Doctoring The Numbers: Learning the Game

I’ll never forget where I was ten years ago this Monday.

It was not the kind of earth-shattering, life-changing moment that would
forever alter the world in which we live, like the moment I heard about the
Challenger disaster, or watching the horrors unfold six weeks ago. But like
the first time I saw my future wife, it was a day that would change the
course of my life.

On October 22, 1991, I saw a Strat-O-Matic card for the first time. My life
as a baseball fan has never been the same.

I had just begun my freshman year of college at Johns Hopkins University.
Even then, my passion for baseball was all-encompassing, but having spent
the previous seven years living overseas my relationship with the game was a
lonely one. My first days of college were no different; hard-core baseball
fans who had read Bill James were in short supply, and that September I went
to a half-dozen games at old Memorial Stadium, just a few blocks from the
university, mostly on my own.

So when I saw a flyer advertising for a "simulation baseball
league" that would be drafted the following day, I tore off a stub and
quickly called the contact number to learn more. At the time, I had never
heard of Strat-O-Matic. I was familiar with Rotisserie-style fantasy
leagues, but the concept of simulation gaming was completely foreign. Every
player’s performance is re-created on a card? Dice are used to reproduce
at-bats?

But it was baseball, and it was a group of guys who loved the game. When you’re
a college freshman, one of your top priorities is to fit in. This was a
group in which I would. So with a ticket to the Friday night game on
Memorial Stadium’s final weekend already in my possession, I passed up the
real thing for an imitation. Nearly a dozen guys showed up, and about half
were as new to the game as I was. The Strat veterans in the room went over
the ground rules, and by the time we finished the 26-round draft (25 players
and a ballpark) around midnight, I was hooked.

Ten years later, the relationship shows no sign of ending. Along the way I
started my own league, participated in tournaments
[http://www.startournaments.com] all over the country, and met people of
every different stripe, bonded only by their love of the game.

Along the way, something completely unexpected happened: the more I played
Strat-O-Matic baseball, the more I learned about the real thing. The
following truths about baseball were burned into my brain:

• The heart of the game lies in the matchup between the pitcher and the
  batter. The importance of that statement can not be overestimated. The
  result of every at-bat is determined by how these two players stack up. In
  Strat-O-Matic, each at-bat’s result is selected from one of six columns:
  three of these columns exist solely on the hitter’s card, and three on the
  pitcher’s card. The result is a natural 50/50 balance between offense and
  defense.

  This 50/50 balance has been proven correct by study after study–teams with
  good offenses but poor pitching staffs do just as well as teams with great
  pitching but an impotent lineup–but a strong sentiment that pitching is
  somehow more important still lingers in baseball. In Strat, as in real
  baseball, good hitting can beat good pitching. And vice versa.

• Performance is independent of context. Strat-O-Matic devises each
  playing card to perform as closely as possible to the player’s real-life
  performance, but in doing so the game company assumes that the context for
  each player remains the same as in real life.

  For example, in 1990, Joe Carter had one of the great fluke RBI
  seasons of all time: despite a .232 average and just 24 home runs in more
  than 600 at-bats, he amassed 115 RBIs and cemented his reputation as a run
  producer. Those naive enough to fall for this failed to notice that batting
  ahead of Carter were Bip Roberts, Tony Gwynn, and Roberto
  Alomar, all of whom got on-base at a terrific clip, were pretty fast,
  and rarely cleared the bases with a homer themselves.

  Strat-O-Matic correctly figured that in creating Joe Carter’s card, there
  was no need to make a drastic adjustment for his RBI total. Used in the same
  role as he was used in the majors, batting fourth and fifth for the Padres,
  Carter would be likely to drive in close to 115 runs even with his terrible
  performance. But taken out of that context–in a draft league–Carter’s RBI
  total would be more in line with the rest of his numbers: unacceptable.

  This wasn’t true just with RBIs. Bobby Thigpen wasn’t anything
  special because he had 57 saves; his Strat card mimicked his performance in
  real life. If he was used as the closer for the White Sox in a replay of the
  1990 season, he no doubt would have racked up tons of saves, pitching for a
  team that had lots of small leads to protect in the ninth inning. Taken out
  of that context and evaluated on his own merits, he was no better than any
  other closer with a similar ERA and peripheral numbers. Bob Welch
  might very well have won 27 games in a Strat league had he pitched in front
  of the A’s offense all season, but taken out of that context, he couldn’t
  hold a candle to Roger Clemens‘s better ERA and baserunners-allowed
  totals.

• On-base is king. Strat made a change to the visual appearance of
  the cards beginning with the 1990 card set, which was the first one I saw.
  All results in which the batter reached base safely were in bold print and
  all caps. All outs were left unbolded and in lower case. You can’t play
  Strat for long before realizing that, in essence, only two things can happen
  in an at-bat: either the batter reaches base or he makes an out. Nothing
  drives home the point like seeing WALK on a card with the same rousing font
  as HOMERUN: reaching base safely, regardless of how, is ultimately a victory
  for the hitter. Making an out, no matter how many times you move a runner
  from second base over to third, is ultimately a failure.

I learned other things as well. I almost instantly became more knowledgeable
about individual players. Before I played Strat for the first time, I would
have been hard-pressed to tell you whether Barry Bonds batted from
the left side or right side. I had no idea about lesser players who didn’t
appear regularly on ESPN’s first season of "Baseball Tonight." I
knew that Eddie Murray was a switch-hitter, but Mark Lemke?

By playing Strat, I learned that Cal Ripken grounded into an enormous
number of double plays, and that John Wetteland was easy to steal
bases on. Because every card is separated into performance versus
left-handers and right-handers, I learned that Shane Mack was a
lefty-killer of the highest order, that Roberto Alomar hit better
from the left side than from the right, and that the right-handed Gregg
Olson was actually much tougher on lefties than on right-handed hitters.

I already knew that Wrigley Field was a great hitters’ park and that Dodger
Stadium was a pitcher’s dream. But because Strat incorporated ballpark
effects on homers and singles into the game, I learned that Riverfront
Stadium was one of the most underrated home-run parks in baseball, while
Jack Murphy Stadium was a terrible place to hit for average.

I learned to test my own theories about the game. Having watched numerous
managers devise their own cockamamie lineups, the ones with Darren
Lewis leading off and George Bell batting cleanup, I had the
chance to create my own. I started by selecting the best leadoff hitter and
then moving down the lineup, and suffered through years of dysfunctional
lineups until I learned that you have to start with your #3 and #4 hitters,
and then build the lineup outward from there. I learned that I would have a
tactical advantage by alternating left- and right-handed hitters as much as
possible. The first time I played in a Strat tournament (where they eschewed
the DH), I was completely intimidated by the art of the double-switch.
Eventually I became not only comfortable with the move, but by learning to
slot an outfielder in the #6 or #7 spot and stash a good fourth outfielder
on the bench, I learned to anticipate it before the game even began.

Bill James once wrote–in all seriousness–that every prospective new
manager should be forced to play a simulation game like Strat-O-Matic or
APBA until he mastered it. As silly as that may sound, the fact is that too
many managers get wrapped up in the personal relations of the job (keeping
25 personalities in check, talking to the media, staying on the GM’s good
side) and overlook the personnel relations, continuing to put Joe Hacker and
his .280 OBP in the leadoff spot. A few hundred games of Strat might help
managers take a step back and coldly evaluate their players as resources.
They might come to see the importance of high OBPs at the top of the lineup,
the futility of using a "contact hitter" with no other redeeming
features in the #2 hole, or that the costs of trying to steal with anyone
other than the best base-stealers outweigh the benefits.

Once I had learned things playing simulation baseball on a hardwood table, I
found that this new knowledge made the experience of watching the real thing
more intense. Before, I could watch Tommy Lasorda bring in a lefty to face
Andy Van Slyke and have no idea how that move altered the game. Now,
knowing that Duane Ward didn’t hold runners well, I could see the
added importance of Harold Reynolds working a full count against him.
During the playoffs, I could look at the Braves’ assortment of bench players
and figure out for myself that Bobby Cox, whatever his strengths, was
tactically unprepared for a close game.

A year after I picked up the game, I started my own play-by-mail (now
play-by-Internet) Strat league. Now, instead of drafting cards which were
merely representations of how a player performed, I was drafting the players
themselves; the league was a perennial one, and as in the era of the reserve
clause, once you acquired a player you owned his rights permanently. I was
GM of a franchise, scouting the best minor-league talent, scouring
major-league rosters for overlooked ability, reading the injury reports on
sore-armed pitchers.

In this role, Strat-O-Matic is no different than any other fantasy baseball
league that carries players over from year to year. Except it is. In a
Rotisserie league, the strict salary cap means that the relative value of a
player is as dependent on his salary as on his talent. You might argue that
the lack of salary considerations in a Strat league is hardly a realistic
model of modern baseball economics, but it helps to keep the focus on two
things: how to spot talent, and how to mold that talent into the framework
of a winning baseball team.

Strat players have known for years what major-league teams are only
beginning to discover for themselves: you’re either contending or
rebuilding. If you’re not going to win this year, the best way to win next
year is to move every aging veteran, lock, stock, and barrel, to a
contending team for the best collection of young talent you can extort.
There is no glory in third place, while a pennant flies forever. Better to
win 50 games this year and win 95 games the next than to finish at .500 in
each season.

Strat-O-Matic’s greatest weakness is that, unlike Rotisserie or other
fantasy leagues (and unlike Scoresheet Baseball, the premier in-season
simulation game on the market), Strat is not played in real time. The game
company releases the cards during the winter, after the season is complete
and the statistics are all known. When you sit down to play, you know what
your players can do; you know if you’ve got the .350-hitting Darin
Erstad or the .250-hitting version. This is, of course, completely
unrealistic, and strips the game of one of baseball’s greatest treasures,
its unpredictability.

The saving grace is that, by displaying all the numbers in plain sight and
baring the inner workings of the game, Strat creates a transparency which
gives someone who plays the game the illusion that he is in control of his
own fortune. In real life, it is impossible to know the exact odds of any
given move, because you can never track every single variable that can
affect the outcome: wind speed, the glare of the sun off the third-deck
facade, whether the opposing pitcher has a tweak in his shoulder which he
only feels when he throws his slider, whether the home-plate umpire decides
to expand his strike zone an inch so he can get home early to watch
"ER"…. In Strat, it is theoretically possible to determine the
costs and benefits of any particular decision. You really do know if
pinch-hitting Dave Magadan for Damian Jackson will increase
your chances of tying the game with a single. You know the exact probability
that Eric Young will steal a base successfully.

While on the surface this seems like a flaw in the game’s design, the result
is that by allowing managers to make an informed decision, the consequences
of those decisions provide powerful and instant feedback. It’s easy for Don
Baylor to give Young the green light with Sammy Sosa at the plate and
the wind blowing out at Wrigley, or for Bobby Valentine to let Rey
Ordonez hit with the Mets down by a run in the eighth. There’s no
accountability, because there’s no way to know that Young had a 40% chance
of being thrown out, or that Lenny Harris would have tied the game
with a double.

In Strat, there is. Once you’ve played enough games and had enough
baserunners thrown out trying to steal when they had a 60% chance of
success, you learn to be more conservative. Once you’ve had your
light-hitting shortstop come to bat in the middle of enough ninth-inning
rallies, you learn to pinch-hit aggressively in order to maximize
run-scoring opportunities. More importantly, you learn to stock your bench
with players whose strengths coincide with your starters’ weaknesses, so
that you might have someone to call on to pinch-hit with the game on the
line.

My suspicion is that this transparency is the reason why so many people
continue to play Strat instead of another simulation game like Diamond Mind,
which is a more realistic game run by people who understand sabermetrics.
For all its advantages, when Diamond Mind stopped printing cards in order to
focus exclusively on its computer game, it lost any chance of convincing
long-time Strat players to switch. (Strat-o-Matic’s computer game, while it
adds an impressive array of bells and whistles, is at its heart simply a
faithful re-creation of the card game.) When every conceivable outcome is
printed before you on a three-by-five piece of card stock, it’s easy to
think that you’re in control. When the results are generated behind the
curtain of a computer program, the illusion of control is lost.

In the end, though, it is just an illusion. Perhaps the most important
lesson to be learned from Strat is this: no matter how well you prepare, no
matter how much the odds are in your favor, Chance can sweep in at any
moment and trump your best-laid plans. Not everyone accepts this
gracefully–I’ve lost count of the number of 20-sided dice I’ve seen hurled
out of windows from a great height–but eventually you have no choice but to
accept that there is only so much you can control. Randomness lurks behind
every corner, ready to ambush your team at a moment’s notice.

And this, too, is like real baseball, for in baseball, more than in any
other major sport, the best team doesn’t always win, even against the worst
team in the league. In 2000, every team in the majors won between 40% and
60% of their games. The Chicago Cubs (with the worst record in baseball)
faced the Atlanta Braves (who were tied for the second-best record in the
game) nine times–and won five of the matchups. In a sport where the margin
between greatness and mediocrity is so slim, where the winningest team of
all time can lose in the World Series in six games, talent will only take
you so far.

Rany Jazayerli, M.D., is an author of Baseball Prospectus. His league, the
North American Strat-O-Matic Association, is looking for several expansion
managers for next season. You can contact him at
ranyj@baseballprospectus.com
if you’re interested in joining.