|
|
|
|
Menu |
|
|
____________________
Eric Byrnes We have character. The other guys got lucky. ___________________
John Boles Victim of
illusion ____________________ “
|
The Luck of the 2007 Diamondbacks (Oct-17-07)
Eric Byrnes, the eccentric left fielder of the Arizona Diamondbacks, said this
after losing the first two games of the NLCS: "I think
we're a good team. I also don't think the There is more
than a little irony in those words considering a strong case can be made that
the Diamondbacks were extraordinarily lucky in simply getting to the
play-offs. The notion
that good teams dominate in 1-run games is pretty much nonsense once you
weed out the self-fulfilling prophecy. If you had a team that won 90 games by
being 12 games over .500 in 1-run games, and it was facing another 90-win
team that played only .500 ball in its 1-run games, the smart money is always
– always – going to be on
the team that won 90 without the spectacular record in 1-run games. That's
the team that will have won based more on its ability rather than luck.
And in case you didn't recognize them, I just described When the
window for seeing who is the better team is narrowed down to games decided by
a single run, the dominance - or lack of dominance - of one team over another
has been largely nullified that day, and so the game is more likely to turn
on chance or a lucky break rather than the skills of the team involved.
Please note that I am not saying that good teams don't do better in
1-run games than poor teams. They generally do, but the opportunity to
demonstrate the quality of the teams involved is hugely muted in games
decided by a single run. In the last 20
years about 28% of the games have been decided by a single run. If you wanted
to predict the winning percentage in 1-run games from the win-loss record in
the other 72% of the games it would go something like this. Express the
winning percentage in games decided by more than 1 run by dividing that
winning percentage by .500. Now take the square root of that, and then do it
again. Now convert that relationship back to a context of .500 by multiplying
it by .500. Here would be the example for a team that played .600 ball in games decided by more than 1-run. Divide .600 by
.500 which gives you 1.2. The square root of 1.2 is 1.0954. The square root
of 1.0954 is 1.0466. If you multiply that by .500 you will essentially have
the center of expectation for that team's success in 1-run games. Even though
that team is playing .600 baseball in its other games, it would be expected
to have only a .523 winning percentage in 1-run games. In this light,
we can see that rather than But we know
there are teams that have deviated significantly from the predicted center of
their performance in 1-run games. Doesn't that suggest some teams
have skills and strengths that make them unusually capable of winning one-run
games? And vice versa, that there are teams that are weak in the skills that
would help them win one run games? Hey, That's all
quite true, but when you look closer you start to wonder how much impact
that really has. All teams model each other to some extent, and the degree of
difference in style – separate from ability – is rarely significant. The
truth is that There is a
lack of predictability in win percentage in 1-run games that reflects the
reality that (A) we are talking about a small sample of games, and (B) that sample
is aimed at exactly the games that are most vulnerable to chance overriding
the ability of the teams involved. If an unusual winning percentage in 1-run
games was really a reflection of things like a meaningful differences in a
team's style of play or their character or whatever, you would find
greater predictability in this category from one season to the next. I
have all the necessary team data on this subject going back to 1989, which
covers over 540 team seasons. Here are the five teams that most
exceeded their expected win percentage in 1-run games, along with how they
did in 1-run games the very next year. One-Run Wonders
Very Next Year 1) 2002
Dodgers 33-15 when 24-24 was expected 26-23 when 25-24 was expected 2) 1994
Marlins 23-13 when 17-19 was expected 18-17 when 17-18 was expected 3) 2003
Giants 28-12 when 21-19 was expected 18-25 when 23-20 was expected 4) 2002 5) 1994
Brewers 18-10 when 13-15 was expected
16-17 when 16-17 was expected
Totals 134-64 when 99-99 was expected 103-102 when 105-100 was expected These are the
most extreme examples of teams that seemed to have a special knack for
winning 1-run games and the very next season four of the five are very close
to the expected norm and the one exception actually goes in the opposite direction from their super
success the year before. After being the best team in 1-run games in 2003,
the Giants were one of the poorest 1-run teams in 2004. Now let’s do the
same thing looking at the worst 1-run teams. One-Run Failures
Very Next Year 1) 1999
Royals 11-32 when 21-22 was
expected
21-26 when 23-24 was expected 2) 1994
Phillies 12-26 when 19-19 was
expected
20-23 when 21-22 was expected 3) 2004
Tigers 12-27 when 19-20
was expected
22-26 when 23-25 was expected 4) 1992
Dodgers 17-40 when 28-29 was
expected
27-27 when 27-27 was expected 5) 1989
Red Sox 13-25 when 20-18 was
expected
31-22 when 27-26 was expected Totals 65-150 when 107-108 was
expected 121-124 when
121-124 was expected Could it be
any less predictive? Incidentally,
until the last few weeks of the season And now you
might ask, "So what?" Why put in all the time to study an
issue like this and write thousands of words on such a minor subject? Fairness
and justice rely on clear vision, the truth. A perfect perception of the
truth is rarely possible, but anything that brings us nearer the truth also
brings us nearer to justice. Very real things are influenced by
illusions exactly like this. During my career I routinely did season reviews
for my clients where the primary purpose was to try to break illusions
that would color our strategies for the off-season. I've seen mistakes of
millions and millions of dollars turn on whether illusions like this were
broken or not. Careers were unfairly lost or shortened on such illusions.
Others were unfairly extended to the detriment of the team and to the careers
of other players. Looking that
list of "1-Run Wonders" I am reminded of one team that just missed
the cut, the 2000 Marlins. That 2000 team created a wave of euphoria by
improving 15 wins. Unfortunately that huge improvement was largely built off
a wildly misleading record in 1-run games (32-20 when 25-27 was
expected). The winds of chance blew the other way in 2001, and this time the
team was 10 wins under .500 in 1-run games. John Boles, one of
the most underrated managers of his generation, was fired 48 games into the
2001 season when folks were looking to lynch someone because the Marlins had
gotten worse. The truth is
they hadn’t. They were not as good as they appeared in 2000 nor were they as
bad as they appeared in 2001. When you break the illusion you can see that
Boles had taken over a gutted team that had won only 54 games in 1998, and with
normalized performance in 1-run games the next three seasons should be
evaluated as 66 wins, 72 wins, 81 wins. He didn't
deserve to be fired, never to manage again. He deserved to be celebrated as a
manager who was steadily improving their young team. Without the harm of that
illusion I expect Boles would have managed at least another five years in It matters. |
|
The Diamond Appraised baseball column is dedicated to
Eddie Robinson |
|
