhibit A: his discussion of the 1987 Tigers
in his chapter on defense. Perry’s analysis centers on how errors are overrated because the crucial
part of defense is getting to the ball, rather than avoiding kicking it around. To this end, he
reasonably champions park adjusted Defensive Efficiency as a superior metric for defensive
performance. In the chapter he provides a list of the worst defensive play-off teams since
1980 according to park adjusted DER. The 1987 Tigers are one of these teams. He then
spends the last fourth of the chapter discussing how many players on that Tiger team are historically
underrated. I have no disagreement with his assessment of Trammell as a deserving Hall of Famer,
or that it is a joke that Lou Whitaker fell off the ballot on his first try, or
that Chet Lemon and Darrell Evans are underrated. Also, he does a good job telling their
stories – heck, his depiction of Evans and the UFO is one of the highlights of the book.
BUT – if his main point in the chapter is to point out the importance of getting to the ball,
why does he spend a half-dozen pages singing the praises for the players on a team
who, by his very own defensive metric of choice, did an exceedingly bad job of getting to
the ball? The ’87 Detroit stories not only don’t back up his point, if anything they
undermine his key argument in the chapter.
Admittedly, the tale of the ’87 Tigers is an extreme example as normally his stories do not
cut against his main points, but far too often his stories seem irrelevant to his points. His
stories entertain, but he does not do very much to link them to the points he is attempting to
make except in the most basic way of “this chapter is about hitters, now here’s a three page
synopsis of the career and life of a guy who could hit.” Immediately after reading some
chapters, one has to flip back through it to find Perry's original point as it got lost in the
thick descriptions of Ted Turner’s rise to billionare-dom, or Pedro Guerrero’s pathetic post-career
life. Hell, his chapter on middle relievers may not have had a point beyond arguing that middle
relievers who get guys out are better than ones who do not. There is nothing wrong with a having
a book which does little more than tell tales. Everyone reading this review can think of at least one
personal favorite baseball book composed of just good stories but no real point. BUT, if an author
structures a book and presents its material in a way that leads the reader to believe there is
a central point, the stories dang well better be more than enjoyable larks filling up space.
Aiming for a wider audience is no excuse for not relating his tales to his main point.
There is a difference between loosely making an argument, and barely making a point.
Stories can be interesting and serve a larger goal. A writer should take the occasional tangent.
Much of the best writing from Bill James or Tom Boswell and countless other writers come when
they momentarily veer off, but off-topics discursions only work when there is a main point consistently
being made. In Winners, nearly eighty percent of each chapter is stories, while the remaining
twenty percent is the supplementary analysis, with the two only tying together in the broadest possible way.
Mostly, each chapters read like two separate articles written on the same general topic, but with no
unifying theme. It makes for a frustrating read. The stories are not always unrelated, but rarely clearly
pertain to whatever point Perry is trying to make. Too often, one can finish the book and not quite
be clear what his main ideas where. One would easily be able to talk about Kevin Mitchell,
and Bert Blyleven, and Rickey Henderson, and Dave Dravecky, and a host of others
though.
Winners is a tale told by a wordsmith, full of verve and wit, but signifying little.
Ultimately, it is worth a read from the library or while perusing in a Barnes & Noble, but the reader may
wonder if he should have really forked over the cover price.
In what appears be a very busy year for baseball books, this one probably
should not be your top priority unless you really like reading stories about championship teams. I don’t want to sound too harsh, as I did genuinely enjoy reading it. Winners" is
good, but not great.
Chris Jaffe
Posted: May 08, 2006 at 07:22 AM |
10 comment(s)
Related News:
Books
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
Did Perry get into "playoff experience"? We hear this alot from analysts that teams with experience in the playoff do better. Also, I am curious is there any correlation to an individuals' experience to performance. Do playoff virgins flop?
- IIRC, the teams analyzed are from 1980-2003, not 2004 as stated in the article.
- The SLG>OBP study is based on a correlation with runs scoring, not winning. I think. My memory's a bit hazy at the moment.
Run Joe Run - Perry pretty much sticks to looking at how the teams did in the regular season.
Probably less in-depth than you would like, but ESPN.com ran an article not long ago that was excerpted from BP's new book about what historically makes teams more successful in the post-season. I don't recall the specifics off-hand, but you could probably search for it on their site. It was titled, I believe, "Why Billy Beane's @#*& Doesn't Work in the Playoffs." It was pretty interesting.
Wouldn't matter. It's true in either case that SLG has a higher correlation with either run scoring OR winnning.
Just doesn't mean what Dayn thinks it does.
Analysis from straight correlation will lead to a lot of nutty conclusions.
Doubles have a higher correlation with run scoring than triples do. Doesn't mean that doubles are more important than triples.
Hell team strikeouts has a substantially higher correlation with team runs scored than triples. (.33 for Ks .10 for triples) Nobody would use this to argue that your better off to strike out than to hit a triple.
I'm simplifying, but the reason that SLG shows up with a better correlation is that when you attempt to model team runs scored with SLG you are in effect guessing OBP and vice-versa and the standard deviation of SLG is more than twice that of OBP.
If you want to know the relative importance of OBP and SLG the right way to do this is by regressions.
And for the period 1955-2001 (what I have handy), a single point of OBP comes out about 1.6 times as important as a single point of SLG. The relative importance rises with offensive levels though I haven't figured out how to model this yet.
This simple model is pretty good. Standard error of 25 runs per 162 games. Just under 93% of the variation in team runs scored is explained by properly weighted OBP and SLG.
Are you accounting for the cross-correlation of OBP and SLG when you do this?
-- MWE
There was a thread about this a month or two ago...
No significant change in the relative value of OBP and SLG. (And the model is a reasonably good one. Slightly worse than using runs per game but that's hardly surprising since OBP and SLG don't perfetly model scoring)
So, since Dayn was talking about good teams I eliminated all teams with a winning percentage less than .556 (IE less than 90 win teams)
The relative value of OBP to SLG was higher for the 90+ win teams.
So then I eliminated the 90 win teams (and only them) that didn't make the playoffs.
And found that for those teams, while a single point of OBP is still substantially more important than a single point of SLG, the relative weight of OBP to SLG is lower for these teams than in general. Almost spot on 10% lower.
Interesting. Don't know what it means (OBP heavy teams that win precisely 90 games are unlucky in one way to put it, but that's only one way to look at it)
You must be Registered and Logged In to post comments.
<< Back to main