The Hidden Game of Baseball, 2015

Hidden Game of Baseball 2015Three decades ago, Pete Palmer and I wrote The Hidden Game of Baseball (Doubleday, 1984), aided greatly by the editing skills of our friend David Reuther. This month The Hidden Game has been reissued by the University of Chicago Press (440 pages, $22.50), in facsimile except for a new tabular appendix, a fine new foreword by Keith Law, and a thirty-years-after introduction that Pete and I wrote together; here it is.

The statistical side of baseball has always gripped me. I believed that in numbers one might uncover truths not visible to the naked eye, in the way that flying at night a pilot will learn things from the instrument panel that his senses can’t show him. In the summer of 1981, I was on assignment for The Sporting News. I went to my first convention of SABR—the Society for American Baseball Research—walked into a reception area, and met Pete Palmer. Pete, I quickly realized, was the best at what he did, which was to think hard about baseball and its numbers. Pete became my dear friend and more or less constant collaborator over the next 20 years.

But our first collaboration was not this book. With David Reuther, Pete and I developed an idea for a new sort of encyclopedia that would provide more revealing stats and tell better stories than the landmark books in the field at the time, which were known as ICI/Macmillan (1969) and Turkin/Thompson (first published in 1951). We called it “Complete Baseball,” I think, and we received a handsome bid for it, but the schedule demanded by the publisher was unworkable. So we walked away from what was at that time very big money and took much less to create The Hidden Game of Baseball, which came out in 1984. (The sort of encyclopedia we proposed did not come out until 1989, as Total Baseball).

We had no idea what impact Hidden Game might have, but our publisher certainly hoped we would enjoy some measure of the success Bill James was having with his first commercially published Baseball Abstract. Bill, of course, was one of the pioneers of what came be known as sabermetrics. He had been releasing his Abstracts annually, focusing on the season just past and the prospects for the next and including essays that articulated his inimitable take on baseball’s statistics and how they might be improved. Like Bill, we had been interested in developing measures that tied runs scored and allowed to player performance—we felt that those numbers were demonstrably related to the outcome of a game or a season. Bill’s best measure, modified over the years, was called Runs Created. Pete’s was Linear Weights, which you can read all about in this book.

I say “Pete’s” rather than “ours” because he was the statistician while I was the historian; he was the genius, I was the explainer. The conventional wisdom about Hidden Game has been that Pete did the numbers and I did the writing. That notion is more right than wrong, but Pete’s words are presented and reflected throughout the book and, oddly, so is some of my statistical noodling. As with any successful collaboration, presumed areas of specialty don’t stay sharply defined for long. Still, none of the innovative measures in Hidden Game may be called mine. I have never been a statistician, though I have been called one. All the same, Thorn & Palmer or Palmer & Thorn have endured as a pioneering sabermetric tandem because of Hidden Game and our subsequent work together.

The hidden game is the one played with statistics. It raises important questions about why we measure, what we think we are measuring, what we are truly measuring, and, most important, what the measurement means. Such questions informed our thinking throughout this book more than thirty years ago, and, even as Big Data and refined statistics sharpen our focus with each new season, sabermetricians today still cannot stray far from them. We were not the first to think unconventionally about baseball statistics, and we were careful to lay out their history from the 1840s on, and to credit those who had innovated in our field long before us. In the original acknowledgments, we even invoke Bernard of Chartres.

Mike Schmidt and Hidden Game in SPORT, 1984.

Mike Schmidt and Hidden Game in SPORT, 1984.

Bill James has remarked that a meeting of sabermetricians at, say, a SABR convention in the early 1980s could have been—and more or less was—held in a hotel room. We were barely a tributary, miles from the mainstream. The chapter titles we chose then reflect the windmills we felt compelled to tilt at. It was much harder back then to convince baseball professionals and beat writers that what we were saying held any water. And yet, now it’s hard to find a baseball professional who does not see the value of analyzing all the data that are available to us.

As general managers and managers came to understand that outs and runs are the currency of the game, as they always have been, they began to value on-base percentage, which measures not just the hits that a batter gets but all the ways he gets on base—and the hidden value of not using up an out and permitting another man to bat with a runner(s) on base. Keeping track of pitch counts was not merely a way to preserve your own pitchers’ arms—it was also a weapon: By having his batters work counts, a manager might force the hand of his opposing number and sooner get to the middle relievers, who are the soft underbelly of every pitching staff.

Today, the thinking in baseball has changed so much from thirty years ago that it is probable that we now overvalue walks where formerly they had been undervalued. Similarly, we scorn risky baserunning, when once it was the prime delight of players and fans. The charm of the grand old game is that it appears to be the same as it ever was, or at least the same as in President McKinley’s day, but of course it has changed radically. In terms of strategy the game is now hardly about baserunning and fielding at all, though recent sabermetric work in these areas may alter the balance yet again.

As much as things have changed, we do think this book can still boast of its own achievements and lasting contributions. Tying individual statistics to team accomplishment—restating batting, pitching, and fielding records in runs scored or saved—still seems worthwhile. Restoring baseball statistical thinking to the 1860s core of the game—securing or conserving outs—was good. Pete came up with the first “Unified Field Theory” of baseball: The Total Player Rating, with all players’ offensive and defensive contributions measured in runs above or below average, with league average performance defined as that which, when aggregated, would produce a .500 record for a team. This baseline troubled some of our colleagues, who contended that Hall of Fame players like Lloyd Waner or Tommy McCarthy could not possibly have been worse than league average over their long careers, as our calculations revealed. The current sabermetric standard is Wins Above Replacement, with some differing notions of what a replacement player (i.e., a somewhat below average one that any team might employ) might look like. Call us old fogies, but Pete and I still think a team of league-average players producing a league-average result (81-81 over the course of a modern season) sounds about right.

We have entertained offers over time to update and revise the original edition of this book, but we think it is better to leave it as it was, a stone along the road to a much greater understanding of how the game might best be played and who has played it best. (Pete has provided a list of the top 500 players of all time as of 2014, though, which appears as an appendix.) The updating, revising, and improving has been better left to the formerly tiny but now vast sabermetric community.

Bill James Baseball Abstract, 1982.

Bill James Baseball Abstract, 1982.

Still, how might we have approached Hidden Game differently—say, if we were to write it afresh today? When we wrote this book, play-by-play data were only beginning to be kept by the Elias Sports Bureau, and retrospective play-by-play had not yet been compiled by Retrosheet. We were compelled to develop our measures based on computer simulations and partial play-by-play. We would benefit from the work reflected at, Baseball Prospectus, FanGraphs,,, and so many other websites. We could not ignore the advances of the digital age: live data capture through time-stamped video. PITCHf/x provides pitch trajectory, velocity, and location data, and FIELDf/x tracks all moving objects on the field: fielders, runners, umpires, balls. Our run values were the product of simulations; today those values may be tested against reams of play-by-play data, and they would be slightly different—not so different, however, as to alter any of our basic findings and tenets. More data bits may be available after a single game today than were available to us as of 1984 for all baseball history, but is our understanding of the game radically altered? Or is the way we play it substantially different? Unbalanced defensive alignments—shifting infielders around to compensate for hitters’ directional tendencies—are a novel reaction to data, for which in time there will be a counterreaction. Baseball is an entropic game.

Yet analytics are here to stay, and it is fair to say that the best constructed clubs—the ones that are in contention year after year—are not just the teams with the most money to lavish upon talent, but the teams that spend wisely and exhibit patience with their young players. It has been ever thus. The backlash against sabermetrics, present to some degree as soon as Bill James began to be widely read, is different from the one we experienced in the 1980s.

Most fans believe the game’s useful history begins with when they first started playing it or watching it. In my household, as my three sons grew up in the game, there was always talk at the dinner table about Ken Griffey Jr. and Greg Maddux and Mike Schmidt—and Babe Ruth and Cy Young and Ty Cobb, too. They were all part of the game. Indeed, they were all part of the family—more so than distant cousins and aunts and uncles. We talked about who was better than whom, what Cobb might do if he had to face Maddux, how many homers Ruth would hit today, what Griffey’s OPS might have been against 1920s pitching staffs, that sort of thing.

Baseball fans of earlier generations had fewer statistics at their disposal, but a simpler game perhaps had less need of them. Ultimately, the statistical fragments that were once saved in scrapbooks, or the new measures devised by ingenious fans, become relics that remind us at every moment that our youth was a wonderful if remote time.

Hidden Game of Baseball, 1984.

Hidden Game of Baseball, 1984.

Cory Schwartz of Major League Baseball Advanced Media has said: “I’m old enough to remember when we had to wait two days to find West Coast box scores in the newspaper, and wait until the Monday and Tuesday editions of USA Today.” Pete and I are older than that, and we recall some of the individuals who were tilling this field before us. We are in a bold new Age of Enlightenment, but fans and writers are not unanimous in believing that we are in a new Age of Enjoyment.

Stats contain and crystallize stories but are not stories in themselves. They are something of a fetish, an encapsulation of a thing once alive. A stat serves to recall and revivify the past, and sometimes to transform the future. As fans, Pete and I both follow baseball as closely as we ever did. But sabermetric writing lies more behind us than ahead, and not only because we are nearer to life’s ninth inning. Amid today’s mix of straight-on game account and metric analysis of who is better than whom, we miss the fun that made us come to love the game in the first place.

For this we could blame Bill James, and ourselves too. Early on, what interested us more than fiddling with formulas or lobbying for Dick Allen to enter the Hall of Fame was the web of illusion that stats created for fans and players alike, evading more interesting theoretical or philosophical questions. Read Hidden Game in that spirit, the one that spurred us thirty years ago, and we think you will be rewarded. Others may say better than Pete and I what Hidden Game has meant, but for us it may be simply that it continues to be sought and cited, all these years later. With this reissue, no longer will fans need to scour antiquarian book sites to luck upon a copy.


As always, a very informative and fascinating read. It leads me to wonder if y’all have ever considered — let alone had the data to at least partially analyze — the extent to which Connie Mack with his scorecard and John McGraw among others had an intuitive feel for Sabremetric type impact on the game and the extent to which they used it?

Oh, sure.
Apart from McGraw’s use of platooning and pioneer use of relief pitching, he deal tin probabilities just as Mack did, with his sometimes daring selections of starting pitchers in the postseason. They used data for sure, even if not overtly, just as Earl Weaver, that famous basher of sabermetrics, did.

In the Managers Roster in Total Baseball we used the concept of Expected Wins, based on a team’s won-lost record as predicted by its run differential (not its actual W-L record). Pete Palmer wrote:

W-EXP Expected Wins Calculated for the team based on its actual runs scored and allowed, not its predicted runs scored and runs allowed.

A team that allows exactly as many runs as it scores is predicted to play .500 ball. The equation for expected wins is:

(Runs Scored – Runs Allowed + 81)/Runs Per Win

A-E Actual Wins Minus Expected Wins (A measure of the extent to which a team outperformed or underperformed its talent; for a single season or two a high figure may be attributable to chance, but over time one must credit good managing.)

BTW, we updated this formula midway in the run of Total Baseball to:

Expected Wins (Calculated for the team based on its record in the previous three years, according to Bill James’ formula [1/2 * last year’s winning percentage + 1/8 * winning percentage the year before that + 1/8 * winning percentage the year before that + ¼ * a .500 record], all multiplied by the number of games played.)

Thanks for the responses. I confess I have not been totally familiar with y’all’s ranking work. I have reviewed a few articles on the internet and might have some intuitive comments if I had more time to delve into it. And while very interesting, my question, poorly worded as it might be, is aimed more toward your opinion as to whether the decisions of the old time managers might have been guided by the managers’ feel for players’ statistical propensities — Mack’s famous defensive positionings being a ready example.

I loved reading your take on statistics in today’s player and fan world. You bring that lovely under-stated humor to every paragraph. Thank you.

Thanks, Fran!

Great read, John. Seeing “1984” brought back some memories. I recently saw an ESPN Classic showing of the Padres-Cubs 1984 Game 1 NLCS. The umpires were on strike, and Peter Ueberoth was in his first day(s) as commissioner. I remembered all of this because I left school early that day for a doctor’s appointment, thrilled at the timing because it meant I would just get home in time to see the game. In the waiting room was a Reader’s Digest with the title “1984 is here: Where is Big Brother?” and that was the first time I learned about any of that theme. Of course Apple’s famous Super Bowl ad had appeared earlier that year, and the beginning of the PC had begun. To think your book was originally written during that time, and all that has changed since then, is fascinating. Even back then, very few markets broadcasted every game of their local team, and I grew up with the radio and NBC’s Game of the Week, and I depended on This Week in Baseball for highlights. And looking up career stats was something I did using the back of my baseball cards.

This probably doesn’t add much to the discussion, but it brings back memories and makes me appreciate books like yours that were well ahead of their time.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: